EPA-600/1-77-030
June 1977
Environmental Health Effects Research Series
HEALTH EFFECTS OF NITRATES IN WATER
Health Effects Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconbmic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL HEALTH EFFECTS RE-
SEARCH series. This series describes projects and studies relating to the toler-
ances of man for unhealthful substances or conditions. This work is generally
assessed from a medical viewpoint, including physiological or psychological
studies. In addition to toxicology and other medical specialities, study areas in-
clude biomedical instrumentation and health research techniques utilizing ani-
mals — but always with intended application to human health measures.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/1-77-030
June 1977
HEALTH EFFECTS OF NITRATES IN WATER
by
Hillel I. Shuval
Nachman Gruener
Environmental Health Laboratory
Hebrew University - Hadassah Medical School
Jerusalem, Israel
Grant No. 06-012-3
Project Officer
Leland J. MeCabe
Water Quality Division
Health Effects Research Laboratory
Cincinnati, Ohio 45268
HEALTH EFFECTS RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
DISCLAIMER
The Health Effects Research Laboratory-Cincinnati has reviewed this
report and approved its publication. Approval does not signify that the
contents necessarily reflect the views and policies of the U. S. Environmental
Protection Agency, nor does mention of trade names or commercial.products
constitute endorsement or recommendation for use.
ii
-------
FOREWORD
The United States drinking water standards listed a recommended limit
for nitrate in 1962 because of the occurrence of cases of Methemoglobinemia
in infants that consumed water with a high nitrate concentration. But many
infants were healthy even when their water supply was high in nitrate and
more information was needed to support a specific limit for nitrate in
drinking water.
An opportunity to expand research on the health effects of nitrate oc-
curred when counterpart funds were available in Israel where it was known
that many water supplies had high nitrate concentrations. A rather balanced
research effort has been carried out comprising the development of clinical
chemistry techniques, toxicological studies, and epidemiological surveys.
The first epidemiological survey demonstrated that nitrate in drinking
water did not pose a public health problem because the water was not con-
sumed by the infants, but subsequently a study population was obtained in
the Gaza strip where the infants were exposed to the water with an excessive
nitrate concentration.
The nitrite concentration of drinking water is low or nonexistant but
this ion was used in most of the toxicology studies because the animal model
did not allow for the nitrate-to-nitrite conversion in the gut. Thus the
findings may be more applicable to the food preservation problem than to
drinking water.
Many interesting findings suggest further research.
Leland J. McCabe
Water Quality Division
Health Effects Research Laboratory
iii
-------
ABSTRACT
This report represents the results of a series of field and laboratory
studies designed to evaluate the health effects of nitrates in drinking
water.
The results of the epidemiological studies indicate that infants con-
suming appreciable amounts of water high in nitrates in the form of powdered
milk formula show significantly raised Methemoglobin levels. This is also
true for infants consuming tap water having a nitrate concentration ranging
from 45-55 ppm nitrate. It is felt that this latter finding provides direct
epidemiological evidence in support of the current nitrate standard in drink-
ing water of 45 ppm. Other field studies showed that even breast-fed infants
or those receiving pasteurized milk can consume under Israel conditions con-
siderable amounts of tap.water as liquid supplements during the hottest months
of the year.
Laboratory studies on the acute and chronic toxic effects of nitrites
indicate among others, that nitrites can pass the rat's placenta and cause
raised Methemoglobin levels in the fetus; that pregnant rats are particularly
sensitive to exposure to nitrites, and that pups born to dams exposed to
nitrites during gestation show poor growth and development; that rats chroni-
cally exposed to sodium nitrate and sodium nitrite in their drinking water for
18 months show distinct deviations in heart blood vessels even at the level
of 200 ppm of NaNO_. Exposure of mice to nitrites in drinking water causes
behavioral effects such as lowered motor activity and an increase in isola-
tion induced aggression. A number of sensitive analytical micro-methods re-
quired for these studies were developed.
The results of the epidemiological and the toxicological studies provide
little basis for a liberalization of the current drinking water standard for
nitrates. If anything, evidence is presented which may raise some questions
as to whether the current standard provides a sufficient margin of safety
below the detectable effect level.
This report was submitted in fulfillment of Research Grant No. 06-012-3
by the Environmental Health Laboratory of the Hebrew University of Jerusalem
under the partial sponsorship of the Environmental Protection Agency. Work
was completed as of October 1973.
iv
-------
CONTENTS
Foreword
Abstract iv
Figures vi
Tables viii
Acknowledgment xii
I. Conclusions 1
II. Recommendations 4
III. Introduction 5
IV. Determination of Methemoglobin in Blood 12
V. Determination of Nitrite in Blood 17
VI. Determination of Nitrate in Water 26
VII. Measurement of Ascorbic Acid in Blood 37
VIII. Epidemiological Study - Rehovot Area 40
IX. Epidemiological Study - Gaza Area 50
X. A Controlled Hospital Study 59
XI. Survey of Liquid Intake in Infants 64
XII. Development of Methemoglobin Reductase 69
XIII. Methemoglobinemia Induced by Transplacental
Passage of Nitrites 77
XIV. Effects of Nitrites on Pregnant Rats and
their Newborn 82
XV. Influence of Ascorbic Acid on Methemoglobinemia 85
XVI. Chronic Toxicity of Nitrates and Nitrites 93
XVII. The Reduction of Methemoglobin in the
Human Erythrocyte Ill
XVIII. Changes in Motor Activity of Mice given Nitrites 118
XIX. Effect of Nitrites on Isolation - Induced
Agression 128
XX. Effect of Nitrites on E.E.G. of Rats 132
Appendicies
I. Questionnaire 144
II. List of Publications 147
-------
FIGURES
V - 1 Influence of Deproteinizing Reagents on Nitrite
Recovery 19
2 Standard Curves for Nitrate Determination 20
3 Effect of Diazotization Time and Coupling Time 21
VIII - 1 Nitrate Content of Pumped Water in Central Israel 41
IX - 1 Methemoglobin Levies in Infants by Age - Gaza 51
2 Methemoglobin Levels in Infants on Different Milk
Regimes - Gaza ' 55
XI - 1 Total Liquid and Water Intake in Infants 1-5 Months Old
According to Months of the Year 65
XII - 1 Changes in Methemoglobin Reductase Levels with
Age - Humans : 73
2 Changes in Methemoglobin Reductase Levels with
Age - Rats 74
XIII - 1 Kinetics of Nitrite and Methemoglobin in Blood of a
Pregnant Rat and Fetuses 79
XV - 1 Mean Ascorbic Acit, Methemoglobin and Nitrite Levels in
Six Rats Administered NaNO- 87
>
,^
2 Mean Difference Between Experimental and Control Animals in
Methemoglobin Levels 88
3 Mean Difference Between Experimental and Control Animals in
Nitrite Levels 89
XVI - 1 Methemoglobin Reduction Rate in Rats 103
2 Heart-Control 18 Month Old Rat 107
3 Heart, Rat, after 18 Months of Drinking Water Containing
1000 ppm NaNO. 107
vi
-------
XVII - 1 Effect of MHb Concentration on Methemoglobin
Reductase Activity 112
2 Effect of the MHb Concentration on Methemoglobin
Reductase Activity 113
3 Changes in Enzyme Activity After Exposure to MHb 113
4 The Inhibitory Effects of Several Anions on
Methemoglobin Reductase Activity 115
XVIII - 1 Barrier Activity Box 121
2 Relationship Between Motor Activity and Methemoglobin
Level in Mice 124
3 Mean Difference Between Experimental and Control
Animals 126
XX - 1 E.E.G. of Control - Group A 134
2 Change in Background Activity After 2 1/2 Weeks of
NaNCL Group B 134
3 Appearance of General Paroxysmal Outbursts(Group B) . . . . 136
4 Background Activity (Group B) 136
5 Appearance of Slowed Background Activity (Group C) 133
6 Background Activity (Group C) . 138
7 Background Activity (Group D) 140
8 Slow Background Activity (Group D) 140
vii
-------
TABLES
Page
IV - 1 Precision of Mehtod 14
2 Recovery of Methemoglobin Added to Hemolysates 14
3 Stability of Methemoglobin with Time Under
Different Conditions 14
4 Oxidation of Hemoglobin During Storage at - 20°C 15
5 The Effect of Storage at Low Temperatures on the
Oxidation of Hemoglobin 15
V - 1 Precision of the Method for Nitrite Determination
in Blood 23
2 Recovery of Nitrite from Blood Plasma 23
3 Residual Nitrite Partition in Blood and Plasma 24
VI - 1 Determination of Nitrate by Selective Electrode Method
at Various Ionic Strengths 30
2 Anions Interference in Nitrate-Nitrogen Determination by
Specific Ion Electrode 30
3 Comparison of Specific Ion Electrode and Phenoldisulphonic
Acid Method for Nitrate-Nitrogen Determination in Drinking
Water Samples 33
4 The Mean Differences Between the Results 33
VII - 1 Human Ascorbic Acid Levels 38
VIII - 1 Methemoglobin in Infants in Areas with High and Low
Nitrate concentrations in Drinking Water 43
2 Distribution of Hemoglobin Levels Amoung Infants 43
3 Methemoglobin in Intants Drinking Powdered Milk Formula
and Only Other Forms of Milk 45
viii
-------
Page
VIII - 4 Methemoglobin in Infants With and Without Citrus
or Tomato Juice in Diet 45
5 Methemoglobin in Infants With or Without Diarrhea 45
6 Methemoglobin as a Function of Age 46
7 Methemoglobin by Present Weight 46
IX - 1 Methemoglobin Levels in Infants by Age 52
2 Mean Methemoglobin Levels in Infants on
Different Milk Regimes 53
3 Distribution of Methemoglobin Levels in Infants Exposed
to High and Low Concentrations of Nitrates-in Drinking
Water 56
X - 1 Methemoglobin Levels in Infants Exposed to High and
Low Levels of Nitrates 60
2 Distribution of .-Changes in Methemoglobin Levels Between
Days of Varying Exposure to Nitrates in Water 60
XI - 1 Liquid Intake Infant Age 1-5 Months 65
2 Liquid Intake and Tap Water Intake in Infants According
to Month, of Year 66
3 Liquid Intake injtafants According to Ethnic Group .... 67
4 Liquid Intake in Infants According to Nutritional
Regime 67
XII - 1 Level of Methemoglobin Reductase Activity in Humans
at Different Ages 71
2 Methemoglobin Reductase Activity in Rats as a Function
of Age 72
XIII - 1 Blood Nijirite and Methemoglobin Levels After Injection
of Different Doses of Sodium Nitrite 78
ix
-------
Page
XIII - 2 Methemoglobin Reductase Activity in the Foetus and
the Pregnant Female Rat and Humans 80
XIV - 1 Distribution of Hemoglobin Determination in Pregnant
Rats Chronically Exposed to Sodium Nitrite in Drinking
Water 82
2 Pyruvate/Lactate Ratio in Pregnant Rats Chronically
Exposed to Sodium Nitrite in Drinking Water 83
XVI - 1 Liquid Intake of Rats Drinking Sodium Nitrite and
Nitrate Solutions 96
2 Changes in Body Weight in Rats Drinking Sodium Nitrite
and Nitrate 98
3 Nitrite and Nitrate Intake in Chronically Exposed
Rats 99
4 Methemoglobin Levels of Rats Drinking Sodium Nitrite
and Nitrate Solution 100
5 Nitrate Level in Rat Blood Exposed to Sodium Nitrite
and Nitrate 101
6 Methemoglobin in Rats After a Single Dose of
NaN02 - I.P 101
>
7 Methemoglobin Reductase Activity in Rats Exposed to
NaN02 and NaNO 102
8 Diphosphoglycerate Level in Erythrocytes of Rats Exposed
to NaN02 and NaNO 105
9 Glutamate Levels in Blood of Rats 105
10 Death of Unknown Origin Among Rats Treated with Sodium
Nitrite and Nitrate 105
11 Effect of Sodium Nitrite and Nitrate on Coronary Blood
Vessels of Rats 106
-------
Page
XVIII - 1 Relationship Between Water Intake in Mice and
Exposure to NaNCL 120
2 Comparison of Mean Variations in Motor Activity
and MetHb 123
3 Relation Between MetHb and Motor Activity in Mice Consuming
NaN02 125
XIX - 1 The Effect of Sodium Nitrite on Aggression in Mice . . . 129
2 Aggression in Mice After Returning to Tap Water 130
XX - 1 Relationship Between Mean Frequency of Background Brain
Electrical Activity, Amplitude and MetHB in Rats . . . 135
xi
-------
ACKNOWLEDGMENTS
This project involved extensive field work and could not have been suc-
cessfully carried out without the active support and cooperation of the staff
and field personnel of the Ministry of Health; the Hillel Joffa Hospital;
Kupat Holim-Sick Fund of the General Labor Federation; The Public Health
Department of the City of Jerusalem and the Health Services of the Gaza Area.
It would not be possible to list here the dozens of doctors and nurses in the
above organizations who contributed to the project each in his own way. Their
fine efforts are gratefully acknowledged.
Many individuals participated directly in the work of the project staff,
some during the entire three-year period, others for shorter periods. However
special mention must be made in particular of the devoted work of Mrs. Sara
Cohen, M.Sc., who served as a key member of the project scientific staff for
the entire period. Each one of the following contributed devotedly to this
true team effort:
H. Adan - Technician A. Ram - M.Sc.
K. Behroozi M.P.H. S. Scharf ~ B.Sc.
S. Cohen M.Sc. D. Scheuermann - R.N.
M. Dalit - M.Sc. H. Shechter - Ph.D.
P. Ever-Hadani - M.P.H. T. Shklanka - B.Sc.
A. Friedberg - R.N. D. Starkman M.Sc.
M. Gutnik - Ph.D. N. Tamir - M.Sc.
I. Ivriani - M.Sc. A. Tysman - M.Sc.
A. Metuki - B.Sc. R. Toeplitz M.P.H.
In addition to the project staff a number of scientists served as consult-
ants or supervised* specific studies. We wish to thank them for their valuable
contributions:
Prof. M. Davies - M.D. E. Hegesh - Ph.D.
Prof. L. Gordis - M.D. S. Robinson - M.D.
Prof. G. Izak - M.D. R. Yarom - M.D.
R. Guttman - Ph.D.
Without the cooperative efforts of all those who worked devotedly on
the various aspects of this complex project it could not have been success-
fully carried out.
xii
-------
SECTION I
CONCLUSIONS
1. The analytical methods developed in this project provide sensitive and
reliable micromethods for epidemiological studies in which methemoglobin,
nitrites and ascorbic acid levels in blood must be determined from samples
taken in the field and brought to the laboratory for later analysis.
2. Evidence is provided that the ion selective electrode method devel-
oped to determine nitrate concentrations in water can lead to considerable
inaccuracies when testing highly mineralized water. Chlorides can, in particu-
lar, lead to inaccurate results. The ion selective electrode method can at
best be considered a screening test under field conditions and then only for
water with nitrate concentrations over 45 ppm (as N03).
3. The results of the two epidemicIgical studies which encompassed 2891
infants up to 24 months of age indicate that there is a relationship between
concentration of nitrates in drinking water consumed mainly as powdered milk
formula and raised methemoglobin (MetHb) levels. The effect was detectable and
significant even in the group of infants exposed to water containing 45-55 ppm
of N03- Even though no clinical cases of methemoglobinemia were detected, it
is felt that the appearance of a significant increase above the normal MetHb
levels in infants when exposed to water with nitrate concentrations slightly
above the current standard of 45 ppm is sufficient to provide direct epidemiolo-
gical support for the current standard. The health significance of such sub-
clinical levels of MetHb is unclear and it is still unknown why only a small
number of infants exposed to such levels of nitrates in water develop clinical
cases of the disease.
4. Methemoglobin is higher in the first three months of life and among
infants with gastrointestinal disturbances irrespective of nitrate intake.
5. The results of the carefully controlled study of infants consuming
nitrates in their formula in a hospital ward provided results very similar to
those of the field studies. However, some evidence indicating an adaptive
mechanism was obtained. This is not yet understood.
6. A study of liquid intake among 104 infants age 1-5 months indicated
that while during the cool months 90% of the total liquid intake is made up
of milk, as much as 50% can be in the form of tap water supplements during
the hottest month. This finding may lead to considerable intake of nitrates
from tap water even among infants not receiving tap water as part of their
-------
powdered milk formula. This finding is in apparent conflict with the
findings that even in high nitrate areas infants receiving no powdered milk
formula in their diet had no higher levels of MetHb than those in low nitrate
areas. No seasonal pattern of MetHb levels was determined. The specific
role that powdered milk formula possibly plays in the development of raised
MetHb in high nitrate areas still remains a mute point.
7. The low level of MetHb reductase (M.R.) in infants at birth and for
the first few months of life may provide a partial explanation of the particu-
lar susceptibility of this age group. The role that partial M.R. deficiencies
of a genetic form play in the development of clinical disease when exposed
to nitrates in water is still an open question worth exploring.
8. The passage of nitrites from pregnant rats to the fetus through the
placenta with the 'development of raised MetHb in the fetus opens the question
as to whether only infants up to six months of age should be provided with
low nitrate water in areas where the nitrate concentrations are high in the
general water supply? This phenomenon was detected even at very low nitrite
doses to the pregnant rat, (2.5 mg/kg of sodium nitrite). That is not much
above doses that could occur under not too extreme environmental conditions.
The potential significance of this finding to humans is emphasized by the
fact that the MetHb reductase levels in the human fetus is only one-tenth
of that found in newborn rats.
9. The higher sensitivity of pregnant rats to nitrites and the poor
growth and development of pups born to dams exposed to nitrites during
gestation suggests that a careful evaluation of the health effects of expo-
sure to nitrites and nitrates during pregnancy is required.
10. Ascorbic acid was demonstrated to provide partial protection against
methemoglobinemia in rats. The mode of action of ascorbic acid is still
unclear.
11. The studies involving chronic exposure of rats to various concen-
trations of sodium nitrate and nitrite in drinking water indicate that even
at very high dose levels few gross effects are detectable. However, clear
deviations in heart blood vessels thickness were discernible after 12 months
and became clear cut after 18months of exposure to both nitrates and nitrites.
The thin balloon-like blood vessels not typical in rats of this age were
clearly detectable even at a dose of 200 mg/1 of sodium nitrite. The patho-
logical significance for humans of this finding is not clear. It appears
that this effect may be due to direct toxicity of the nitrites rather than
as a side effect of methemoglobinemia since in the group of rats exposed to
nitrates no raised MetHb was detectable.
12. Evidence of an adaptive mechanism to nitrite exposure was detected
in the chronic studies. DPG levels increase within 2 days of exposure to
nitrites and remain elevated throughout the period of exposure. The effect
of this is to increase the oxygen release from blood to the tissues.
-------
13. The behavioral studies also indicate the possibility that nitrites
have a direct toxic effect since the lowered levels of motor activity in mice
and increased isolation-induced aggression occurred even in mice showing only
very slight increases in MetHb. While the reduced motor activity might be
associated with the known quality of nitrites as muscle relaxants, the
increased aggression suggests a more central effect. While the studies on
the effects of nitrites on brain electrical activity of rats remain incon-
clusive, the recent findings of Russian researchers in this area suggests
that there is more to this matter than we have been able to clearly confirm.
14. Finally, it must be pointed out that from both the field and
laboratory studies evidence has been gathered that nitrates and nitrites
may be more toxic than generally considered. The fact that significant
effect can be detected in infants consuming water having only slightly more
nitrates than the current standard raises the question as to the margin of
safety provided by that standard. The other toxic effects revealed in this
study which may be independent of the methemoglobin increases resulting
from exposure to nitrates raises many questions as to the proper basis for
establishing the nitrate standard in drinking water which has until now
been based on the possibility of infant methemoglobinemia alone. Certainly
in light of this information there can be little grounds for relaxing the
current standard, particularly in those areas where infants consume consid-
erable amounts of tap water in the form of milk powder formula.
-------
SECTION II
RECOMMENDATIONS
While the nitrate standard for drinking water has been based originally
on the association with infant methemoglobinemia, a number of findings in this
study point to possible direct toxic effects not previously considered. The
full significance of the pathology in heart blood vessels in rats exposed to
nitrates and nitrites for 18 months should be thoroughly investigated both
in the laboratory and in the field. The behavioral changes in mice such as
reduced motor activitiy and increased isolation induced-aggression, also not
directly associated with raised levels of MetHb, should be studied further.
A complete rechecking of the findings of changes in the E.E.G. in rats exposed
to nitrites in water should be urgently made.
The high degree of sensitivity of pregnant rats to methemoglobin induc-
ing agents should be evaluated, since to date only infants up to 6 months in
age have been considered at risk. The question of transplacental passage
of nitrites and raised MetHb in the fetus of rats requires further evaluation
as to its significance to humans.
There is still much to be learned about the association of nitrates
in drinking water and raised MetHb levels in infants. What role does the
powdered milk formula actually play in the development of MetHb? How
effective are vitamin C rich foods such as citrus juice and tomato juice
in preventing raised MetHb? What is their mode of action? Why do only a
small percent of infants exposed to high nitrates in water develop clinical
cases of methemoglobinemia? A fuller understanding of these questions is
still required.
However, despite the many questions that remain unanswered it is
apparent that nitrates and nitrites are potentially more toxic than gener-
ally assumed. The information gained from this project provides a basis
for recommending that no relaxation be made in the nitrate standard for
drinking water at this time despite the infrequency of clinical cases of
the disease. Other direct toxic effects of nitrates may prove to be of
equal or greater importance than the problem of infant methemoglobinemia.
Until those questions are fully elucidated the standard would best be main-
tained as is while being kept under constant scrutiny and review.
-------
SECTION III
INTRODUCTION
GENERAL
The standard for nitrates in drinking water was initially established
based on limited epidemiological evidence that indicated that no cases of
infant methemoglobinemia occurred in areas with less than 45 ppm of nitrates
(as NOj) in the water. Since the standard was established there has been
considerable controversy on the subject. Some European researchers have
reported that they were able to detect raised methemoglobin levels in so-
called normal infants in areas where occasional clinical cases of the disease
were reported. In addition, clinical disease was also reported among infants
exposed to water having less than 45 ppm of nitrates. Suggestions that a
stricter standard be enforced have been made as a result of such studies.
On the other hand, extensive areas in the United States supplying water
showing nitrate levels above the standard have reported little or no clinical
cases of the diseases. Since nitrates are difficult to remove from water by
economically feasible means, moves have been made for a more liberal standard
reflecting the lack of clinical disease in such areas.
Since the epidemiological and toxicological base for establishing the
nitrate standard was relatively limited the project reported upon here was
initiated to clarify some of the basic questions concerned with the standard
and provide a basis for confirming it or changing it as need be. What follows
is a review of some of the general aspects and known pathogenesis of methemo-
globinemia which served as the basic consideration for establishing the
standard initially.
-------
GENERAL ASPECTS OF METHEMOGLOBINEMIA
Methemoglobinemia can be caused by several chemicals such as nitrates,
perchlorates , aminophenols, anilin, sulfonamides and others (1,2). Various
endogenous forms of the disease are known. Nitrates do not directly convert
hemoglobin to methemoglobin but can be converted to nitrites by intestinal
microflora, with subsequent formation of methemoglobin. Hemoglobin (Hb) is
the oxygen carrier of the blood. It is a protein of molecular weight 68,000,
consisting of four identical sub-units, each containing a polypeptide chain
(globin) and a heme group. Every heme contains at its center an atom of iron
which, in the oxygenated form of Hb, oxyhemoglobin (HbO?) , is in the bivalent
state (Fe2+ ferrous) . Methemoglobin (MetHb) is the oxidized product of Hb in
which the iron is in the trivalent state (ferric
On transition from the ferrous to the ferric state the Hb loses its
ability to combine with 02- The exact molecular mechanism of the conversion
of Hb to MetHb is still obscure.
Conversion of Hb to MetHb takes place all the time in the body, but the
quantity of the latter is maintained at a low, steady- state level mainly by
the action of an enzymatic system (3) . Several enzymes with MetHb reduction
capability have been purified, but the exact mechanism by which they operate
in vivo has not yet been resolved. A direct; nonenzymatic reduction of
methemoglobin is carried out by glutathion or ascorbic acid(4) . The normal
concentration of methemoglobin is still a matter of dispute.
Methemoglobinemia constitutes a potential impairment of the proper supply
of oxygen to the tissues. For example, it was found that in trained subjects
who underwent work tests, 10-12% MetHb resulted in impaired oxygenation of
muscles (5). This phenomenon is a result of both less hemoglobin and the
greater affinity of the residual Hb for oxygen.
PATHOGENESIS OF METHEMOGLOBINEMIA
It is accepted by many workers that under normal circumstances that about
1% of the total hemoglobin exists as MetHb (6, 7) . No external signs or symptoms
are generally noted under 5%. The first signs of cyanosis can be seen between
5 and 10% (8).
The presence of high concentrations of nitrates in water is the principal
determinant of the occurrence of methemoglobinemia in infants; however, it is
not the only one (9). Other factors important in the pathogenesis of the disease
are:
Age: Most of the cases of nitrate methemoglobinemia occur in infants
below one year of age. In a review of 146 cases of methemoglobinemia
in Minnesota, 90% of them were found to have occurred by the age of
eight weeks. The youngest case was seven days and the oldest five
months (10). Schmidt and Knotek(ll) reported on a survey carried out
in Czechoslovakia, in which 52% of the infants, 0-3 months old, in a
high-nitrate area had elevated methemoglobin levels; in the age group
from 3-12 months old the percentage reached 13. There are also reports
-------
on elevated MetHb levels in older children and adults who had consumed
water with high nitrate concentrations but showed no symptoms of the
disease(12). In a Russian study(13), of 800 children in day nurseries, it
was found that 92.2% of them who ingested water which contained 20-40 mg
N03/1 had elevated MetHb levels. In 50% of the cases the level was
higher than 5%. Previously it was found that 8 mg/1 nitrate did not
raise the methemoglobin level(13).
Presence of bacteria; Cornblath and Hartman(14) emphasized the import-
ant role of bacteria in the production of methemoglobinemia. They
studied the gastrointestinal flora with regard to nitrate reduction,
gastric acidity, age and the level of intestinal absorption of nitrate.
It was found that all micro-organisms isolated from the mouth and from
the gastrointestinal tract were capable of reducing nitrate to nitrite
and grew in media of pH 5-7.
The invading bacteria must adapt themselves to nitrite formation
if they were not previously in contact with nitrate. Consequently,
they felt it may take from four to five days after the first ingestion
of nitrate until the full nitrate-reducing efficiency is reached. This
may explain the commonly observed lag period of 1-3 weeks before the
onset of the illness(12). However, in our own studies we found raised
MetHb in infants 24 hours after exposure to powdered milk formula made
with high nitrate water.
Gastric acidity: Examinations of gastric juice of infants who devel-
oped appreciable levels of MetHb revealed that the pH was usually higher
than 4. It was found(15) that the normal pH of infant gastric juice
varied between 2-5, but with unspecific diarrhea, gastric pH increased
and ranged between 4.6 and 6.5; Mucha(16) examined the pH and bacterial
flora of gastric juice from children. They found the gastric juice
was sterile at pH 4.6.
Gastrointestinal disturbances: It has been shown that all the members
of the family Enterobacteriaceae are able to reduce nitrate to nitrite.
Such organisms can gain access to the upper intestine during gastro-
intestinal disturbances(14). However, their ability to become estab-
lished in the stomach is dependent on the pH, as mentioned in the pre-
ceding section. In the absence of nitrate-reducing bacteria in the
stomach or upper intestine, most of the nitrate is probably absorbed
as nitrate before reaching the colon in which the nitrate-reducing
bacteria are normally found.
Type of powdered milk product: Studies in Czechoslovakia(8,ll) indi-
cated that the use of certain types of milk preparations has been sus-
pected as the main cause for the development of methemoglobinemia in
areas with high levels of nitrates in water. They reported on cases
of methemoglobinemia in infants due to feeding with various brands of
regular powdered milk which contained spores of B. subtilis, a nitrate-
reducing bacteria; acidified milk powders which are often prepared by
-------
fermentation with St. lactis did not cause any disease. The acidi-
fied milk preparation can contain an antibiotic substance called nisin
which inhibits the growth of nitrate-reducing bacteria. On the other
hand, Mucha et al(16) claimed that the main source of nitrate-reducing
microorganisms could be eliminated by providing bacteria-free food to
infants; however, under certain conditions, the ascent of bacteria from
the colon to the duodenum and stomach cannot be prevented. B. subtilis
spores are, however, not destroyed by normal milk pasteurization and
drying processes. Little nitrate was found in the milk of cows drink-
ing water with up to 800 mg N03/l(17). Many hold the opinion that
mother's milk or cow's milk cannot be a cause of methemoglobinemia.
High fluid intake; Infants with an average fluid intake would ingest
several times more nitrate per gram of hemoglobin than an adult due
to their higher fluid intake per unit of body weight (7) . Burden (1 8) ,
assuming that all nitrate is reduced to nitrite, sets the permissible
level of nitrate for adults at 1,056 mg/1 in England and 198 mg/1
in the tropics; for infants the permissible levels would be 88 mg/1
and 26 mg/1 respectively. His estimates are based on 13.2 mgNOs/kg
as the maximum daily amount which can be tolerated without giving rise
to toxic symptoms. The governing factor would be, in his estimate, the
relative daily fluid intake which for infants would be 0.5 liter in
England and 2.0 liters in the tropics.
Effect of nutrition: Food composition can also affect the severity of
the illness. On the one hand, there are certain nutrients such as
vitamin C that can cure or prevent methemoglobinemia. High vitamin C
intake among infants in some areas may explain the scarcity of the
disease even when waters rich in nitrates are consumed. On the other
hand, certain vegetables such as spinach, rhubarb, etc., contain con-
siderable amounts of nitrates. Several cases of nitrate poisoning
in infants after eating spinach were reported (19) . In our own studies
in Jerusalem we detected as much as 4850 ppm of N03 and 233 ppm of
N02 in a local variety of spinach (20) . Such concentrations in food
can become particularly toxic if exogenic bacterial activity converts
most of the N03 to
Fetal hemoglobin; Hemoglobin F is oxidized more readily to methemo-
globin(21) . The fact that blood of newborn babies consists of more than
80% hemoglobin F might explain their increased tendency to develop
methemoglobinemia .
Methemoglobin reduction: Methemoglobin reduction velocity in the
presence of lactate or glucose is lower in cord blood erythrocytes than
in adult blood. The lower methemoglobin reduction velocity in cord
blood is explained by a temporary deficiency of DPNH - the methemo-
globin reductase cofactor(22) or by low activity of the enzyme itself.
A positive correlation between DPNH diaphorase activity and methemo-
globin reduction was shown in adult blood but not detected in cord
blood(23) .
-------
OBJECTIVES
The following general questions served as guidelines in carrying out
the project objectives of evaluating the suitability of the current nitrate
standard in drinking water.
1. Can a dose-response relationship be established between intake
of nitrates in drinking water and the development of raised meth-
emoglobin levels in infants?
2. At what threshold level of nitrates in water is the first effect
detectable?
3. What environmental, nutritional, physiological or genetic factors
influence this relationship?
4. What is the health significance, if any, of chronic sub-clinical
or slightly raised levels of MetHb?
5. Can new sensitive parameters be used to detect health effects due
to exposure to nitrates other than raised MetHb levels?
6. Are there direct toxic effects of exposure to nitrates and/or
nitrites other than raised MetHb?
PROGRAM
In order to answer the above questions an extensive series of field
and laboratory studies was initiated. The field work included studies of
some 3000 infants in various parts of the country exposed to various levels
of nitrates in water and on various dietary regimes.
Laboratory studies on various aspects of acute and chronic toxicity
were carried out in parallel. A number of special investigations on sensi-
tive methods of evaluating toxic effects were initiated including behavioral
and neurobiological effects.
-------
REFERENCES
1. Bodansky, 0., Methemoglobinemia and Methemoglobin-producing Compounds,
Pharmacol. Rev., 3, 144, 1951.
2. Lee, D.H.K., Nitrates, Nitrites, and Methemoglobinemia. Environ. Res.
3:484-511, 1970.
3. Jaffe, E.R., The Red Blood Cell (C.W. Bishop, ed.), p. 397, Academic
Press, New York, 1964.
4. Lemberg, R. and Legge, J.W., Hematin Compounds and Bile Pigments, p. 228.
Interscience Publishers, Inc., New York, 1949.
5. Tepperman, J., Bodansky, 0. and Jandorf, B.J., Effect of Paraminopro-
piophenone-induced Methemoglobinemia on Oxygenation of Working Muscles
in Human Subjects, Am. J. Physiol., 146, 702, 1946.
6. Ferrant, M., Methemoglobinemia: Two Cases in Newborn Infants Caused
by Nitrates in Well Water, J. Pediat., 29, 585, 1946.
7. Gross, E., Arch., Hyg., 148, 288, 1964.
8. Knotek, Z., and Schmidt, P., Pathogenesis, Incidence and Possibilities
of Preventing Alimentary Nitrate Methemoglobinemia in Infants, Pediatrics,
34, 78, 1964.
9. Winton, E.F., Tardiff, R.G., and McCabe, L.J., Nitrate in Drinking
Water. Jour. Amer. Water Works Assoc. 65:95-98, 1970.
\
10. Rosenfield, A.B. and Huston, R., Infant Methemoglobinemia in Minnesota
due to Nitrates in Well-water, Minn. Med. 33, 787, 1950.
11. • Schmidt, P. and Knotek, Z., Problems of Nitrate Food Methemoglobinemia
of Infants in Czechoslovakia. Gig, i San., 31, 290, 1966.
12. Teerhaag, L. and Eyer, A., Nitrate in Trinkwasser, Off. Gesundheits-
dienst, 20, 1, 1958.
13. Subbotin, F.N., The Nitrates of Drinking Water and Their Effect on
the Formation of Methemoglobin, Gig, i San., 2, 13, 1961.
14. Cornblath, M. and Hartman, A.F., Methemoglobinemia in Young Infants,
J. Pediatrics, 33, 421, 1948.
15. Marriott, W.M., Hartman, A.F. and Senn, M.J., Observations on the
Nature and Treatment of Diarrhea and the Associated Systemic Distur-
bances. J. Pediatrics, 3, 181, 1933.
16. Mucha, V., Kamensky, P. and Keleti, J., Genesis and Prevention of Ali-
mentary Nitrate Methemoglobinemia in Babies, Hygiene and Sanitation,
30, 185, 1965.
10
-------
17. Davison, K., Nitrate Toxicity in Dairy Heifers, Dairy Science, 47,
1065.
18. Burden, E.H., The Analyst, 86, 429, 1961.
19. Sinios, A. and Wodsak, W., Die Spinatvergiftung des Sauglins, Dtsch.
med. Wschr., 90, 1856, 1965.
20. Eisenberg, A., Wisenberg, E., and Shuval, H.I., The Public Health
Significance of Nitrate and Nitrite in Food Products. Ministry of
Health, Jerusalem, 1970.
21. Boi-Doku, F.S. and Pick, C., Spontaneous Oxidation of the Chain of
the Foetal Haemoglobin Component of Cord Blood, Biochim Biophys. Acta
115, 495, 1966.
22. Ross, J.D. and Desforges, J.F., Reduction of Methemoglobin by Erthro-
cytes from Cord Blood, Pediatrics, 23, 718, 1959.
23. Kamazawa, Y., Hattori, M., Kosaka, K. and Nakao, K., The Relationship
of NADH-dependent Diaphorase Activity and Methemoglobin Reduction in
Human Erythrocytes, Clin. Chim. Acta, 19, 524, 1968.
11
-------
SECTION IV
MICROMETHOD FOR THE DETERMINATION OF METHEMOGLOBIN IN BLOOD
INTRODUCTION
In connection with the study to determine methemoglobin (MetHb) levels
in infants exposed to high nitrate content in water the need arose
for a precise and accurate micromethod which would assure adequate stability
of the sample to allow a field survey to be conducted in communities distant
from the laboratory.
The method of Evelyn and Malloy(l) , which up to date has been the
standard procedure for the assay of MetHb in blood, is considered not suffi-
ciently sensitive to determine low concentrations of the pigment, present in
normal blood samples and to differentiate between normal and slightly ele-
vated levels which might be indicative of subclinical cases of methemoglo-
binemia caused by nitrate ingestion. Due to the instability of the MetHb
in the drawn blood sample it is usually recommended in normal clinical proce-
dures to carry out the test as rapidly as possible. This is generally not
feasible under field survey conditions.
The procedure presented below is a modification of the Evelyn and
Malloy method aimed to augment its precision. The size of the blood sample
is kept at a minimum so that finger-tip blood samples from infnats should
suffice. The procedure also assures sample stability.
PRINCIPLE
MetHb is reacted with cyanide and the change in light absorption at
632 nm is measured.
REAGENTS
1. K3Fe(CN)6-5% (w/v) in water. Store in a dark bottle. Prepare monthly.
2. Phosphate buffer - 0.5 M, pH 6.6.
3. Ferricyanide-phosphate mixture: 0.2 ml of KsFeCCN)^ solution
(reagent 1) and 2.5 ml of phosphate buffer (reagent 2) are made up
10 to ml with water. Prepare daily.
4. Sodium cyanide- 10% (w/v). Prepare monthly.
5. Acetic acid-12% (w/v).
6. Neutralized cyanide solution. Mix 1.0 ml of reagent 4 with 0.9 ml
of reagent 5. Use within 6 h.
12
-------
INSTRUMENTS
1. Highspeed microcentrifuge, capacity 10,000 rev./min. The Sorvall
SS-1 centrifuge or the Eppendorf Microfuge with the corresponding
1.5-2 ml-microtubes were used.
2. Microuvettes with a path length of 1 cm, inside diameter 3-4 mm
and a minimal working capacity of 0.6 ml. The Zeiss MT4 cuvettes
were used.
3. A Zeiss PMQ II or another equivalent sensitive spectrophotometer
adaptable for the use of microcuvettes.
METHOD
Pipette 200 yl of freshly drawn, heparinized blood into a 1.5-2.0 ml
micro centrifuge tube and hemolyse by adding 550 yl of water. Mix. After
3 min, add 250 yl of phosphate buffer (reagent 2) and mix again. Unless
test is completed within 30 min, cool hemolysed and buffered sample rapidly
to 2° and keep refrigerated until assayed. Centrifuge at 10,000 rev./min
for 15 min. Centrifugation at high speed is absolutely necessary. Trans-
fer the completely clear supernatant into a small tube to the first (1) of
two microcuvettes. Transfer to the second (2) cuvette 50 yl of the super-
natant and mix with 550 yl of the ferricynaide-phosphate mixture (reagent
3). In this step the total hemoglobin in cuvette No. 2 is oxidized to
MetHb. Measure the absorbances A^ and ^ at 632 nm against air. Add
20 yl of the neutralized cyanide solution (reagent 6) to each of the two
cuvettes. Mix gently and let stand for 1 min. Read again absorbances
A and A..
Calculation
A! - A3
- A4)x 12
xlOO = MetHb, in percent of total hemoglobin.
RESULTS
Precision
In Table 1 data on the precision of the method are presented. Repli-
cate blood samples drawn from three normal adults were assayed. The
concentration of MetHb was around 0.5% of the total pigment. The standard
error was in the range 0.019-0.025.
Accuracy
The accuracy of the method was investigated by recovery experiments.
A solution of MetHb was prepared from nitrite-treated erythrocytes with
13
-------
a MetHb content of 96%(2). The assay was modified by incorporating into
a series of microcentrifuge tubes, each one containing 200 yl of normal
blood, increasing amounts of the methemoglobinemic solution which
substituted part of the 550 ul of water used for hemolysis. The results
are summarized in Table 2. It can be seen that 94-99% of the added
"Nitrite-methemoglobin" is recovered.
Subject
A
B
C
Table IV-1 PRECISION OF METHOD
Number of
Replicates
10
7
6
Mean Percent
Methemoglobin
0.49
0.38
0.63
S.E.
0.020
0.019
0.025
Table IV-2 RECOVERY OF METHEMOGLOBIN ADDED TO HEMOLYSATES
Methemoglobin
added*
none
1.2
2.4
4.8
9.6
Percent of total pigment
Found Expected
0.46
1.60
2.84
5.23
9.88
1.66
2.86
5.26
10.46
Recovery
Percent
96.4
99.4
99.4
94.5
*The added methemoglobin was prepared from nitrite-treated erythrocytes.
*
Stability of MetHb
The stability of methemoglobin in blood samples was investigated
in both whole and hemolysed blood from normal adults (Table IV-3).
Storage of whole blood kept at room temperature (24°) or refrigerated
at 2° or below, results in the reduction of MetHb to Hb. Hemolysates,
held at 24° for few hours show a considerable increase to MetHb. How-
ever, in buffered hemolysates kept at 2° for not more than 24 h the
rate of auto-oxidation of hemoglobin is insignificant.
Table IV-3 STABILITY OF METHEMOGLOBIN WITH TIME UNDER DIFFERENT CONDITIONS
Condition of preservation
Hemolysate (2°C)
Hemolysate (2 C)
Hemolysate (2 C)
Hemolysate (24OC)
Whole blood (2°C)
Whole blood (24°C)
Whole blood (2°C)
Percent of Methemoglobin
at zero time after 24 hrs
1.2
4.8
43.1
1.2
1.2
1.2
43.1
1.3
4.7
41.9
24.5
0.9
0.8
18.5
14
-------
Conversion of hemoglobin to methemoglobin at low temperatures
Initial attempts to keep blood samples until assay in the frozen
state (at -20°C) failed as they showed considerable increases in MetHb.
This phenomenon has not been previously reported.
Table IV-4 OXIDATION OF HEMOGLOBIN DURING STORAGE AT -20°C
Conditions Percent of methemoglobin
0 3 hrs 24 hrs
Whole blood 0.6 5.7 12.4
Hemolysate 0.6 - 4.1
As can be seen from Table IV-4 this is a progressive reaction with
time. Controls run simultaneously at +2°c and -4°C (not frozen) did not
show any change in MegHb level. This reaction did not occur after freez-
ing at -192°C. The possibility that this oxidation occurred only during
slow freezing at -200C was considered. The results presented in Table
IV-5 show that rapid freezing of blood samples in liquid air (-192QC)
does not prevent the oxidations of samples held at -20°C. Samples frozen
at -192°C and held at this temperature for 24 hours were stable showing
no increase in MetHb. Storage of blood samples in liquid air was not
considered practical under field conditions.
Table IV-5 THE EFFECT OF STORAGE AT LOW TEMPERATURES ON THE OXIDATION OF
HEMOGLOBIN
Percent of Methemoglobin
Conditions 0 3 hrs 24 hrs
Freezing -20°C; held at -20°C 0.4 - 9.8
Freezing -192OC; held at -2QOC 0.4 5.1 10.8
Freezing -192°C; held at -192QC 0.4 - 0.4
Freezing -192°C; held for 3 hrs. at -20 C
and there for 21 hrs at -192OC 0.4 - 4.8
Freezing experiments done in the presence of dimethylsuIfoxide and
concentrated sucrose solutions (results not presented) prevent Hb oxida-
tion at -20°C. Nitrogen atmosphere during freezing at -20°C did not
influence the reaction. The mechanism involved in this phenomenon is
still obscure. Intracellular substrate may be involved in this reaction
and conformational changes may be part of the mechanism. More experiments
should be done before any conclusion can be drawn.
DISCUSSION
The sensitivity of the Evelyn and Malloy method in which the blood
is diluted 1:100 was assessed by the authors to be sufficient to detect
a minimum of 0.2 g MetHb per 100 ml of blood(3) . This sensitivity is
15
-------
considered low for the assay of MetHb in normal blood samples containing
approximately 0.15 g MetHb per 100 ml (calculated on the basis of total
Hb of 15 g per 100 ml wherein the MetHb content is 1%). By using a con-
siderably more concentrated hemolysate with a dilution ratio of 1:5 in
the modification presented above, sensitivity was increased to detect
approximately 0.02 g of MetHb per 100 ml.
The stability of the MetHb of the hemolysed freshly drawn blood
cooled to 2° allows the carrying out of field surveys in communities
distant from the laboratory, providing the samples are assayed within
24h; the pH of the buffered hemolysate should be approximately 7.35
(at 20°). Its alkalinity retards the auto-oxidation of Hb to MetHb.
If the samples are stored at room temperature, they should nevertheless
be assayed within 30 min.
REFERENCES
1. Evelyn, K.A., and Malloy, H.T. Micromethod of Oxyhemoglobin,
Methemoglobin and Sulfhemoglobin in a single sample of blood, ,J.
Biol. Chem. 126:655, 1938.
2. Rossi-Fannelli, A., Antonini, E., and Mondivi, B. Ferrihemoglobin
reduction in normal and methemoglobinemic subjects. Clin. Chim. Acta.
2:476, 1957. t
*
3. Henry, R.E., Clinical Chemistry, Harper and Row, Philadelphia, 1964,
p. 757.
16
-------
SECTION V
THE DETERMINATION OF NITRITE IN BLOOD
Investigations carried out in our laboratory concerning the relationship
between the amount of nitrate in drinking water and the content of methemo-
globin in blood emphasized the need for a method of estimating low nitrite
levels in the circulatory blood of laboratory animals fed with nitrate or
nitrite. The method was also required for planned field studies of infants
consuming water high in nitrate concentration.
The existing methods for nitrite determination in blood(l-3) were not
found to be satisfactory for the determination of very low concentrations
of nitrite, especially in the small blood sample volumes available in these
investigations.
In this work, different factors were investigated that may influence
nitrite determination in blood by a spectrophotometric method, based on
diazotization and coupling reactions, involving sulphanilic acid and
Cleve's acid. The modifications and refinements developed in order to
reach a detection limit of 0.01 ug N in 0.1 ml blood are described in
this paper. The advantages of the micromethod presented are high selec-
tivity, wide range, accuracy and lack of photosensitivity of the final
color obtained.
EXPERIMENTS
Apparatus
A Zeiss PMQ II spectrophotometer equipped with 10-mm cells (Hellma
105/4 microcell) and an Eppendorf Microfuge 3200 centrifuge providing
15,000 rev min"* with the special 1.5-ml microtubes were used. A conven-
tional clinical centrifuge may also be used.
Reagents
All chemicals used were of analytical-reagent grade. Nitrite-free
distilled water was used in the preparation of all solutions.
Sulphanilic acid solution. Dissolve 0.5 g of sulphanilic acid in
150 ml of a 20% (w/v) solution of glacial acetic acid in water.
Store in a brown bottle.
Cleve's acid solution. Dissolve 0.2 g of Cleve's acid (1-naphthy-
lamine-7-sulphonic acid) in 120 ml of water, warming in a water-
bath. Filter the solution, cool and add 30 ml of glacial acetic
acid. Store in a brown bottle in a refrigerator.
17
-------
Stock nitrite solution. Dissolve 0.4928 g of sodium nitrite and di-
lute to 1 liter with water (1 ml = 100 yg of nitrite as nitrogen).
Preserve with 1 ml of purified chloroform. Interfering substances
present in the chloroform can be removed by extracting 100 ml of
chloroform with four 20-ml portions of 0.1 M hydrochloric acid.
Standard nitrite solution. Dilute 10 ml of stock solution to 1
liter (1 ml * 1 yg NO~-N). Prepare this solution immediately before
use.
Zinc sulphate solution. Dissolve 4.31 g of zinc sulphate hepta-
hydrate in water and make up to 100 ml.
Recommended procedure
Collection of samples. Adequate precautions must be taken during
collection of blood samples for the determination of nitrite because
it is rapidly oxidizable in_ vitro. Blood samples should be analyzed
within the shortest possibTe time. However, the sample may be
stored at 40 for about 1 h without significant loss of nitrite
content.
Deproteinization. Add 0.1 ml of blood to a 1.5 ml microtube con-
taining 0.6 ml of zinc sulphate solution. Add 0.4 ml of bidistilled
water and mix well, preferably using a Vortex-Genie apparatus. To
this add 0.1 ml of aqueous 4% (w/v) sodium hydroxide solution and
mix again. Keep on ice for 1 h. Remove the precipitated proteins
by centrifugation (2 min at 15,000 rev min-1).
Color reaction*. Take 0.6 ml of clear supernate in a 100/10 test-tube
and add 0.4 ml of bidistilled water; mix well, add 0.1 ml of sul-
phanilic acid solution and leave on ice for 15 min for the diazo-
tization to take place. Add 0.1 ml of Cleve's acid solution and
leave for 60 min at room temperature for the coupling reaction. If
nitrite is present a red-violet diazo complex is formed. The absorb-
ance is read at 520 nm.
Preparation of standard curve. Add several aliquots (up to 0.5 ml)
of standard sodium nitrite solution (1 ml ~ Ivg NOj-N) to Eppendorf
centrifuge microtubes containing 0.6 ml of the zinc sulphate solu-
tion. Thereafter follow the technique as described under Recommended
procedure.
Double-distilled water was substituted for nitrite solution for the
reagent blank.
RESULTS AND DISCUSSION
The variables which may influence the results of nitrite determina-
tion in such a complex matrix as blood were investigated and different
procedures for removing blood proteins were tested. The precision and
accuracy of the method were evaluated statistically.
18
-------
Removal of blood proteins
One of the problems of the method for nitrite determination in blood
is the choice of a deproteinizing agent that will not inhibit color devel-
opment. Most protein precipitants are active in acidic media, which are
not suitable here, because of the instability of the nitrite ion. There-
fore, only protein precipitants acting in slightly alkaline or neutral
media may be considered. In thie present work, two deproteinizing systems
were compared: zinc sulphate with either barium hydroxide or sodium
hydroxide. Figure V-l shows the recovery curves between 0 and 0.5 yg of
nitrite-nitrogen in standard solutions and after treatment with both the
above-mentioned protein precipitants. It appears that the zinc sulphate-
sodium hydroxide deproteinizing system is superior to the other system and
does not inhibit color development.
0.1 0.2 0-3 0.4 0.5
conc.N02 as N (jjg / ml )
FIGURE V-l. INFLUENCE OF DEPROTEINIZING REAGENTS ON NITRITE RECOVERY (A) Bank-
(B) Ba(OH>2ZnS04; (C) NaOH ZnS04
19
-------
The color reaction
Coupling reagent. The classical coupling reagent in the Griess-
Ilosvay reaction is a-naphthylamine(4). This substance has been shown
to be carcinogenic(5). Among the other known non-carcinogenic coupling
agents, the folloiwng two were chosen for investigation: 1-naphthyl-
ethylenediamine(6) and 1-naphthylamine sulphonic acid (Cleve's acid(7)).
A comparison of these two coupling agents under the proposed experimental
conditions showed that both reagents gave good results, but Cleve's acid
was preferable; it provides a color of higher intensity, Beer's law is
obeyed over a wider range and the color is less dependent on the tempera-
ture. The results are presented in Figure V-2.
Diazotization time and coupling time. The time factor is important
in the" reactions involved in this method. Different diazotization times
were tried: 0, 5, 10, 15 and 20 min; and the coupling time was varied
concomitantly between 20 and 90 min. The results obtained under these
various conditions for a concentration of 0.2 ug NO^-N per ml are pre-
sented in Figure V-3 and are summarized as follows: with a diazotization
time of zero, when the coupling agent is added immediately after the
diazotization agent, the color developed is of low intensity, and increases
with time without levelling off after 60 min; with diazotization times
of 5, 10, 15, or 20 min, the strength of the color formed increases with
the increasing diazotization time and becomes more stable after 40-60 min.
Diazotization times longer than 15 min led to no further improvement.
Thus, the optimal diazotization time was established to be 15 min and the
coupling time 60 min.
1.500 -
1.000 -
w
u
e
to
0.500 -
0.1 0.2
0.3 0.4 0.5 0.6 0.7 0.8
cone. NO? as N (>jg/ml)
0.9
FIGURE V-2. STANDARD CURVES FOR NITRITE DETERMINATION, COMPARISON OF TWO
COUPLING AGNETS, (a) Cleve's Acid; (B) Naphthylethylenediamine
20
-------
ISJ
0)
O
c
O
(0
0.7
0-6
0.5
0.4
0-3
0.2
0.1
1
20 40 60
Coupling Time (minutes)
90
FIGURE V-3. EFFECT OF DIAZOTIZATION TIME AND COUPLING TIME. DIAZOTIAZTION TIME: (A) 0 min; (B) 5 min;
(C) 10 min; (D) 15 min; (E) 20 min.
-------
The effect of temperature. The effect of temperature on the final
color development was studied. During the diazotization time a tempera-
ture of about 1° is necessary. When the diazotization is done at 20°,
the final absorbance is decreased by about 10%. During the coupling
time, temperature variation between 20° and 30° does not affect the
intensity of absorbance. In parallel determinations with naphthyle-
thylenediamine and Cleve's acid, the influence of temperature was found
to be somewhat greater with the former reagent.
Evaluation of method
Limit of detection. Under the experimental conditions and with
the aforementioned optical system, the limit of detection is 0.01 yg
of nitrite-nitrogen in 0.1 ml of blood, which yields an absorbance
of 0.012. This sensitivity was adequate for the research on residual
nitrite in the blood of rats with induced methemoglobinemia. Other
known methods(1-3) are less sensitive and need at least 1 ml of blood
for each determination. The method of Litchfield(3), which requires
less than 1 ml of blood, has a sensitivity similar to the proposed
method, but needs special apparatus.
Precision. Precision was determined by 36 replicate analyses of
standard solutions in distilled water with 1, 4, and 5.5 ug of ni-
trite-nitrogen per ml. Standard deviations and 95% confidence limits
were calculated at the various levels. The results are shown in
Table V-l. The standard deviation varied between 0.025 and 0.110,
and the 95% confidence limits between ±0.01 and ±0.06.
Accuracy. The accuracy of the method was evaluated by determin-
ing the recovery pf different amounts of nitrite added to plasma.
Recovery data from whole blood were not evaluated because hemoglobin
reacts immediately with nitrite. Ten determinations were carried out
at each concentration in plasma and in distilled water; the results
are presented in Table V-2. The mean recovery varied between 97.5
and 100.2%.
22
-------
Table V-l PRECISION OF THE METHOD FOR NITRITE DETERMINATION IN BLOOD
(12 Determinations were done at each level)
N02 -N
Taken .
(yg ml )
1.00
4.00
5.50
Mean Value ,
Found (yg ml )
0.99
3.99
5.52
Range
0.97-1.04
3.96-4.04
5.40-5.68
Standard
Deviation
0.025
0.025
0.110
95% Confi-
dence Limit
0.99 ± 0.01
3.99 ± 0.01
5.52 ± 0.06
Table V-2 RECOVERY OF NITRITE FROM BLOOD PLASMA
(10 Determinations were done at each level)
N02 -N
Added
(yg ml-1)
0.40
1.00
2.50
4.00
6.00
Mean N02 -N
Found
(yg ml-1)
0.39
0.99
2.48
4.01
5.95
Standard
Deviation
0.06
0.04
0.07
0.08
0.16
Mean
Recovery
Percent
97.5
99.0
99.4
100.2
99.1
Recovery
Range
Percent
82-112
95-103
96.104
99.102
96.102
23
-------
Some applications of the method
The proposed method was specially developed for the determination
of small amounts of free nitrite in the blood of experimentally
nitrite-intoxicated rats in connection with the levels of methemo-
globin formed. Since hemoglobin reacts with nitrite, the determina-
tion of residual nitrite in blood is a necessary parameter in the
study of this reaction. At the first contact between nitrite and
blood, some nitrite reacts and hemoglobin is partially converted to
methemoglobin. The partition of residual nitrite among the differ-
ent blood components is a particular feature of interest.
Different amounts of nitrite-nitrogen between 1 and 100 yg ml"1
were added to human blood in vitro and simultaneously, for reference,
to distilled water. After the blood had been mixed, part of it was
centrifuged and the plasma separated. Residual nitrite was determined
in total blood and, in parallel, in plasma. The results obtained are
presented in Table V-3. With nitrite concentrations varying between
the above limits, the initial drop of nitrite in blood represents
76-90% of the amount added. The concentration of nitrite in plasma
was about twice that found in blood. In Table V-3 the amount of
nitrite was calculated for 0.53 ml of plasma, since plasma represents
53 ± 5% of the total blood volume(8). The total residual nitrite in
0.53 ml of plasma indicates figures very close to those obtained in
1 ml of whole blood. Therefore, it can be assumed that most of the
residual nitrite in blood is bound in the plasma fraction.
Table V-3 RESIDUAL NITRITE PARTITION IN BLOOD AND PLASMA
««- v, AJJ j
NO^-N Added
.* ni ,
to Blood
C g ml"1)
100.0
60.0
30.0
21.0
10.0
1.0
m««j
In Blood
(ug ml-*) (percent)
24.0
13.0
7.0
4.0
1.8
0.1
24.0
21.
23.
19.0
18.0
10.0
Found:
ln Plasma
(vg ml'1)
46.0
25.0
14.0
7.8
3.6
0.2
,-,._ n r-z
(yg 0.53
ml~la)
24.40
13.20
40
7.
4.
1.
10
90
0.11
Plasma represents 53% of the total blood volume 13%.
24
-------
REFERENCES
1. Stieglitz, E.J., and Palmer, A.E. "A Colorimetric Method for the
Determination of Nitrite in Blood," J. Pharmacol, Exp. Ther. 11,
398, 1934.
2. Diven, R.H., Pistor, W.J., Reed, R.E., Trautman, R.J., and Watts,
R.E. "The Determination of Serum or Plasma Nitrate and Nitrite,"
Am. J. Vet. Res. 25, 94, 1962.
3. Litchfield, M.A. "The Automated Analysis of Nitrite and Nitrate
in Blood," Analyst 92, 132, 1967.
4. Standard Methods for the Examination of Water and Wastewater, 12th
Ed., American Public Health Association, New York, p. 205, 1965.
5. Precautions for Laboratory Workers who Handle Aeromatic Amines,
Chester Beatty Research Institute, London, 1965.
6. Saltzman, B.E., "Colorimetric Microdetermination of Nitrogen Dioxide
in the Atmosphere," Anal. Chem., 26, 1949, 1954.
7. Bunton, N.G., Crosby, N.T., and Patterson, S.J. "Factors Influenc-
ing the Colorimetric Determination of Nitrite with Cleve's Acid,"
Analyst, 94, 585, 1969.
8. Wintrobe, M.M. "Clinical Hematology," 6th Ed., Lea Febiger, Phila-
delphia, p. 86, 1967.
25
-------
SECTION VI
DETERMINATION OF NITRATE IN HIGH MINERALIZED
DRINKING WATER USING THE ION SELECTIVE ELECTRODE
INTRODUCTION
A great variety of methods for nitrate determination in water have
been described. The majority of them fall into the following types:
nitration by the nitrate radical(1), reduction of the nitrate radical(2),
oxidation by the nitrate radical(3), ultraviolet absorption(4) and polaro-
graphic technique(5). Several serious interferences are commonly encount-
ered and they are often time consuming. All these methods are presented
in "Standard Methods for the Examination, of Water and Wastewater"(6) and
it is not surprising that all the methods are still listed as tentative
ones, because the nitrate determination is one of the most difficult tests
and its accurate measurement remains yet one of the unsolved problems
in environmental chemistry.
During the past few years an electrode selective for nitrate ion
has been developed, which is intended to give a direct measure of the
concentration of nitrate ion in solution. Potterton and Shults(7) have
reported on the theory and operation of this electrode. The selective
ion electrode technique for nitrate analysis has already found wide-
spread use in both laboratory and field determinations(8-10). It has
been used in soil and plant extracts(11,12), in microbial media(13), in
measuring sodium nitrite as an impurity(14) and in some aqueous systems
(15,16). The nitrate selective electrode appears to be convenient as a
simple and rapid method for determining nitrates in water.
•<
The object of this study was to establish the analytical efficacy
of the nitrate specific electrode in analysis of some drinking waters,
characterized by high mineralization and large variations in concentra-
tion of interfering ions. The electrode measurements were compared with
results of the phenoldisulphonic acid method, the most commonly used
spectrophotometric standard procedure.
EXPERIMENTAL
Instruments
Potential measurements were read from a battery powered pH/mV
specific ion meter (Orion Model 401). Both absolute and relative milli-
volt modes are provided. The nitrate indicator electrode was an Orion
26
-------
Model 9207 and the reference electrode was a single function electrode
Orion Model 9001 filled with the reference solution described later.
A Zeiss PMQ II spectrophotometer with 1 cm cells was used for mea-
surements in the phenol disulfonic acid method.
Reference Electrode
The reference electrode filling solution was prepared from a satur-
ated sodium sulphate solution, stirred and filtered. The original Orion
filling solution 900001, provided with the reference electrode, contains
nitrate ion(17) and it was proved unsuitable because contamination of
the test solutions may occur particularly when measuring dilute samples.
The composition of the electrolyte of the reference electrode causes inte]
ference if it contains ions'for which the ion selective electrode is sens:
tive. Recently Orion recommended a double function reference electrode
bridged with saturated sodium sulphate solution(18). The use of specific
ion flouride electrode as a nitrate ion reference electrode was also re-
ported (19).
Reagents
All reagents were of analytical reagent grade. Distilled water was
always glass distilled.
Sodium nitrate stock solution 200 ppm nitrate as nitrogen. Weigh
1.2140 gr. sodium nitrate which has been previously dried at 120°C.
Transfer to a 1 liter volumetric flask, dissolve. Add 1 ml of 0.1%
phenylmercurie acetate preservative solution, to inhibit biological growth.
Sodium nitrate standard solutions were prepared by diluting appro-
priate amounts of stock solution. All diluted standards must be pre-
pared daily.
Sodium sulphate stock solution 4000 ppm as sulphate. Add 5.9184 gr
sodium sulphate to a 1 liter volumetric flask. The conductivity of this
solution is 3.4 mmhos. Dilute the stock solution as needed.
Interfering anions solutions. Chloride, carbonate, bicarbonate,
nitrite, sulphate, all sodium salts, were used as sources of interfering
anions.
METHOD
Calibration
For greater precision in reading electrode potential the expanded
M.V. position ("M.V. EXP") is used. Measurements made on this scale are
27
-------
relative and can be adjusted with the calibration control. The lowest
concentrated nitrate standard is set to give a near full scale positive
reading.
For drinking water the maximum recommended level is 10 ppm nitrate
as nitrogen; the scale meter is calibrated with 10 ppm nitrate-nitrogen
in the middle and the upper point is set on 100 ppm. The lower limit
should then be 1 ppm, since the meter covers a range of two logs in
concentration at any one setting. The potential of each standard solu-
tion containing 1.0, 5.0, 10.0, 25.0, 50.0 and 100.0 ppm nitrate-nitrogen
was measured by transferring about 50 ml of solution, respectively, into a
100 ml beaker, immersing the electrodes and stirring uniformly with a
small teflon stirring bar for one minute before reading the mV value.
A high degree of uniformity in stirring and temperature for all standard
solutions is critical. In this study all determinations were made at
23° C ± 1°. Temperature fluctuations due to heat from stirring motor
were eliminated by using a thermal insulating barrier between the beaker
and the stirrer.
The relationship between nitrate concentration of the standardizing
solutions and the equilibrium cell potential E read is plotted on semilog
paper, the values of the concentration in ppm being plotted on the log
axis. Absolute calibration curves (red millivolt scale) are used to
verify correct electrode operation.
Sample Analysis - Taking into account both ionic strength and inter-
fering anions, the procedure for nitrate analysis with the electrode is
as follows:
The concentration-1 of chlorides is determined and the chlorides are
removed from solution by quantitative precipitation with silver sulphate.
The ionic strength of the unknown sample is estimated from specifc
conductivity measurements.
Before reading the potential of sample solution or a group of samples
with about the same ionic strength, calibrate the meter with two or three
standard nitrate solutions differing in concentration by a factor of 10
(1 ppm, 10 ppm and 100 ppm nitrate-nitrogen) and having the same total
ionic strength as the unknown sample. This can be done by diluting 1:1
standard nitrate solution with sodium sulphate solution, prepared in the
needed concentration to obtain the same conductivity in standard as in
sample. (In preparing all dilutions pipets and volumetric glassware must
be used). Immerse electrodes successively in each standard calibration
solution and adjust with Calibration Control so that potential reading
will correspond exactly to the calibration curve millivolt previously
obtained for the respective standard. Finally, place the electrodes in
the unknown sample and read, at the same position, the potential developed.
Allow reading to stabilize after one minute stirring, at the same
rate and temperature as the standard. The potential measurement of
28
-------
the sample is converted to nitrate concentration by using the standard
calibration curve. Between measurements electrodes should be blotted
dry with absorbent tissue.
RESULTS AND DISCUSSIONS
Interfering Factors
The measured potential of the nitrate activity depends on the ionic
strength of the solution because the activity coefficient decreases with
increasing ionic strength. On the other hand, electrode response is in-
creased by interference from many common anions and this increased poten-
tial can counteract the influence of the ionic strength. The net inter-
ference is the result of these two opposing effects.
Ionic Strength In principle the influence of ionic strength can be
avoided in electrode determinations by the addition of a relatively con-
centrated salt solution of a non-interfering electrolyte, such that all
the samples and standards are analysed at the same ionic strength.
Sodium sulphate solution is the best ionic strength adjuster be-
cause it has the properties of high ionic strength, good equitransference
and is free of nitrate interfering anions, its selectivity constant being
K = 3.10~5(18). The ionic strength of samples can be estimated by measur-
ing the electrical conductance. In standard and unknown solutions of the
same ionic strength, nitrate ion activity is assumed to be in constant
proportion to concentration. In practice, this assumption involves deter-
mination first of the conductivity of the unknown sample solution; then
calibration of the electrode with a variety of mixed standard solutions
prepared in sodium sulphate solution with a concentration of identical
conductivity value as the sample.
The effect of total ionic strength has been verified by the increase
in slope of the calibration line at high salt concentration and by increase
in nitrate content found, when measurements are made in the same diluted
solution.
Table VI-1 presents results obtained for a solution of 10 ppm nitrate-
nitrogen and varying sodium sulphate concentrations. The calibration was
made with 500 ppm sodium sulphate solution containing various amounts of
nitrate. The measured nitrate content is smaller when the sulphate con-
centration increases; this is the effect of ionic strength, which influ-
ences the activity of nitrate ion. The interference from sulphate anion
is insignificant.
Anion Interference - Theoretically the magnitude of anions inter-
ference can be calculated by substituting the concentrations (Ci) and
29
-------
Table VI-1. DETERMINATION OF NITRATE BY SELECTIVE
ELECTRODE METHOD AT VARIOUS IONIC STRENGTHS
Nitrate nitrogen
added
ppm
Sodium sulphate
ppm
Conductivity
mmhos lcm~*
Nitrate nitrogen
detected by
electrode
ppm
10
10
10
10
10
10
0
50
200
500
1000
2000
0.085
0.135
0.300
0.600
1.100
1.900
11.8
11.2
10.5
10.1
9.0
8.3
Table Vl-2. ANIONS INTERFERENCE IN NITRATE-NITROGEN
DETERMINATION BY SPECIFIC ION ELECTRODE
Anion added, ppm
Nitrate - as nitrogen added, ppm
1 5 10 25
Nitrate as nitrogen detected, ppm
50
Nitrite
Bicarbonate
Carbonate
Chloride
Sulphate
5
10
50
50
250
500
50
250
500
50
250
500
1000
50
200
500
1000
1.20
1.50
4.40
0.25
0.40
0.50
1.00
1.15
1.20
1.30
*
2.50
3.60
5.20
0.20
0.30
0.60
0.70
1.30
1.70
4.50
1.10
1.50
1.60
1.50
1.70
1.70
1.30
2.10
3.00
4.80
1.00
1.10
1.00
0.90
5.00
5.60
8.40
5.80
7.30
7.50
6.00
6.40
6.70
5.80
7.00
8.00
9.80
4.90
5.00
5.00
4.80
10.10
10.50
13.50
11.50
12.00
13.80
11.40
12.00
12.50
11.50
13.20
15.20
17.00
10.40
10.10
10.00
9.80
26.10
28.20
28.70
25.00
26.10
27.00
27.00
29.00
29.00
27.00
28.00
30.00
34.00
25.00
25.10
25.20
24.90
53.20
58.10
60.40
50.00
51.50
52.10
54.00
58.00
59.50
52.00
56.00
59.10
64.20
50.00
50.00
49.70
49.30
30
-------
selectivity constant (Ki) values of interfering anions (i) into the elec-
trode response equations. The precise evaluation and empirical correc-
tion of anion interferences is difficult to establish, because the elec-
trode response is influenced not only by the concentration of the inter-
fering anion, but also by total solution compositions and by the nitrate
level. Therefore the approach to a systematic correction has important
limitations.
To evaluate the net influence of most common anions occuring in
water, electrode measurements were made in a variety of synthetic solu-
tions containing from 1 to 50 ppm nitrate-nitrogen and varying amounts of
nitrite, hydrogen carbonate, carbonate, chloride and sulphate. The results
are presented in Table VI-2.
It is observed that the level of interfering ion, the total ionic
strength and the level of nitrate do influence the measured potential.
For most anions occuring in water, additive influences from nitrite and
chloride are major. Nitrite is seldom present in high concentrations in
drinking waters. As to the strongly interfering chloride, it must be
initially removed by precipitation with silver sulphate if accurate deter-
minations are to be made. The elimination of chlorides was also performed
by treatment with cationic ion exchange resin, prepared in the silver form;
in this study better results were obtained by Ag2SQit precipitation.
Removal of interferences is a problem: an excess of removing reagent
may increase sample ionic strength and may deviate the pH from the range
of proper electrod response. To avoid this excess, it is necessary to
know the concentration of interfering chlorides prior to their removal by
precipitation. The advantage is that after this precipitation, sodium
sulphate is formed in the sample medium solution -the same ionic strength
adjuster as added in standard solutions prepared for calibration.
Langmuir and Jacobson(16) have calculated the approximate amounts of
different anions which, if present together with 10 ppm of NOi, produce a
+1% error in the nitrate determination by electrode method. The most
interfering anion was found to be nitrite; for this anion the presence of
1.2 ppm is sufficient to cause an error of +1%. The chloride anion causes
a similar error when present in concentration of 2 ppm. The sulphate anion
causes little interference. There must be 5000 ppm to cause an error of
+1% in the 10 ppm NOa determination.
Operating Procedure Conditions
Temperature Effects - The Nernst equation is temperature dependent
and the slope of the calibration plot E versus nitrate concentration changes
with temperature. This dependence is not the same for different membrane
materials, therefore a universal temperature compensator is not feasible.
Precise results' require meticulous temperature control and samples and
standards must be measured at the same temperature.
31
-------
Electrode Equilibration - The response time is mainly determined by
the measuring conditions. Continuous and reproducible stirring of the
samples and standards is necessary. Non-uniform stirring causes insta-
bility of the potential at equilibrium. The degree of stirring of the
solution affects the equilibration time of the electrode; stirring of the
solution reduces the time taken to reach equilibrium. High stirring rates
introduce air bubbles into the solution, which causes unstable electrode
potential response, because the air bubbles accumulate on the surface of
the electrode and decrease contact with the solution. In the usual stand-
ardized conditions equilibrium is reached after 1 minute.
The depth of immersion of the electrodes in solution must always be
the same. On changing over from one sample to another, the equilibrium
at the surface has to be re-established; the time required for this depends
on the difference between the concentrations in the samples. With large
concentration jumps, when going from a high to a low concentration, more
time is required before the measurement is stable.
Between determinations, electrodes should be wiped dry with absorbent
tissue before immersion in the next solution.
Comparison with Phenoldisulfonic Acid Method
Parallel determinations of nitrates were performed on 148 spring and
well water samples, by both the standard phenoldisulfonic acid method
(P.D.S.) and the described specific electrode method. The ranges of nitrates
as found by the P.D.S. method, varied between 1.4 to 40.7 ppm, most of the
samples (78%) having between 5 to 20 ppm. The mineralization of the samples
varied largely and was high in most of them. The conductivity varied be-
tween 0.45 to 3.2 mmhcfs and the chlorides between 40 to 750 ppm; most of
the samples (65.6%) contained more than 300 ppm chlorides.
The results of this comparison are summarized in Table VI-3. A
good agreement was generally observed between the results of both methods.
The correlation coefficient (r) for the relationship of phenoldisulphonic
acid and electrode values was found .98. The linear regression equation
is y = 0.99x + 0.96.
The electrode measurements values are in most cases higher than those
of the phenoldisulphonic acid method. Perhaps this is because this pro-
cedure does not correct for bicarbonate interfering anion. The mean dif-
ferences do not exceed 10%.
The results of nitrate determination by electrode method were com-
pared with those of phenoldisulfonic acid method, as a function of con-
32
-------
Table VI-3. COMPARISON OF SPECIFIC ION ELECTRODE AND
PHENOLDISULPHINIC ACID METHOD FOR NITRATE-NITROGEN
DETERMINATION IN DRINKING WATER SAMPLES
Conductivity
mmhos.
mm.
max.
Nitrate-nitrogen Number of Nitrate-nitrogen, ppm
level range Electrode P.D.S.*
ppm samples mean values
2.92
8.78
15.24
26.15
35.94
*Phenoldisulphonic Acid
Table VI-4. THE MEAN DIFFERENCES (PERCENTAGE) BETWEEN THE
RESULTS OBTAINED BY P.D.S.* AND E* METHODS, CORRELATED
WITH LEVEL OF NITRATE AND CONDUCTIVITY VALUES
Chlorides
ppm
min. max.
< 5.0
5.0 -10.0
10.1 -20.0
20.1 -30.0
>30.0
TOTAL
4
45
71
12
16
148
2.60
8.02
14.45
25.00
34.86
0.45
0.50
0.70
0.80
1.80
1.10
3.15
3.15
2.70
2.90
76
40
111
95
500
140
750
740
600
580
Nitrate-nitrogen
level range
ppm
1.4 - 9.9
10.0 -19.9
20.0 -40.7
Number of
samples
39
25
5
Mean difference
percent
6.6
5.9
4.0
Number of
samples
10
46
23
Mean difference
percent
20.1
7.4
3.5
*P.D.S. phenoldisulphonic acid
E electrode
33
-------
ductivity and of nitrate level. The percentage differences between the
two methods are presented in Table VI-4. The differences are more ac-
centuated in the samples with low nitrate content (<10 ppm) and high
salt content, (conductivity > 2.0 mmhos.) The differences are lower in
samples with high nitrate concentration.
CONCLUSIONS
The results obtained in nitrate determination in highly mineralized
waters by the electrode method performed under meticulously controlled
laboratory conditions including removal of chlorides are in good agreement
with those obtained by the standard phenoldisulphonic acid method. The
electrode measurements values are in general somewhat higher than those
of the phenoldisulphpnic acid method, especially in the range of low
nitrate concentrations and high salt contents.
The electrode method for nitrate determination is not a field method,
because of detailed calibration requirements and the complex controls
required to obtain reliable results. When nitrate concentrations are low
and chlorides are high the electrode results may be many fold higher than
in reality.
In water of low mineralization it may be used as a screening method
to identify the content of nitrate with a minimum of special procedures.
In water of high mineralization it needs standarization of curves
with solutions of the same order of ionic strengths as the samples and
elimination of interfering anions. Therefore>it is laborious when dif-
ferent types of water are to be examined in the same series.(20)
>
The electrode method is practical and rapid, when many samples be-
longing to a same source are to be compared or for continuous monitoring
of a water supply with known mineralization. It cannot be relied upon
as a field method of universal applicability unless degree of accuracy
required can allow for deviations of some 20%, or more.
34
-------
REFERENCES
1. Taras, M.J. Phenoldisulphonic Acid method of determining nitrate in
water. Anal. Chem. 22:1020, 1950.
2. Edwards, G.P. Determination of nitrates in wastewater effluents and
water. J.W.P.C.F. 34:1112, 1962.
3. Jenkins, D. § Medsker, L.L. A brucine method for the determination
of nitrate in ocean, estuarine and fresh waters. Anal. Chem. 36:610, 1964.
4. Navone, R. Proposed method for nitrate in potable waters. J.A.W.W.S.
56:781, 1964.
5. Frazier, R. E. Polarographic determination of nitrate nitrogen in
water. J.A.W.W.A. 55:624, 1963.
6. Standard Methods for the Examination of Water and Wastewater 13th
edn. APHA, AWWA and WPCF New York, 1971.
7. Potterton, S.S. § Shults, W.P. Analyt. Lett. 1, 11, 1967.
8. Andelman, J.B. Ion Selective Electrodes theory and applications
in water analysis. J.W.P.C.F. 40:1844, 1968.
9. McClelland, N. $ Mancy, K.H. (1972) Water Quality monitoring in
distribution systems. J.A.W.W.A. 12, 795.
10. Durst, R.A. editor, Ion Selective Electrodes. National Bureau of
Standards Special Publication 314, 1969.
11. Oien, A. & Selmer Olsen, A.R., Nitrate determination in soil
extracts with nitrate electrode. Analyst 94:888, 1969.
12. Paul, J. L. § Carlson, R.M., Nitrate determination in plant extracts
by the nitrate electrode. J. Agr. Food Chem. 16:837, 1968.
13. Manaman, S.E., Nitrate ion selective electrode in microbial media.
Applied Microbiol. 18:479, 1969.
14. Gehring, G.D., Dippel, W.A. S Boucher, R.S., Determination of sodium
nitrate in sodium nitrite by selective ion electrode measurement.
Anal. Chem. 14:1686, 1970
15. Keeney, D.R., Byrnes, B.H. § Genson, J.J., Determination of nitrate in
waters with the nitrate selective in electrode. Analyst 95:383, 1970.
16. Langmuir, D. § lacobson, D.L., Specific Ion electrode determination
of nitrate in some fresh waters and sewage effluents. Env. Sci. Tech.
4, 834, 1970.
35
-------
17. Orion Research Inc., (1970) Instruction sheet single function
reference electrode.
18. Orion Research Inc., (1972) Analytical methods guide. Fourth edn.
19. Manahan, S.E., Anal. Chem. 42:128, 1970.
20. Shechter, H. & Gruener, N. An evaluation of the ion selective
electrode method for determination of nitrate in highly mineral-
ized drinking water. J.A.W.W.A. 68:543,1976.
36
-------
SECTION VII
MEASUREMENT OF ASCORBIC ACID IN BLOOD
INTRODUCTION
Ascorbic acid is considered to be one of the natural antimethemoglo-
binemie agents and is also used clinically as therapeutic agent for that
purpose (1) .
The importance of this compound from an epidemiological and nutri-
tional point of view has become apparent from our field studies (2).
In order to check ascorbic acid levels among infants which are the
susceptible section in the population, we developed a sensitive micro-
method for ascorbic acid in blood.
Existing methods for ascorbic acid measurement in blood suffer from
several disadvantages:
1) Low sensitivity.
2) Lack of specificity.
3) Large volume samples.
PRINCIPLE
Ascorbic acid reduces Fe to Fe which is complexed by a specific
iron* reagent to a red compound which is determined by photometry.
REAGENTS
3M Acetate buffer pH 4.5.
10% Trichloroacetic acid CTCA) .
0.05% Bathophenantroline disulphate (BPAS) .
0.05% Ferric chloride.
85% Orthophorphoric acid.
METHOD
All reagents added consecutively, in order with thorough mixing at
each addition.
1) 0.25 ml of TCA was added to 0.25 of clear plasma. 0.25 ml of
TCA supernatant was used.
2) 0.25 ml
3) 0.20 ml orthophosphoric acid.
4) 0.30 ml BPAS.
5) 0.20 ml acetate buffer.
37
-------
Hold in dark. After 30 min. read against ascorbic acid standards (in
5% TCA) at range between 0-1 mg/100 cc. Blank contains 0.25 ml 5% TCA
in place of supernatant. Read at 534 my in Coleman spectrophotometer.
Linearity up to 10 mg% in plasma.
RESULTS
Table VII-1. HUMAN ASCORBIC ACID LEVELS
mg/100 cc of plasma
Source mean n
Cord blood 1.73 120
Infants (1-6 months) 0.84 24
Pregnant women 1.08 33
The results represented here show that the fetus has a higher level
of ascorbic acid than the mother and that this level fell to half within
the first months of life. This fact correlates with our findings that
MetHb levels are low in the fetus and are the highest at 0-3 months then
fall to adult level. This phenomenon may hint that ascorbic acid may
be a factor in determining MetHb levels in_ vivo.
COMMENTS
1) TCA did not have any effect in complex formation. Perchloric
acid as precipitant can replace TCA.
2) Kinetic study* on effect of time and temperature gave best
linearity at 30-9*0 at 40°C.
Whan a water-bath is not available incubation can be at room
temperature as well.
3) FeCls, optimal concentration was found to be 0.5%. Recovery
of ascorbic acid based on Fe+2 concentration found under these
conditions was 93%.
4) Highly cleaned laboratory glassware is absolutely necessary.
5) When whole blood was used a small .steady increase was observed.
This may be because of glutathion presence.
6) Storage of samples (Plasma or deproteinized plasma in TCA)
did not show any loss for 24 hrs. for long storage freezing is
recommended .
38
-------
REFERENCES
1. Tonz, 0. The Congenital Methemoglobinemia, Bibliothea Hematologica
No. 28, S. Kruger, Basel, N.Y., p. 72-74, 1968.
2. Shuval, H. T. and Gruener, N., Am. J. Pub. Health 62:1045, 1972.
39
-------
SECTION VIII
EPIDEMIOLOGICAL STUDY OF THE ASSOCIATION BETWEEN NITRATES IN
DRINKING WATER AND METHEMOGLOBIN LEVELS IN INFANTS
REHOVOT AREA
INTRODUCTION
Infant methemoglobinemia resulting from consumption of water with
high nitrate concentrations was first recognized clinically by Comly in
1945(1). Since then about 2,000 cases, including many fatal poisonings
in infants resulting from ingestion of water containing nitrates have
been recorded in various countries throughout the world(2). The U.S.
Public Health Service Drinking Water Standards(3) recommend that water
consumed by infants not contain over 45 mg/1 of nitrates as NQ$. The
World Health Organization has made a similar recommendation. These
standards were originally based on limited epidemiclogical evidence
gathered by Walton in 1951(4) indicating that no cases of methmoglo-
binemia had been reported in the United States when water containing
less than 45 mg/1 of NO, was consumed. More recent studies in Europe
have, however, reported on clinical and subclinical cases of the disease
in infants consuming water having less than 45 mg/1 of N0_(5).
Recently there has been pressure to relax the nitrate standard in
drinking water in areas exposed to increasing nitrate pollution of ground
water in which overt cases of infant methemoglobinemia are rarely re-
ported. The Department of health in California has issued administrative
orders which require surveillance of infants' health in communities
supplying water containing up to 90 mg/1 NOj while recommending the
discontinuation of the source only at higher concentrations. The W.H.O.
Recommended Drinking Water Standard for Europe-1970 more or less have
followed suit. The 1971 International Drinking Water Standards of the
W.H.O. retained, however, the recommended maximum limit of 45 mg/1. The
standard set in Israel is 45 ppm N0_ as the recommended limit but allow
the use of water up to 90 ppm on condition that low nitrate water be
made available to infants up to one year of age.
The epidemiological study reported upon here is part of a broad spec-
trum program to re-evaluate the current nitrate standard for water. Of
particular interest was the possible presence of chronic sub-clinical
methemoglobinemia in infants in the areas of Israel having medium-high
concentrations of nitrates in the water supply but few overt clinical
cases, as reported on by Knotek and Schmidt in Czechoslovakia(5).
METHODS
Two thousand four hundred seventy-three infants were studied in
communities with medium-high and low concentrations of nitrates in the
drinking water. The study area is on the coastal plain of Israel and
40
-------
Figure 1—Nitrate Radical (N03) / <.
Conteflt of Pumped Water In
Cerftral Israel as of 1968
Ltacxo
• NO> CONTCXr «BOVt «3P*M
• MCACASE N MTHtfE CONTCNT
WDM TMW 2 WM M IN WILLS
UNCC* •
. OTHCn WELLS
L
r
Figure VIII-1. Nitrate Radical (N0_) Content of
Pumped Water in Central Israel as
of 1968.
41
-------
includes the towns of Rehovot, Rishon le Zion, and Nes Ziona as well as
a number of small agricultural settlements near Natanya (see Figure VIII-1).
The nitrate concentration in most of the water supply wells in these
communities varied from 50-90 mg/1. However, since the integrated water
supply systems in each community are served by a number of wells of
different nitrate content, the actual NO, concentration at the tap in a
given home may vary from hour to hour depending on the pumping rates and
water demand. Infants were examined at well-baby clinics run on a rou-
tine basis by the Ministry of Health, or by the Kupat Holim-Medical
Insurance scheme, so that it was not practical to determine the exact
concentration of nitrates at the home tap of the infant examined at the
time of examination. For the purposes of this analysis all infants from
the study area are considered as one group with a mean of about 70 mg/1.
Jerusalem, with 5 mg/1 of nitrates in the water supply, was chosen as
a control area, and 758 infants were examined at Municipal Health Depart-
ment well-baby clinics.
A 0.2 ml fingertip blood sample was taken from each infant studied
by project staff nurses and examined for hemoglobin and methemoglobin
(MetHb) in the laboratory in Jerusalem. An accurate and precise micro-
method for testing MetHb was developed to meet the needs of this study
and similar field studies. The method is a modification of the classical
Evelyn and Malloy(6) method and has been reported upon previously(7).
Blood samples are stabilized so that they can be taken in distant communi-
ties and transported to the laboratory for assay within 24 hours rather
than within less than an hour as required by the conventional procedure.
This method can show differences in MetHb as small as .1% which is essen-
tial in a study aimed at detecting slightly raised levels of MetHb in
a healthy population where normal MetHb levels are expected to be about
1% of total hemoglobin.^
A detailed questionnaire was filled in by the project nurse for
each infant, based on information supplied by the mother and the files of
the well-baby clinic (appendix 1). Details on age, sex, weight, ethnic
background, health status and nutritional regime were recorded and together
with the laboratory results were punched on cards for computer analysis.
Particular attention was given to intake of tap water with milk formula
or from other sources.
RESULTS
There are essentially no differences between the mean MetHb levels
in the study group and the control group as shown by Table VIII-1. The
mean MetHb level for the entire population studied was 1.04% with a
standard deviation of .72%. For the most susceptible age group of 1-60
days the mean MetHb for the study group is 1.38% while being 1.30% for
the controls. The mean for both groups is 1.33%. The differences are
not significant. For all ages, the mean MetHb percent in the control
population is actually higher than that of the study population, but
this is not statistically significant. However, on further analysis it
was shown that 31.5% of the infants up to 60 days of age in the study
42
-------
Table VIII-1. METHERMOGLOBIN IN INFANTS IN AREAS WITH HIGH
AND LOW NITRATE CONCENTRATIONS IN DRINKING WATER
Mean
N03
Mg/1
Control area 5
Study area 50-90
Total
1-60 days 61-90 days
n MetHb% n MetHb%
91+ days
n MetHb%
All ages
n MetHb% S.D.
96 1.30 75 1.24 556
71 1.38 188 1.14 1426
167 1.33 263 1.17 1995
* Total includes infants with age unknown.
.97 758*
.99 1702*
1.11
1.01
.98 2473*' 1.04
.72
.72
.72
Table VIII-2. DISTRIBUTION OF HEMOGLOBIN LEVELS AMONG INFANTS
Hb
up to 9.0
9.1 - 10.0
10.1 - 11.0
11.1 - 12.0
12.1 - 13.0
13.1 - 14.0
14.1 +
unknowns
numbers of
infants
Hb mean
Jerusalem
Jews
1.5%
8.6
21.9
29.7
21.3
10.1
3.3
3.6
606
11.6
Jerusalem
Arabs
5.5%
13.1
28.0
28.0
16.9
3.4
3.7
1.3
236
11.2
(Age, 1-18 months)
Nes-Ziona Rehovoth Rishion Le Zion
4.1%
9.3
26.9
29.2
19.0
5.4
2.9
3.1
484
11.4
1.0%
6.5
18.1
33.5
25.0
13.0
2.
0.
,1
.7
828
11.8
0.9%
5.6
18.3
37.6
25.3
8.2
3.8
0.4
466
11.8
43
-------
population have MetHb levels higher than 1.7% as against 21.9% for the
controls. Table VIII-2 shows the hemoglobin levels of the groups studied.
The means of each group are almost identical.
Only 6% of the infants studied received powdered milk formula made up
with tap water. The remaining 94% of the infants were either breast-
fed and/or received whole cow's milk. Those infants fed with powdered
milk formula in the study area showed somewhat higher MetHb percent than
those fed exclusively on other types of milk (see Table VIII-3). There
are only 36 infants in this category and the differences are not signifi-
cant. Detailed analysis of the effect of age, vitamin C intake, diarrhea
on MetHb levels did not reveal any significant differences in this group
as compared to the control.
Forty-nine percent of the infants in the 1-60 day age group consumed
either citrus or tomato juice or both. For the 61-90 days group, 80%,
and over 91 days, 91% consumed such vitamin C rich foods. The mean con-
sumption for the total population studied was 87%. In Table VIII-4 the
MetHb levels in infants with or without vitamin C rich foods in their
diet are presented. The 1-90 day age group of the study population con-
suming citrus or tomato juice show somewhat lower MetHb levels as com-
pared to the non-consumers. In the control group and in those over 91
days in age there were no differences.
In the 1-90 day age group infants reported to be suffering from
diarrhea on the day of examination or within the last month showed higher
MetHb in both the study and control area (see Table VIII-5). In the
study area, 1-90 day infants with diarrhea had a mean of 1.78% MetHb as
compared to 1.16% in infants not suffering. Infants of 1-90 days with
diarrhea from the study area showed higher MetHb than those from the
control area. The numbers in this category are small and the differences
are not significant.
There were essentially no differences among infants above 91 days
in age.
MetHb is significantly higher in the first 60 days of life in both
study and control populations (Table VIII-1). Table VIII-6 shows that
the mean for both groups is 1.33% for the 1-60 day group, and 1.17% for
the 61-90 day group. Infants over 90 days show MetHb levels of about 1%,
with the mean 6f-.1.04% for the total population studied. The same patterns
are found with present weight as shown in Table VIII-7.
DISCUSSION
The fact that there are no significant differences between the mean
MetHb levels in the 1,720 infants in the study area with a mean nitrate
concentration of 70 mg/1 in the drinking water as compared to the 758
infants in the control area does not support the findings of subclinical
44
-------
Table VIII-3. METHEMOGLOBIN IN INFANTS DRINKING
POWDERED MILK FORMULA AND ONLY OTHER FORMS OF MILK
Control area
Study area
Total
Powdered Milk
n MetHb%
111
36
147
.98
1.17
1.01
n
664
1666
2310
Only other
Forms of Milk
MetHb%
1.14
1.00
1.04
Table VIII-4. METHEMOGLOBIN IN INFANTS WITH AND WITHOUT
CITRUS OR TOMATO JUICE IN DIET
Control area
Study area
Total
Age 1-90 days
Age 91+ days
with
n MetHb%
65
226
1.28
1.19
without
n MetHb%
106
33
1.27
1.30
with
n MetHb%
452
1366
1.03
.97
without
n MetHb%
104
73
1.09
.98
291 1.21 139 1.28 1818
.98
177
1.04
Table VIII-5. METHEMOGLOBIN IN INFANTS WITH AND WITHOUT DIARRHEA
Control area
Study area
Total
Age 1-90 days
Age 91+ days
with
n MetHb%
33
20
1.43
1.78
without
n MetHb%
137
239
1.23
1.16
with
n MetHb%
186
165
.99
1.01
without
n MetHb%
369
1254
1.06
.96
53
1.56 376 1.18
353
1.00 1635
.98
45
-------
Table VIII-6. METHEMOGLOBIN AS A FUNCTION OF AGE
Age in Days
1-60
61 - 90
91 - 120
121 - 180
181 - 270
271 - 360
361 - 450
451 - 540
541+
Unknown
All Ages
Table V^II-7.
Wgt in Kgs.
0 - 4.0
4.1 - 5.0
5.1 - 6.0
6.1 - 7.0
7.1 - 8.0
8.1 - 9.0
9.1 - 10.0
10.1 - 12.0
12
Unknown
Total
n
167
263
235
393
461
411
277
161
57
48
2473
METHEMOGLOBIN
n
38
204
303
341
408
431
317
323
38
70
2437
MetHb%
1.33
1.17
1.07
.95
1.00
.94
.99
.97
.86
-
1.04
BY PRESENT WEIGHT
MetHb%
1.37
1.30
1.13 -
1.08
.99
.96
.93
.99
.97
-
1.04
46
-------
methemoglobinemia reported by Knotek and Schmidt(7) in Czechoslovakia.
A possible explanation for this lies in the differences in infant feeding
practices found in Israel where only 6% of the infants included in the
study received appreciable amounts of tap water together with formula
prepared from powdered milk. Measurements of fluid intake in infants
1-5 months of age indicate that during hot dry periods their intake of
tap water given as a supplement is about 20%-50% of their total daily
liquid intake, which is mainly made up of mothers' milk or cows' milk.
However, there were no indications of higher MetHb levels during the
summer months despite somewhat higher intake of tap water. The specific
role that powdered milk per se may play in raised MetHb levels as suggested
by Knotek and Schmidt(5) is still a mute point. Another factor may be the
widely practiced feeding of vitamin C rich foods such as citrus and
tomato juice to infants. Both foods are inexpensive and available during
most of the year in Israel. Eighty-seven percent of the infant popula-
tion studied consumed such vitamin C rich foods which are known to be
an effective antidote to methemoglobinemia. In Central Europe it can
be assumed that such vitamin C rich food supplements are expensive and
not always in supply so that they are much less frequently used.
The finding that infants in the study area receiving powdered milk
formula or no vitamin C rich food supplements had slightly higher MetHb
than their controls is supportive of this hypothesis, although the number
of infants in each category was small and the differences are not statis-
tically significant.
The higher MetHb levels in infants (1-90 days) reported to be suffer-
ing from diarrhea in both the study and control area supports clinical
findings in Israel of an association between infant diarrhea and methemo-
globinemia (8) . The fact that in the study area the level of MetHb in
infants with diarrhea was higher than similar infants in the control area
may be suggestive of an additive effect resulting from the higher expo-
sure to nitrates in water. But there again the number of infants in
those categories is small and the differences not statistically signifi-
cant.
The finding of higher MetHb levels in the first 60 days of life in
both study and control groups supports similar findings by Shearer
et al(9).
One conclusion from this study is that no apparent public health
problem associated with infant methemoglobinemia was detected in the
study area, despite the fact that most of the wells were supplying water
with nitrate concentrations above that generally recommended by public
health authorities. However, it would be incautious to extrapolate from
these findings as to the situation that may exist in other areas where
infants consume appreciable amounts of tap water together with milk formula
and where vitamin C rich foods are not widely given as diet supplements
to such young infants(10).
47
-------
SUMMARY AND CONCLUSIONS
Two thousand four hundred seventy-three infants in the Rehovot area
with medium-high (50-90 mg/1 as NO,) and in Jerusalem with low (5 mg/1
as NO-) concentrations of nitrate in drinking water were studied in an
effort to determine whether there is any association between methemoglo-
bin (MetHb) levels and nitrates in drinking water.
MetHb levels are highest in the first 60 days of life with a mean
of 1.33%. The mean MetHb level for the entire infant population studies
was 1.04±.72%.
No differences were found between the mean MetHb level in the study
and control areas.
A possible explanation for this finding lies in the fact that only
6% of the infants consumed appreciable amounts of tap water together
with powdered milk formula. The remainder were breast-fed or consumed
whole cow's milk. Eighty-seven percent of the infants were fed vitamin
C rich foods such as citrus or tomato juice which are known for their
anti-MetHb effects. It cannot be assumed that there is no danger from
nitrates in drinking water in areas where infants consume larger amounts
of tap water with powdered milk formula and/or little vitamin C rich
foods.
48
-------
REFERENCES
1. Comly, H.H., Cyanosis in infants caused by nitrates in well water.
J.A.M.A. 129, 112, 1945.
2. Gruener, N. and Shuval, H.I., Health aspects of nitrates in drinking
water. In Developments in Water Quality Research (p 89-106), Ann
Arbor Humphrey Science Publishers, Ann Arbor: London, 1970.
3. U.S. Public Health Service Drinking Water Standards, PHS Publ. No.
956, 1962.
4. Walton, B. Survey of the literature relating to infant methemoglo-
binemia due to nitrate contaminated water. A.J.P.H. 41:986, 1951.
5. Knotek, Z. and Schmidt, P. Pathogenesis, incidence and possibilities
of preventing alimentary nitrate methemoglobinemia in infants.
Pediatrics 34,78, 1964.
6. Evelyn, K.A. and Malloy, H.T. Microdetermination of oxyhemoglobin,
methemoglobin and sulfhemoglobin in a single sample of blood. J_.
Biol. Chem., 126, 655, 1938.
7. Hegesh, E., Gruener, N., Cohen, S., Bochkovsky, R., and Shuval, H.I.
A sensitive micromethod for the determination of methemoglobin in
blood. Clin. Chem. Acta., 30:679-682, 1970.
8. Levine, S., Chief of Pediatrics, Kaplan Hospital, Rehovot, person?!
communications.
9. Shearer, L.A., Goldsmith, J.R., Young, C., Kearns, O.A., and Tomp-
lin, B. Methemoglobin Levels in infants in an area with high nitrate
water supply. Am. J. Pub. Health 62:1174-480 (1972).
10. Shuval, H.E., and Gruener, N. Epidemiological and toxicological
aspects of nitrates and nitrites in the environment, Am. J. Pub.
Healh 62:1945-1052. (1972).
49
-------
SECTION IX
EPIDEMIOLOGICAL STUDY OF THE ASSOCIATION BETWEEN NITRATES IN
DRINKING WATER AND METHEMOGLOBIN LEVELS IN INFANTS
GAZA AREA
INTRODUCTION
On completion of the epidemiological study in the Rehovot area it
became apparent that very few infants in that area were actually exposed
to nitrates in drinking water in the form of powdered milk formula since
nutritional regime is rarely practiced in Israel today.
Children receiving only cows' milk or breast fed did however receive
tap water supplements amounting to about 10% of their total liquid intake
in winter and as high as 50% in summer.
Previous studies indicate however that only infants exposed to
appreciable quantities of high nitrate tap water made up as powdered
milk formula could be considered as candidates for raised MetHb levels.
It is still a mute question as to the role of the powdered milk per se
in the etiology of methemoglobinemia. In order to study the possible
association between nitrates in drinking water consumed in powdered milk
formula and raised MetHb levels in infants, an area where this type of
infant feeding is practiced had to be found.
The Gaza area was considered to be favorable for such a study
since preliminary investigations indicated that well over 50% of the
infants up to two years of age received powdered milk formula with tap
water. In addition there were many local areas in Gaza with high nitrate
concentrations in the drinking water. A survey of the local wells indi-
cated that many of them had as much as 135 mg/1 of nitrates as NO or
three times that of the recommended maximum concentration.
METHODS
In 1972 and 1973 a field study was mounted in the Gaza region with
the full cooperation of the local public health services, after numerous
administrative difficulties which prevented an earlier start had been
overcome.
50
-------
1.51—
1.0
CD
O
X
I-
Ul
0.5
0-3 4-6 7-9 10-12 12
AGE IN MONTHS
Figure IX-1. METHEMOGLOBIN LEVELS IN INFANTS BY AGE GAZA
51
-------
RESULTS
Age
As had been found in our previous study the MetHb level in the first
three months of life was higher than after that period regardless of ex-
posure to nitrates, as can be seen in Table IX-1 and Figure IX-1. The
MetHb level for the 45 infants three months of age or less is 1.25% as
compared to about .8-.9% after that age. These differences are statis-
tically significant at the p = 0.005 level.
Table IX-1. METHEMOGLOBIN LEVELS IN INFANTS BY AGE
Age in Months n MetHb%
0-3 45 1.26
46 106 0.81
7-9 109 .92
10 - 12 67 .88
13 + 91 .79
All 418 .89
Milk regime
For the purposes of this analyses three categories of milk regimes
were used.
a) PowderedfcmiIk only.
b) Powdered*miIk and other forms of milk.
c) Other forms of milk only.
It is assumed that infants consuming only powdered milk would have
the highest nitrate intake from tap water while those on the mixed powd-
ered milk and other milk regime would receive somewhat less tap water
and therefore less nitrates. Those receiving no powdered milk would
consume the least amount of tap water and would not therefore be exposed
to nitrates if present in the water.
Table IX-2 (total column 1-3) shows that infants on "powdered milk
only" regime have higher MetHb levels than those on the mixed regime,
while the latter show higher MetHb levels than those consuming no
powdered milk. However, a fuller analysis requires looking at each
milk regime category in terms of the potential nitrate exposure from
tap water in each individual case (see next paragraph).
52
-------
Ul
CO
Table IX-2. MEAN METHEMOGLOBIN LEVELS IN INFANTS ON DIFFERENT MTLK
REGIMES CONSUMING WATER OF VARYING NITRATE CONCENTRATIONS
GAZA
Milk Regime
a) Powdered
milk only
b) Powdered
plus other
a+b) Powdered
milk only
or some
c) No powdered
milk
a-c All
1
Low
Nitrates
n
7
90
97
26
.23
MetHb
.69
*(.25)
.73
(.46)
.73
(.45)
.79
(.80)
.74
(.54)
2 3
Medium High
Nitrates Nitrates
n MetHb n
12
(
58 1
(1
70 1
(1
15
(
85
(1
.84 22
.55)
.08 124
.26)
.04 146
.17)
.58 54
.28)
.96 200
.08)
MetHb
1
(2
(1
1
(1
(
(1
.37
.80)
.94
.08)
.01
•47)
.83
.55)
.96
.28)
2 + 3
Medium + High
Nitrates
n MetHb
34 1
(2
182
(1
216 1
(1
69
285
(1
.18
.27)
.99
.14)
.02
.37)
.77
.51)
.96
.23)
1-3
Total
n
41
272
313
95
**
418
MetHb
1.12
(2.03)
.90
(.97)
.93
(1.16)
.78
(.60)
.89
(1.06)
* Number in C ) is standard deviation
** Includes 10 cases of nitrates unknown
Low nitrates - 44 mg/1 NOs or less (mean =32.5 mg/1)
Medium nitrates - 45-55 mg/1 N03 (mean = 50 mg/1)
High nitrates - 56 mg/1 N03 and greater (mean * 87 mg/1)
-------
Nitrates in drinking water
For the purposes of this analysis three nitrate concentration cate-
gories were established:
1. Low nitrates: 44 mg/1 N03 or less (mean =32.5 mg/1)
2. Medium nitrates: 45-55 mg/1 N0_ (mean = 50 mg/1)
3. High nitrates: 56 mg/1 NO, and greater (mean 87 mg/1)
The low nitrate category represents water within the limits of the
current recommended maximum for drinking water, i.e. 45 mg/1 N0_. The
mean given is the actual mean nitrate level in the drinking water of the
123 cases examined in this category. The medium nitrates category repre-
sents the border area immediately above the present standard, while the
high nitrates category includes all 200 cases above 56 mg/1 NO,, with a
mean twice that of the standard. In that category 58 cases were exposed
to nitrate concentrations greater than 88 mg/1 with a mean of 135 mg/1.
As can be seen from Table IX-2 and Figure IX-2, there are essentially
no differences in MetHb levels among infants in the low nitrate category
regardless of the milk regime. The mean MetHb for the low nitrate group
being .74%, the small differences seen are not significant statistically.
In the medium nitrate group there appears to be a raised level of
MetHb in the two groups consuming powdered milk as compared to the group
consuming only "other milk" in the medium nitrate category and as compared
to the low nitrate group.
The trend that is first indicated in the medium nitrate group becomes
clearer in the high nitrate group with those infants consuming "powdered
milk only" showing a raised MetHb level of 1.37%, almost twice that of
the low nitrate group. Here it also appears that those consumine only
powdered milk show higher MetHb levels than those on the mixed regime,
while those on the mixed regime have higher MetHb levels than infants
consuming only "other" forms of milk.
For the purposes of statistical analysis it was found necessary,
however, to combine the groups consuming only powdered milk (a) with that
consuming powdered milk as well as other forms of milk (b) so that the
numbers of infants in each category would be larger. This was considered
feasible since there were no statistically significant differences be-
tween the two groups. Statistical tests show that infants consuming some
amount of powdered milk made up with water containing high concentrations
of nitrates (3 a+b) had significantly raised MetHb levels as compared to
those in the low nitrate area regardless of the type of milk consumed
(1 a-c) with p= 0.035.
There were similar differences in the medium nitrate group. Those
consuming "powdered milk" (2 a+b) were compared to all infants in the
low nitrate group (1 a-c) with p = 0.04.
54
-------
1.5
1.0
ffi
O
_i
to
O
X
t-
UJ
0.5
Powdered milk only
Powdered milk + other milk
Other milk only
(408)
MEAN: 32.5 mg/l
a b c
MEDIUM
NITRATES
50 mg/l
\\
FIGURE :lX-2.
MEAN METHEMOGLOBIN LEVELS IN INFANTS ON DIFFERENT MILK
REGIMES, CONSUMING OF LOW AND MEDIUM TO HIGH CONCENTRA-
TIONS OF NITRATES - GAZA.
55
-------
The infants in the medium and high nitrate group were also pooled
for further statistical analysis since there were no statistical differ-
ences between those two groups.
Infants consuming powdered milk in the medium-high nitrate group
(2+3-a+b) had a significantly higher mean MetHb than those consuming no
powdered milk in the same nitrate group 2+3 c) with p = 0.02; as well as
when compared with all infants in the low nitrate group (1 a-c) with
p = 0.003. The high level of significance in the difference between
these last two groups is noteworthy.
DISTRIBUTION OF METHEMOGLOBIN IN INFANTS
An analysis of the distribution of MetHb levels of infants was made
according to milk regime and exposure to nitrates. For the low nitrate
group all milk categories were combined while for medium+high nitrate
group all infants consuming powdered milk to any extent were combined.
The low nitrate category is the same as in Table IX-3 while medium and
high nitrate groups were pooled since both groups showed raised MetHb
levels.
The MetHb levels are divided into two categories. The first 0-1.7%
and the second 1.8% and over. This second group was selected on the
assumption that 1.8% MetHb or over represents significantly raised levels
of MetHb. The mean MetHb levels found in all infants in the low nitrate
group was .74% with a standard deviation of .54%. It can be assumed that
infants with MetHb levels of 1.8 or greater (i.e., the mean plus 2 S.D.)
can be categorized as having significantly raised levels of MetHb. From
Table IX-3 it can be seen that 3.2% of the infants in the low nitrate
group show raised MetHb while in the medium and high nitrates group 10.7%
of the infants consuming powdered milk have raised MetHb levels.
Only 4.6 of those in the medium and high nitrate groups consuming
no powdered milk show raised MetHb levels. These differences are signifi-
cant at the 0.05 level.
Table IX-3 DISTRIBUTION OF METHEMOGLOBIN LEVELS IN INFANTS EXPOSED
TO HIGH AND LOW CONCENTRATIONS OF NITRATES IN DRINKING
WATER - GAZA
Milk
Regime
Low Nitrates all
Medium + Powdered
High Milk
Nitrates No powdered
Milk
All
0-1.7% MetHb
n %
1.8%+ MetHb
n %
Total
117
193
65
259
96.8
89.3
95.4
90.9
6
23
3
26
3.2
10.7
4.6
9.1
n
123
216
69
285
100
100
100
100
56
-------
OTHER VARIABLES
Information was gathered for each infant concerning diet, in particu-
lar the consumption of high nitrate foods such as spinach and foods rich
in ascorbic acid which is an antidote for methemoglobinemia. No signifi-
cant findings evolved from an analysis of such dietary variables.
A further analysis controlling for age of the infants could not
provide significant findings.
In another phase of this project we were able to show that tap water
intake is higher in the hot summer months. An analysis of MetHb levels
by month of the year did not however provide any significant findings.
As an additional control group, 195 infants were tested for MetHb
levels in Jerusalem where nitrate levels in the water supply are very low.
The mean MetHb found was .79% which is similar to the MetHb of infants
in low nitrate areas in Gaza - .74%.
It is noted that the mean MetHb levels of the low nitrate group
in Gaza and the control group in Jerusalem are lower than the means found
in the Rehovot area study. These slightly lower MetHb levels are a re-
sult of refinements in the test method which developed in the period
since the original study was initiated in 1970 and 1971.
DISCUSSION
The Gaza study has provided some further confirmation that infants
in areas with water supplies having concentrations over 45 mg/1 of
nitrates (as NO ) who consume appreciable amounts of tap water in powd-
ered milk formula show raised MetHb levels. Although no clinical cases
of infant methemoglobinemia were revealed among 285 infants in the medium
and high nitrate areas, 26 of them showed significantly raised MetHb
levels of over 1.8%, of them 23 (10.7%) recieved either only powdered
milk formula or powdered milk formula in addition to other types of milk.
Only 3.2% of the infants in the low nitrate group showed raised MetHb.
Infants in the low nitrate group show similar MetHb levels regardless
of milk regime while infants who consume powdered milk in the medium
nitrate group with a mean nitrate concentration of 50 mg/1 in the water
have a significantly raised mean MetHb level. This difference appears
more clearly in the high nitrate group. It is worthy to note that the
mean MetHb level of the 22 infants on "milk powder only" and who received
water having high nitrate concentrations (mean 87 mg/1 of NO,) had a mean
MetHb level of 1.37% as compared with a mean of 1.30% for the 104 hospital-
ized infants exposed to approximately the same level of nitrates in their
powdered milk formula reported upon elsewhere in this report.
The possibility that infants in the medium and high nitrate areas
consuming powdered milk were younger than the population studied and
57
-------
therefore normally might be expected to have higher MetHb levels was
examined. Of the infants in the medium and high nitrate area 10.6% were
three months or younger and 33% were six months or younger as compared
with 10.8% and 36% for the total sample studied. Thus, there is no
indication that the higher MetHb levels in the infants consuming powdered
milk mixed with medium or high nitrate water could be explained by age
differences.
It was impossible in this study to clarify the role of milk powder
per se in conjunction with high nitrate water in the etiology of raised
MetHb levels. Undoubtedly infants on milk powder regimes consumed the
highest amounts of high nitrate tap water. Our studies on the consump-
tion of drinking water by infants do indicate that during the hot summer
months as much as 50% of the total liquid intake could be in the form of
supplemental tap water. In low nitrate areas, milk regime had no influ-
ence in MetHb levels, however.
The Gaza study appears to provide support for the present maximum
recommended standard of 45 mg/1 of NO, in drinking water. The fact that
the first signs of raised MetHb levels clearly appear in infants exposed
to water just above the standard (45-55 mg/1 N0_) suggests that little
if any safety factor is provided by this standard.
The full health significance of slightly raised MetHb levels such as
reported upon here is yet to be established. Whether such sub-clinical
methemoglobinemia is deliterious in itself or whether such exposure is
only of importance to the extent that clinical cases of the disease develop
requires further study.
The question of why; and how subclinical methemoglobinemia develops
in certain individuals into the clinical form of the disease still remains
unanswered. Other possible direct toxic effects of nitrates and nitrites
cannot be overlooked as well. These will be discussed elsewhere in this
report.
In conclusion, we must point out that the association between the
exposure to nitrates in drinking water in the form of powdered milk formula
and raised MetHb levels in infants has been demonstrated even though
certain inconsistencies associated with a field study of this type are
apparent.
Even water containing nitrates slightly above the current standard
is not free from suspicion. Such findings even in the absence of clear-
cut clinical cases of the disease must be considered as further evidence
supporting the current recommended standard of 45 mg/1, in those areas
where infants consume appreciable amounts of tap water in the form of
powdered milk formula. One might even question the degree of protection
provided by that standard when large population groups are involved
since there is evidence from this study that little, if any, safety factor
is provided by it.
58
-------
SECTION X
THE EFFECT OF CHANGES IN NITRATE CONCENTRATION IN DRINKING
WATER ON METHEMOGLOBIN LEVELS IN INFANTS
A CONTROLLED HOSPITAL STUDY
Under normal field conditions many difficulties hamper the possibility of
detecting a dose-response relationship between nitrate intake and methemoglo-
bin levels especially in the low range. To overcome some of these problems, a
controlled experiment was carried out in a hospital. In this study, we attemp-
ted to determine the threshold value of nitrates in water which can cause a
significant increase above "normal" methemoglobin (MetHb) levels in infants.
METHODS
A controlled experiment was carried out in the pediatrics ward of the
Hillel Joffa Hospital, Hadera. In the course of the national survey it was
revealed that this hospital is normally supplied with high nitrate content
water. Patients, however, come from both low and high nitrate areas. For
five days, 104 infants ranging from one week to ten months were exposed ex-
clusively to water whose nitrate content was exactly controlled. As was the
normal practice in the ward, the infants received mainly formula prepared from
milk powder, a fact which increased their water intake as compared with fresh
whole milk feeding which is the routine diet regimen in Israel.
The exposure schedule was as follows:
1. First day: Low nitrate content (mean 15 mg/1)
2. Second day: High nitrate content (mean 108 mg/1)
3. Third day: High nitrate content (mean 108 mg/1)
4. Fourth day: High nitrate content (mean 108 mg/1)
5. Fifth day: Low nitrate content (mean 15 mg/1)
The high nitrate water was from the local well which normally supplied
water to the hospital, and the low nitrate water was specially brought from
a distant one. Water was kept in special containers and the infants were
supplied solely from this source. MetHb levels in infants were measured by
a sensitive method previously reported(l). Finger flood samples were taken
in the morning after the infants had been exposed to the specific water for
24 hours. Each infant was checked and followed individually. The results
were statistically analyzed according to Wilcoxon(2). It should be noted
here that the District Public Health Office had recommended that the hospital
alter its water supply even before this study was initiated, and this has
meanwhile been done.
RESULTS
The infant population of 104 consisted of 56% females and 44% males.
Three quarters were hospitalized because of intestinal problems and one-
third of them received infusions to provide extra fluid intake. Exact
59
-------
Table X-l. METHEMOGLOBIN LEVELS IN INFANTS EXPOSED
TO HIGH AND LOW NITRATES IN WATER
Day Nitrate No. of Infants Mean of MetHb in
percent
0.89
1.30
0.91
0.93
0.80
1
2
3
4
5
low
high
high
high
low
87
93
85
63
75
Table X-2. DISTRIBUTION OF CHANGES IN METHEMOGLOBIN LEVELS
BETWEEN DA^S OF VARYING EXPOSURE TO NITRATES IN WATER
Compared
1
2
3
4
1
- 2
- 3
- 4
- 5
- 5
Number of
Infants
79
80
72
59
66
d
n
4
8
8
5
10
= o
%
5
9
11
8
15
d+
n %
46
35
28
88
22
59
44
37
18
32
d
n
29
38
36
36
34
P
% increase decrease
37 0.02
48
52
61
51
-
0.
0.
0.
0.
24
14
02
26
60
-------
records of the water intake could not be taken. Table X-l represents
the means of MetHb levels during the five days of the experiment. The
first stage of the low level nitrate intake was intended to reveal the
individual level each infant would show when little nitrate was included
in its diet. It is suggested that only the addition to this baseline
level should be ascribed to the exposure to nitrates on the second,
third and fourth day. The additional low level of nitrate of the fifth
day was intended to reveal in what degree MetHb levels in the infants
would react to the cessation of nitrate exposure.
There was a significant rise in the mean MetHg levels in the second
day compared to the first one, i.e., following the first exposure to
nitrate. There were three cases of massive rise in MetHb levels, all
between the first and second day. The rises amounted to 5.3%, 10.4%,
and 14.2%, reaching levels of 6.9%, 13.9%, and 15.9%. The mean MetHb
level on the third day decreased almost to the original level in spite
of the fact that the high exposure continued. It remained constant on
the fourth day (high nitrate intake) but dropped even lower than the
first day on the fifth day (low nitrate again).
The Wilcoxon test was used to analyze whether the shift in MetHb
levels for the series of individual infants on days of varying exposure
to nitrates in drinking water was truly significant. Examining the
difference between the means of each day might lead to misleading results
since a few infants showed relatively large changes in MetHb levels.
By this method each child is scored according to its value relatively
to the other observations. In Table X-2 this analysis is presented and
it can be seen for example that between day one and two 59% of the in-
fants showed an increase in MetHb levels (d+) as compared to only 37%
showing a decrease (d-). The increase in MetHb levels between days one
and two, resulting from the first exposure to nitrates was significant.
The decrease in MetHb levels between days four and five after the infants
were once again supplied to low nitrate water was also significant.
The above observations were made on 104 infants even though not all
infants were tested every day since some left the hospital and others
were missed for various technical reasons. There were 57 infants whose
blood was taken on each of the five days. The findings of these 57 in-
fants reveal the same tendencies as the whole population, though in a
more extreme pattern.
DISCUSSION
Epidemiological field surveys of methemoglobinemia induced by nitrates
in drinking water face a number of difficulties: 1) Nitrate levels in the
water vary over a wide range even at a single source(3). 2) Excessive
boiling of water'to be used for powdered milk formula can increase the
nitrate concentration. 3) Measurement of methmoglobin levels at a short
interval after nitrate intake is generally not feasible. This last point
is particularly important since elevated MetHb induced by nitrates returns
61
-------
to normal relatively quickly. It has been shown that in each 90 min.
period the level drops to 50%(4). We were able to control at least
these three variables in this study carried out in a hospital pediatrics
ward.
The results show that although the nitrate level in the water was
more than double the recommended standard for drinking water and the
infants drank the water for three consecutive days, essentially no clini-
cal methemoglobinemia developed. Most of the infants reacted to the
increase in the nitrate level between the first and second days by a
parallel increase in MetHb level. In spite of the continued high intake
of nitrates on the third and fourth days, there was an apparent drop in
the mean MetHb level on the third and fourth days. More cases showed
a fall in MetHb than an increase between the third and the fourth days.
Although these changes could not be shown to be statistically signifi-
cant which may be due to the large variability and the small sample, it
may be a hint at a mechanism of adaptation which reacts quite rapidly to
increased exposure of nitrates. There is a possibility that such a mech-
anism may be already developed among infants coming from high nitrate
areas and this converts the problem to much more complex question.
The day after the exposure to high nitrates in water stopped, MetHb
levels fell to the initial pre-exposure level. This fact demonstrates
the rather fast reactions of MetHb levels to changes in nitrate levels
in the diet. While previous reports indicate that several weeks of expo-
sure to high nitrate water is required before raised MetHb levels are
detected, our findings indicate that at least under certain conditions,
such long "incubation periods" do not necessarily apply. The prevalence
of gastroenteritis among the hospitalized infants might explain their
predisposition to developing high levels of MetHb after one day of expo-
sure to high nitrate water.
This work indicates that nitrate levels in drinking water of about
100 mg/1 can cause a significant increase in infant MetHb levels. If
this exposure is stopped recovery is rapid. Despite the fact that no
clinical methemoglobinemia developed among the 104 infants studied, the
question as to the possible long-term effects resulting from continuous
exposure to elevated nitrate levels in water used in infant formula has
not yet been determined. The possibility of an adaptation mechanism
remains to be elucidated as well.
62
-------
REFERENCES
1. Hegesh, E., Gruener, N., Cohen, S., Bochkovsky, R., Shuval, H.I.
A Sensitive micromethod for the determination of methemoglobin in
blood. Clin. Chim. Acta. 30:679-685, 1970.
2. Siegel, S. Non-parametric Statistics, McGraw-Hill Book Company,
Inc., New York, p 75-83, 1956.
3. Saliternik, C. Ground water pollution by nitrogen compounds in
Israel, in Advances of Water Pollution Research, Ed. S. H. Jenkins,
Pergamon Press, London, p 171-179, 1973.
4. Gruener, N. and Shuval, H.I., in Environmental Quality and Safety,
Vol. II (ed. F. Coulston and F. Roste), Georg Thiem Verlag, Stutt-
gart, and Academic Press, New York, p 219, 1972.
63
-------
SECTION XI
SURVEY OF LIQUIDS INTAKE IN INFANTS
Limited information exists in the literature concerning water intake
in infants. Such information may be of importance for a variety of rea-
sons: 1) Evaluation of the risks to infants and children from exposure
to harmful chemicals in water. 2) Intake of nutritional constituents in
water which may be of health importance. 3) The physiological importance
of sufficient water intake particularly in hot areas.
One of the main hypothesis in our previous study which indicated no
difference in methemoglobin levels between the study area (high nitrates
in water) and control areas was that in both cases there is essentially
low exposure to nitrates from drinking water due to the normal dietary
practices in Israel today. As was reported, 96% of the infants included
in the Rehovot area study were either breast-fed or received whole cow's
milk as their main source of liquid intake. The use of powdered milk
formula has practically disappeared in the Jewish population of Israel.
Tap water is given to such infants as "tea" or sugar water or plain tap
water as a supplement only. No reliable information was available from
pediatricians, or from pediatric departments at several hospitals as to
the amount of supplemental water usually given to infants in the 1-6
months age group. Therefore it was difficult to evaluate the contribution
of supplemental water as a nitrate source for infants whose main liquid
source contains no tap water. To answer this question we initiated a
survey of water intake among infants in the Rehovot-Rishon-Le Zion area
(high in nitrate).
METHOD
A liquid intake survey was carried out directly with the mothers
under the supervision of our staff nurse. We designed a detailed record
book with one page for each day of the month. The page was divided to
ease the recording of the information for different liquid and food
types and also by hours of the day - three sections and one section for
the night. Before starting the study the mother was interviewed by a staff
nurse and was instructed in the use of the record book.
Each week, each mother was visited by the nurse who inspected the
record book and advised on any problems that may have arisen. The daily
amounts of liquid intake were recorded differentiating between milk and
different "types" of tap water. Other sources of liquids were negligible.
One hundred fifteen infants were included in the survey between the ages
of one to five months, who were followed for one month periods. The pre-
test was made in January 1972 and then the study continued for a 12-month
period starting June 1972.
64
-------
RESULTS AND DISCUSSION
There is a marked decrease in the unit liquid intake (ml/kg of body
weight/day) with age (Table XI-1).
Table XI-1 LIQUID INTAKE IN INFANTS
AGE 1 TO 5 MONTHS
ml/kg/day
12345
Age in Months
Mean intake 104.6
n 24
125.30
37
102.10
33
86.9
8
69.2
13
This decrease parallels the increase in the body weight of the infants.
A relationship between water intake and the monthly average ambient tempera-
ture (Figure XI-1) was noted, with the highest water intake occurring during
the hottest months (July and August) and with the lowest in cold season. The
correlation between age and the specific liquid intake was found to be
significant with a correlation constant of 0.65. Exactly the same figure
was found between liquid intake and body weight.
150
LU
^ -SlOO
0)
LJ
2- £
tl 50
ID
Temperature
>< Mean Total Liquid Intake
^*
\
*S-
Wafer Intake
/x
0 —
VI VII VIII IX X XI XII I II III IV
1972 1973
MONTHS OF YEAR
25
z
o
20 -
LU O
tr
15 UJ
a.
Z.
LJ
10
Figure XI-1.
TOTAL LIQUID AND WATER INTAKE IN INFANTS 1-5 MONTHS OLD,
ACCORDING TO THE MONTHS OF THE YEAR.
65
-------
Month
Table XI-2. LIQUID INTAKE AND TAP WATER INTAKE
IN INFANT ACCORDING TO MONTH OF YEAR
n
Average age
Month
mean liquid intake
ml/kg/day
mean water intake
ml/kg/day
1972
January
June
July
August
September
October
November
December
1973
January
February
March
April
May
10
12
7
5
12
10
7
8
10
9
6
10
9
2.3
3.0
2.6
2.4
2.5
2.4
2.7
3.1
2.5
2.7
3.0
2.3
2.2
113
98
116
129
110
123
106
121
102
128
113
108
118
12
27
57
38
19
17
9
7
10
14
11
19
23
Mean
113
66
-------
The total liquid intake did not change and remained constant through-
out the year. Analysis of the results in relation to ethnic origin showed
a small difference between Ashkeasim and Sepharadim infants which may be
attributed to cultural factors. Only two infants were fed with powdered
milk formula (both new in Israel coming recently from the U.S.A.)- As
can be seen from Table XI-4 their liquid intake was similar to the majority
of the infants who were fed on cow's milk. A minority was breast-fed, got
an extra liquid supply which amounted to a little higher than half of the
intake of bottle fed infants of the same age group. The extreme cases
recorded for the mean liquid intake ml/kg/daily were as high as 192 and
as low as 52. The highest mean water intake was 85 and a baby with no
water supplement was also recorded. The two infants fed on powdered milk
(almost only water as liquid intake) had mean liquid intake of 109 and
145 ml/kg/day, respectively.
Table XI-3 LIQUID INTAKE IN INFANTS ACCORDING TO ETHNIC GROUP
ML/KG/DAY
n mean intake
Ashkenazim 54 119
Sepharadim 46 102
mixed 15 121
Table XI-4 LIQUID INTAKE IN INFANTS ACCORDING TO NUTRITIONAL REGIME
ML/KG/DAY
n liquid intake
Breast Fed* 27 85
Pasteurized milk 86 118
Powdered milk 2 127
*only supplemental liquid intake recorded
.The unit liquid intake in the first months of life is higher than
during any other period of life. This is a physiological necessity and
may be even more extreme in hot dry climates, as infants may suffer a
severe loss of water and even in normal conditions infants insensible
water losses are higher in the first year(1,2). Such high unit liquid
intake may, however, endanger the child's health when the majority of
the liquid is water which contains factors that may be harmful to his
health.
Previous studies showed(3) that water intake in infants under one
year of age were between 20-50% of total liquid intake. This is a
little higher but not far from the results reported in this study. The
water intake throughout the year can differ between the winter and the
summer by a factor of 5. During the six cool months of the year the water
intake amounted -only to around 10% of the total liquid intake while during
67
-------
the four hottest months reached 30% and up to 50% at the maximum. In
colder climates such variations would not be expected as supplemental
water would not be an important constituent of the diet in the first six
months.
The observation that in spite of the large variations in water intake
the total liquid intake remained constant was found also by Galagan et al(4)
in their extensive study.
Such observation should be studied carefully in relation to its
effects on the nutritional status of the infant during the hot seasons,
since the remainder of the liquid intake is usually made up on milk which
for many infants is the sole source of nutrition.
c
REFERENCED
1. Pratt, E.L., and Snyderman, S.E. Pediatrics 11, 65, 1965.
2. Heeley, A.M. and Talbot, N.B. Am. J. Pis. Child., 90, 251, 1955.
3. Margolis, F.J., Teate, H.L., Weill, M.L., and Wilson, H.L. Water
intake of normal children, Science, 140, 890, 1963.
4. Galagan, D.J., Vermillion, J.R., Nevitt, G.A., Stadt, Z.M., and
Dart, R.E. Climated and fluid intake, Pub. Health Report, U.S.
72, 484, 1957.
68
-------
SECTION XII
ONTOGENETIC DEVELOPMENT OF NADH DEPENDENT METHEMOGLOBIN
REDUCTASE IN ERYTHROCYTES OF DIFFERENT SPECIES
INTRODUCTION
NADH methemoglobin reductase (MR) has been suggested as the main
reductive pathway of methemoglobin (MetHb) in the red blood cell(1,3).
This assumption was based mainly on the in_ vitro relative activity in
MetHb reduction. Correlation was also found between lower ability to
reduce MetHb in young infants and their low MR activity(3,4). We have
shown that the "normal" mean level of MetHb in infants up to three
months old is significantly higher than the mean of 6-12 month old in-
fants as well as that of adults(5).
Other studies in this laboratory carried out in different species
show that mice, guinea pigs and rabbits, which have higher MR activity
than rats, will show lower MetHb levels than rats when exposed to equal
doses of sodium nitrite(6).
Epidemiological surveys reveal a relatively low prevalence of
raised methemoglobin levels in infants exposed to nitrates in water.
A number of factors have been indicated as playing a role in the predis-
position to develop methemoglobinemia which include stomach pH, micro-
flora in the upper intestine, liquid intake per unit of body weight and
type of milk powder formula. All these are in addition, of course, to
the amount of nitrates actually ingested. In addition to these factors
we have hypothesized that sporadic clinical cases of methemoglobinemia
in areas with high nitrate concentrations in water may arise from hetero-
zygotes (in relation to MR deficiency). These cases generally can cope
with the normal factors which convert Hb to MetHb but succumb when exposed
to an abnormally large body burden of methemoglobinemia causing factor.
Such individuals would start to show symptoms at lower nitrate dose
levels than normal persons. Cases of both complete MR deficiency which
cause methemoglobinemia, an autosomal recessive disease, and partial MR
deficiency which has been shown to be susceptible to certain drugs, have
been reported(7,8).
This study follows the development MR levels from the fetus to adult
in both humans and rats, in anticipation that detailed information on the
ontogenetic development of this vital enzyme in both species will be of
value in understanding the pathogenesis of methemoglobinemia both under
field conditions arid in controlled laboratory studies.
69
-------
METHOD
Sixty-nine cord-blood samples were taken from regular normal births.
Finger blood samples were taken from 758 (435 Jews and 325 Arabs) in-
fants appearing for routine check-ups at mother and child clinics.
Ninety-nine adult blood samples were taken from blood bank donors.
Heparin or EDTA were used as anticoagulants and found to have no effect,
at concentrations used, on the enzyme activity. Samples were checked
within 24 hours. Storage at 4°C for this length of time did not lead to
changes in the enzyme acitvity. Sabra rats were used for the animal
studies. Enzyme assay was done by a modification of Hegesh method(9).
RESULTS
Methemoglobin Reductase Activity in Man. The known phenomenon of
lower activity in the newborn than adults has been confirmed. However,
in testing numerous infants and children of different ages between birth
and three years of age, a step-like pattern of MR activity increases
appears to exist (Table XII-1). The rather low fetal level increased
rapidly in the first days of life from 1.5 to 2.3 mymole/min/mg.Hb.
This level remains for about 14 days. In the next step, 2 weeks to two
months of age, there was an increase of 20%. Another increase of 15%
brings the enzyme level at the age of 2-6 months to almost adult level.
In the next step, 6-24 months, the enzyme level showed a plateau which
is unexpectedly higher than adult level by 20% and drops down after two
years to the adult level. An additional survey, which included 325
Arabs, showed a similar age profile with an increase to a plateau and a
decrease after about two years. The MR activity in this population was
lower compared to the Jewish infants and the plateau was shifted to
slightly higher age.
fc
•y
Methemoglobin Reductase Activity in the Rat. While in humans the
enzymatic level increases from the fetal to the adult level by 100%,
in the rat the fetus has an MR activity which is about ten times higher
than the adult (Table XII-2). This high level MR at birth decreases
sharply in about two months, to the adult rat level. Some differences
between males and females were observed in the younger age groups but
this apparently disappears in the matured animals. Age profiles were
followed also in the mouse and hen and the same trend as in the rat was
found but in less extreme fashion.
DISCUSSION
Erythrocytic enzymes show different ontogenetic changes. In a
survey done recently in human fetus erythrocytes(9) a number of enzymes
such as hexokinase or Mg Atpase are higher in the fetus than in the
newborn. Others, like NADH (and NADPH) MR and gluthatione peroxidase,
are lower during the intrauterine life than after birth. The reasons
70
-------
Table XII-1. LEVELS OF METHEMOGLOBIN REDUCTASE ACTIVITY
IN HUMANS AT DIFFERENT AGES
Age
(Days)
Chord Blood
0-7
8-14
15
31
46
61
76
91
121
151
181
271
361
541
720
- 30
- 45
- 60
- 75
- 90
- 120
- 150
- 180
- 270
- 360
- 540
- 720
- 1080
Adults
n
69
83
9
17
13
22
18
19
34
28
30
49
34
34
20
23
99
m y moles/min./mg.Hb
Jews Arabs
mean S.D. n mean
1.58
2.30
2.31
2.78
2.68
2.76
3.14
3.41
3.18
3.16
3.24
3.94
3.81
3.96
3.55
3.33
3.34
0.54
1.25
1.29
0.21
0.98
1.35
1.62
1.62
1.47
1.11
1.58
1.60
1.49
2.18
2.07
1.52
0.98
-
-
5
5
3
8
27
25
36
85
54
61
12
4
-
-
2.18
2.24
2.42
2.98
2.92
2.66
2.99
3.29
3.21
3.83
3.43
3.13
S.D.
—
-
0.86
0.92
0.85
1.84
1.32
1.18
1.76
1.45
1.21
2.19
1.00
0.64
-------
Table XII-2. METHEMOGLOBIN REDUCTASE ACTIVITY IN RATS AS FUNCTION OF AGE
S3
Age (Days)
Foetus
3
8
15
25 (males)
25 (females)
40 (males)
52 (males)
Adults (females)
n
12
12
12
11
10
10
9
10
6
Enzyme Activity (Mean) ± S.D.
my moles/min./rog'Hb
23,82 ± 2,26
15,42 ± 2,13
10,36 ± 0,63
6,56 ± 0,92
2,98 ± 0,80
4,78 ± 0,72
2,78 ± 0,35
2,65 ± 0,36
1,91 ± 0,59
-------
OJ
± 4.0
>
>
I-
o
UJ
(A
3.5
u
O
U
DC
m
o
_i
(9
O
z
ui
X
UI
z
JO
I
O)
c
E
in
3.0
2.5
2.0
1.5
Foetus
I
I
9 12 15 18
AGE IN MONTHS
27
FIGURE XII-1. CHANGES IN METHEMOGLOBIN REDUCTASE LEVELS WITH AGE - HUMANS
-------
>
>
p
u—-
26
if
UJ
M
10
o
UJ
u
3:
*
Foetus
(10)
-*-
ess;
_j i i
10 20 30 4
AGE (days)
50
Adults
FIGURE XII-2. CHANGES IN METHEMOGLOBIN REDUCTASE LEVELS WITH AGE - RATS
-------
for these differences are not clear. The elucidation of the mechanism
of these metabolic changes will assist in the understanding of the unique
fetal physiology and in formulating the precautions that should be taken
when pregnant women or newborns are treated with certain drugs.
The phenomenon that MR activity during certain periods of childhood
is higher than among adults has been recently reported(10)- Changes in
Hb susceptibility to oxidation with age or metabolic changes which lead
to increase in production of methemoglobinemia causing factors may lead to
parallel changes in MR as an adaptation defense mechanism. A similar
survey done in Malaysia among different races(11) found similar MR activity
in adults and newborns. The authors checked the possibility of detecting
heterozygotes by total enzyme activity evaluation. They concluded that
this approach is not feasible as the lower limit of normal seems to overlap
with the activity found in trait carriers. We also came to these conclu-
sions as we found a wide range of enzymatic levels in the population. The
lag in the development of MR among Arab children may be associated with
nutritional factors or may be of genetic origin.
Agar and Harley(12) reported recently on an MR survey among differ-
ent species and the change in activity which occurs with age may have
ecological meaning which is still obscure. MR activity was found to
raise from 1.58 in the human fetus erythrocyte to 3.90 at the age of
6-24 months and then decrease again to the level of 3.34 which then
apparently remains constant throughout life. Very similar adult human
MR levels were found in several surveys done in different places of the
world among different races. Different species show different onto-
genetic pathway characters of MR development. In some, the enzyme
activity does not change from fetus to adult; others, like the rat, show
a sharp decrease with maturity. In man, MR activity doubles itself to
about adult levels in the first six months of life. Beyond the interest-
ing comparative question, these differences should be taken into account
when results of toxicological experiments, in respect to the question of
methemoglobinemia, are considered. The possibility that human cases of
clinical methemoglobinemia in infants exposed to high nitrate concentra-
tions in drinking water may be associated in part to MR deficiency still
remains.
75
-------
REFERENCES
1. Jaffe, F.R. In Biochemical Methods in Red Cell Genetics, ed. Yunis.
Academic Press, New York, 1969, p 231-253.
2. Scott, E.M. In Hereditary Disorder of Erythrocytes Metabolism.
E. Beutler, ed. Grune and Stratton, New York. 1968, p 102-113.
3. Ross, J.D. Deficient Activity of DPNH Dependent Methemoglobin
Diaphorase of Human Erythrocytes in Cord Blood Erythrocytes. Blood
21:51-62, 1963.
4. Bartos, H.R., Des Forges, J.F., and Clark, N. Erythrocyte DPNH
Dependent Diaphorase Levels in Infants. Pediatrics 37:991-993,
1966.
5. Shuval, H.I., and Gruener, N. Epidemiological and Toxicological
Aspects of Nitrates and Nitrites in the Environment. Am. J. Pub.
Health 62:1045-1052, 1972.
6. Gruener, N., and Cohen, S. (Unpublished data).
7. Scott, E.M. The Relation of Diaphorase of Human Erythrocytes to
Inheritance of Methemoglobinemia. J. Clin. Investigation 39:1176-1179,
1960.
8. Cohen, R.J., Sachs, J.R., Wicker, D.J., and Cenard, M.E. Methemoglo-
binemia Provoked by Malarial Chemoprophylaxis in Viet Nam. New Eng.
J. Med. 279-.1127-U51, 1968.
.v
9. Hegesh, E., Calmanovici, A., and Avron, M. New Method for Determin-
ing Ferihemoglobin Reductase (Methemoglobin Reductase) in Erythro-
cytes. J. Lab. Clin.Med. 72:339-344, 1968.
10. Vetrella, M., and Barthlemai, W. Erythrocyte Enzymes in the Human
Fetus. Morratshur Kinderheilk 119:265-267, 1971. .
11. Eng, L.L., Loo, M., and Fah, F.K. Diaphorase Activity and Variants
in Normal Adults and Newborns. Brit. J. Haem. 23:419-425. 1972.
12. Agar, N.S., and Harley, J.D. Erythrocytic Methemoglobin Reductase
of Various Mammalian Species. Experimentia 28:1248-1249, 1972.
76
-------
SECTION XIII
METHEMOGLOBINEMIA INDUCED BY TRANSPLACENTAL
PASSAGE OF NITRITES IN RATS
Since clinical methemoglobinemia from nitrates in water apparently
appears only in infants, one possible solution proposed for areas with
high nitrate water is to supply the infants with low nitrate content
water from alternate sources. This measure will not exclude the risk
of the exposure to nitrates in the prenatal state, i.e., the transfer
of nitrates or nitrites through the placenta to the fetus, where methemo-
globinemia may be induced.
The possibility of this occurring was tested on pregnant rats.
Nitrites were given to pregnant rats in their drinking water or by injec-
tion and subsequently nitrite and MetHb were assayed in the fetal blood.
Suckling rats whose dams received nitrites in their drinking water
showed no rise in MetHb levels. By contrast, the dams showed high MetHb
levels. This demonstrates that nitrites are apparently not transferred
in appreciable amounts to the suckling rats via the milk.
The transfer of nitrites to the fetus in_ utero and the production
of MetHb was tested in the following experiment^Pregnant white albino
rats were used. Each pregnant rat was weighed and anesthetized with
ether. From 2.5 - 50 mg/kg of sodium nitrite was given orally or injected
intraperitoneally to the pregnant rat and the kinetics of nitrites and
MetHb in the dams as well as in the fetuses were measured. Blood was
collected from the tail of the dam at regular intervals throughout the
experiment. After opening the abdomen the fetuses were removed serially
from alternate sides, at regular time intervals over a two-hour period,
the umbilical cords being cauterized. The fetuses were washed in saline
solution at 37 C and then decapitated. Blood was collected and the MetHb
and nitrite levels were measured. The micromethods developed by our
group for determining MetHb and nitrites (Sections IV and V) in blood
enabled us to carry out these experiments with the small amount of blood
available from each fetus.
The characteristic picture obtained is shown in Figure XIII-1.
After a 30 mg/kg dose of NaN02 per os to the pregnant rat, nitrite levels
rose in the fetal blood thougn with a lag of about 20 minutes behind the
dam. This rise in nitrite in the fetus was followed by a rise in MetHb.
The possibility that the placenta was damaged during our experiments
leading to increased permeability was excluded when sodium nitrite was
given to normal pregnant rats after labor had started. The first fetus
77
-------
showed a normal MetHb level of 1.2% MetHb while those which were born
after the chemical had been given showed a level of 10.1% MetHb and 1.2
Ug/ml of sodium nitrite in their blood. All births were unassisted.
Different concentrations of nitrites caused similar kinetic pictures
differing only in their timing and MetHb levels. Table XIII-1 shows
that the threshold of the effect was at a sodium nitrite dose of 2.3 mg/kg.
The increase in effect was steep with increased dosage.
Table XIII-1 BLOOD NITRITE AND METHEMOGLOBIN LEVELS AFTER INJECTION
OF DIFFERENT DOSES OF SODIUM NITRITE
dose
Peaks of MetHb Level
as percent of total Hb
Peak of nitrite level in
blood as NaNO_ ug/ml
2.5
5.0
10.0
15.0
20.0
25.0
30.0
mother
0.9
3.4
5.0
11.9
17.0
33,
40.
.2
.4
60.2
fetus
1.2
1.9
2.7
5.1
7.9
13.3
19.2
27.2
mother
0.0
3.9
6.9
8.9
10.8
21.7
25.6
32.5
fetus
0.0
traces
traces
0.4
1.2
5.9
6.9
9.4
Pregnant rats exhibited a higher susceptibility to nitrites than non-
pregnant rats in chronic and acute experiments.
In chronic feeding experiments pregnant rats exposed to 2000 mg/1
of NaN02 in drinking water or about 200-250 mg/kg/day developed severe
anemia with a mean of 10.3±1.5 g% Hb as compared with non-pregnant rats
exposed to the same levels on nitrites having means of 14.2±0.9 gm% Hb. The
control kept on tap water showed a mean of 14.3±0.8 g% Hb. All pregnant
rats (five) died within one hour when injected with doses of 60 mg/kg
while non-pregnant females survived such doses.
In general, nitrites given per os led to somewhat lower levels of
MetHb as compared with comparable doses given sub-cutaneously but the
kinetic picture was similar in both cases. The finding that the MetHb
peak could be detected in the fetus 45-60 minutes after injecting NaNO_
in the dam but not in the newborn several hours after birth indicates
that the MetHb reductive mechanism in the fetus and newborn rat is highly
effective. MetHb recovery rates were measured in rats after ceasing
the consumption of water which contained NaN02. The time which it takes
for the MetHb to be reduced to 50% of its initial level was found to be
around ninety minutes, independent of the initial concentration of MetHb.
NADH-dependerit MetHb reductase is claimed to be the main pathway of
MetHb reduction(2). The enzyme acting in the red blood cells was mea-
sured in adult rats and in fetuses. Similar determinations were made
78
-------
vo
Mithfmoglobin - pregnant rat
Z 2
o
z
f» - pregnant rat
Mtthtmoglobin - fttus»s
A
15
15
10
O)
o
5 E
V
Z
45 60 75 90
Time (min)
U-O«-O..O
105 120 135
Figure XIII-1. KINETICS OF NITRITE AND METHEMOGLOBIN IN BLOOD OF A PREGNANT RAT AND
THE FETUSES (30 mg/kg NaNO per os)
-------
in humans and human cord blood. Table XIII-2 shows that the rat fetuses
have MetHb reductase activity some ten times higher than adult rats have
or that found in human cord blood while human adult blood exhibits
one and one-half times the activity of the cord blood.
Table XIII-2 METHEMOGLOBIN REDUCTASE ACTIVITY IN THE FETUS
AND THE PREGNANT FEMALE - RATS AND HUMANS
Methemoglobin reductase
n mymoles/min/mg Hb
Pregnant rats 6 1.86±0.34
Rat fetus 10 17.4 ±1.3
Human pregnant females 49 2.38±0.78
Human cord blood 69 1.58±0.54
Enzyme activity was checked in a. blood sample of 10 yl which was
incubated in a mixture that contains 50 umoles citrate buffer (pH4.7),
EDTA 300 mymoles, K_Fe(CN), 56 mymoles, MetHb 288 mymoles. Total volume
is 0.6 ml. 6 °
The reaction is started by adding 120 mpmoles of NADH. The reduction
of MetHb to oxyhemoglobin is followed at 577 nm (Unicam SP 1800, band
width of Inm) . Calculations are based on the value of 42.0 for AeM at
577. Molecular weight of hemoglobin is taken as 66.000. Hb was measured
at 540 nm according to the cyanomethemoglobin method.
These findings point to the possibility that the human fetus might
have a weaker defensive mechanism to the intra-uterine exposure to nitrites
than that detected in rats.
The results underline the possible risk of intra-uterine methemo-
globinemia when water or foods containing nitrates are consumed during
pregnancy. It should be noted that certain foods such as spinach can
contain as much as 4000 ppm of nitrates, a major portion of which can.
be reduced to nitrites on storage under certain conditions(3). 200 ppm
of nitrites can legally be added to many "corned" meat products. It is,
however, premature to extrapolate from these acute animal experiments to
the situation that may exist with humans consuming MetHb inducing chemi-
cals in water or food. A study of MetHb levels in infants' cord blood
where the mothers come from areas with high and low nitrate levels in
their drinking water did not reveal any raised MetHb levels among the
150 cases studied.
80
-------
REFERENCES
1. Gruener, N. and Shuval, H.I., in Environmental Quality and Safety,
Vol. II (Ed. by F. Coulston and F. Korte), George Thiem Verlag,
Stuttgart and Academic Press, New York, p. 219, 1972.
2. Scott, E.M. in Hereditary Disorders of Erythrocyte Metabolism
(edit, by Beutler, E.), 102 (Greun and Stratton, New York, 1968).
3. Eisenberg, A., Wisenberg, E. and Shuval, H.I. The Public Health
Significance of Nitrates and Nitrites in Food Products. (Ministry
of Health, Jerusalem, Israel, 1970).
81
-------
SECTION XIV
THE EFFECTS OF THE ADMINISTRATION OF SODIUM NITRITE IN
DRINKING WATER ON PREGNANT RATS AND THEIR NEWBORN
Two experimental groups were used, each containing twelve pregnant
albino "sabra" rats. Group II was given 2000 mg/1 sodium nitrite and Group
III 3000 mg/1 in their drinking water. The control, Group I, of seven
pregnant rats, received tap water without added nitrite.
The pregnant rats that received nitrites showed an increase in meth-
emoglobinemia. Group II had a mean of 5.5% MetHb. The Group III mean was
24.0% and the mean of the Controls was 1.1%.
The pregnant rats that received sodium nitrite suffered from anemia in
a direct relationship to the concentration of the compound in their drinking
water. The Hb results are presented in Table XIV-1.
In view of the marked hematological effect of nitrite on the pregnant
rats, we measured red cell fragility with hypotonic solutions. Erythrocytes
from pregnant rats that received nitrite showed less fragility to hypotonic
than those from the control rats.
Table XIV-1 DISTRIBUTION OF HEMOGLOBIN DETERMINATIONS
IN PREGNANT RATS CHRONICALLY EXPOSED TO
SODIUM NITRITE IN DRINKING WATER
Group I Group II Group III
gm% Hb Percent Percent Percent
10< 0 45.2 52.4
lO.'l - 12.0 31.5 27.2 26.0
12.1 - 14.0 44.6 21.6 13.0
14.1 + 23.9 6.0 8.6
In experiments where the blood of rats was mixed with 0.0 to 0.9%
sodium chloride solutions (i.e., 0 to 100% normal saline) we obtained
typical sigmoidal fragility curves. In red cells, 50% hemolysis corres-
ponded (from the curves) with 25.4% normal saline for Group III,with 30.7%
normal saline for Group II and 37.8% for Group I, the controls. Thus, red
blood cells from the treated groups were more resistant to hypotonicity than
those from the control group.
It is possible that nitrite has some contact effect on the erythrocyte
in which the "weaker" cells only are affected, the "stronger" more resistant
cells remaining unaffected. Incubation of erythrocytes in_ vitro with
82
-------
sodium nitrite does not have any affect on the reaction of hypotonic os-
motic pressure.
MetHb is reduced in the erythrocyte to Hb with the aid of an
enzyme system utilizing DPNH for this purpose:
MetHb + DPNH •*• Hb + DPN
In the next experiments we test the possibility that the increase in
MetHb changes the metabolic picture in the red blood cell. It is assumed
that the DPNH/DPN ratio is small during the MetHb increase, due to the
rapid consumption of DPNH.
The reduction of the MetHb is rapid and uptake of DPNH outbalances
the rate of regeneration of DPNH in the glycolitic pathway. By way of
checking the DPNH/DPN ratio we assayed the pyruvate/lactate ratio in blood
from pregnant rats which, together with lactic dehydrogenase constitutes
the main system in the erythrocyte using DPNH/DPN.
Lactate + DPN -> Pyruvate + DPNH. Competition for DPNH between the
pyruvate/lactate and the MetHb reductase systems may increase pyruvate/
lactate ratio. The results are presented in Table XIV-2. The increase
in the ratio was attributable to increase in pyruvate (up to 4 times
more than in the control) as opposed to the small lactate increase. This
points to possible increased glucose metabolism to provide extra DPNH
supplies to the cells. There was a pronounced effect on mortality *unong
newborn rats of dams receiving 2000 mg/1 (Group II) and 3000 mg/1 (Group III)
in their water, particularly in the three-week period up till weaning.
Table XIV-2 PYRUVATE/LACTATE RATIO IN PREGNANT RATS
CHRONICALLY EXPOSED TO SODIUM NITRITE IN
DRINKING WATER
Group Pyruvate/lactate ratio
I (Controls) 0.013
II (2000 mg/1) 0.038
III (3000 mg/1) 0.042
The average litter in the control group contained 10 fetuses, 9.5
in Group II and 8.5 in Group III The mortality within the first three
weeks was 6% in the control as opposed to 30% in Group II and 53% in
Group III. Birthweights were similar with 5.5 gm for each group. However,
after the birth, newborn rats in Groups II and III lagged behind the con-
trols in their growth rates. For example, after one week the mean weight
was 16.5 gm in the control group, 12.0 gm in Group II and 9.5 gm in Group
III. After 21 days (at the end of the period of giving nitrites to the
dams), 51.5 gm mean weight in the control group, 29.5 gm in Group II
and 18.5 gm in Group III.
83
-------
Apart from the weights, a characteristic difference observed in the
experimental groups was that the fur thinned and lost its luster. After
separation from their dams and being put on water, there was an improve-
ment in growth in the experimental groups. At the age of 32 days, the
mean weight for the control group was 100 gm, for Group II, 67 gm and
for Group III, 39 gm. At 62 days the control group attained a mean weight
of 213 gm, Group I, 181 gm, and Group II, 172 gm.
During the whole period from birth to weaning, the newborn showed no
abnormally high MetHb. The mean hemoglobin of the newborn from the experi-
mental groups was low - about 20% of the control group.
84
-------
SECTION XV
INFLUENCE OF ASCORBIC ACID ON NITRATE-NITRITE
INDUCED METHEMOGLOBINEMIA
INTRODUCTION
In our epidemiological survey carried out in Rehovot, Rishon-Le-
Zion and Nes-Ziona, no significant difference in the mean MetHb level was
found compared to the control area. One of the explanations raised was
that the infants were fed some sort of antidote to methemoglobinemia.
It is known that ascorbic acid is an antidote to methemoglobinemia, and
as such is used in conjunction with the therapy of this disorder, espe-
cially in the congenital variety(l.,2) . It was found in the survey that
87% of the infants received citrus or tomato juice, and that 50% of the
infants up to the age of 60 days received additional vitamin C rich sup-
plements .
Although the possibility that ascorbic acid may be a useful prophy-
lactic agent against methemoglobinemia has been raised before(3), experi-
mental proof of the efficacy of this agent has been lacking.
In one study(4) ascorbic acid administered in physiologic doses (50
mg/kg/day) had no effect on methemoglobin levels of healthy and distressed
newborn infants. However, in this study MetHb levels were quite low to
begin with (less than 1%) even in the perinatally distressed infants whose
methemoglobin levels were slightly higher (0.95% as opposed to 0.755%
in normals). An experimental study by Kociba and Sleight(5) was done on
acute nitrite toxicosis in ascorbic acid deficient guinea pigs. In this
study subcutaneous administration of 50 mg/kg NaNO^ led to higher methemo-
globin levels in ascorbic acid deficient animals (Dlood levels of the order
of 0.35 mg%) than in controls (blood levels of the order of 0.88 mg%).
Blood nitrite levels were not measured in this study.
It was decided to experimentally approach the question of ascorbic
acid prophylaxis. If ascorbic acid can be shown to be a useful anti-
methemoglobin prophylactic agent, ascorbic acid dietary supplementation
may be indicated in areas with high nitrate loading in the environment.
To begin with, it was decided to study the effects of ascorbic acid on acute
nitrite induced methemoglobin, and to study the relations between blood
levels of ascorbic acid, nitrite, and methemoglobin in an uncomplicated,
well-studied experimental preparation.
MATERIALS AND METHODS
In all, six pairs of rats (albino Hebrew University Sabra strain)
were used. They were matched for sex and weight, with weights between
400 and 450 grams-. The basic experiment was done in pairs, each pair on
85
-------
a different day and using new reagents each day. One rat of each pair was a
control, the other the experimental animal, given the ascorbic acid. After
withdrawing a sample of tail blood for ascorbic acid determination, the
experimental animal was given 100 mg ascorbic acid in 1 cc of normal saline,
and the control given 1 cc of normal saline, both given intraperitoneally
(IP). After 1 1/2 to 2 hours, found to be the maximal absorption time in
unpublished observations from this laboratory, tail blood was again with-
drawn. Methemoglobin levels, ascorbic acid levels, and nitrites, were
measured. Both rats were then given IP NaNCL, 20 mg/kg solution. In all 6
pairs, methemoglobin and ascorbic acid levels were determined and in four
out of the six pairs nitrite levels were also determined at various inter-
vals after the injection of nitrites. Total hemoglobin levels were also
taken at the beginning and the end of each experiment from each animal.
MetHb, nitrites, and ascorbic acid were measured by the technique described
elsewhere in this report.
Results were plotted in two ways. 1) The controls were meaned, and the
experimental animals were meaned for each time interval and each variable
studied. 2) Since the experiments were carried out in a paired fashion, the
differences between control and experimental were meaned for each variable
and each time interval. Statistical analysis was carried out on the mean
differences thus determined. Statistical difference was determined by the
"t" test for paired variables.
RESULTS
Figure XV-1 shows the values of the variables at each time interval
after the injection of nitrite, with the controls meaned as a group, and
the experimentals meaned as a group. What can be seen is that the ascorbic
acid values remained relatively high in the experimental group (given the
ascorbic acid), and lo& in the control group. Not shown are the ascorbic
acid values 1-1/2 hours before the injection of nitrites, immediately prior
to the injection of ascorbic acid. These were all similar to the control
values shown on the figure.
It is noted that the highest value in the MetHb curve for the experi-
mental animal occurs later and is lower than the highest value for the
control animal. However, after one hour the two curves become very similar
to each other.
It is further noted that the nitrite curves seem to be parallel in
a way to the methemoglobin curves, except that their peaks are earlier.
The control nitrite peak was higher and earlier than the experimental,
just as in the case of the methmoglobin peaks. After 20-25 minutes the
two curves become quite similar to each other.
The data are shown analyzed in a paired fashion in Figures XV-2 and
XV-3. Figure XV-2 shows the mean differences (control minus experimental)
86
-------
FIG. 1.
MEAN ASCORBIC ACID/ M.ETHEMOGIOBIN AND NITRITE
LEVELS IN 6 RATS ADMINISTERED 20 mg/kg NoN02
AT TIME ZERO
ASCORBIC
ACID
mg 1,
METHEM06LOBIN
(%)
6(
4
3
2
1 1
0
20
15
10
5
r _ ^_________-_
*
-
-
C °"
s "~ o
f ^.\
/ -s^
- 1 /
1 /
1 /
I/
~y
NITRITE
jig N /ml
I
15 30 45 60
TIME (min)
•• EXPT.
o --o CONTROL
90
Figure M-l. MEAN ASCORBIC ACID, METHEMOGLOBIN, ND NITRITE LEVELS
IN RATS ADMINISTERED 20 mg/kg NaN02 AT TIME ZERO.
87
-------
FIG. 2
MEAN DIFFERENCE BETWEEN EXPERIMENTAL
AND CONTROL ANIMALS (CONTROL MINUS
EXPERIMENTAL) IN METHEM06LOBIN LEVELS
AT EACH TIME POINT STUDIES
m
8s
Z
O
o
o
2E
UJ
X
UJ
«
10
8
6
A
2
0
-2
-A
-6
•
H.
.
|
- /
- /
*
*
\
\
~
'/
f
| ._.
L \
-
-
» 1 1 1
«.
1
15 30 AS 60
TIME (min )
90
statistical significance p £ 0.05
shown are means ± 1 SO
Figure XV-2. MEAN DIFFERENCE BETWEEN EXPERIMENTAL AND CONTROL
ANIMALS (CONTROL MINUS EXPERIMENTAL) IN METHEMO-
GLOBIN LEVELS AT EACH TIME POINT STUDIES.
88
-------
FIG. 3
MEAN DIFFERENCE BETWEEN EXPERIMENTAL
AND CONTROL ANIMALS (CONTROL MINUS
EXPERIMENTAL) IN NITRITE LEVELS AT
EACH TIME POINT STUDIES
oc
H-
15 30 45
TIME (min)
* s fa f / s f ica I s ignificance p £0-05
shown are means + 1SD
Figure XV-3.
MEAN DIFFERENCE BETWEEN EXPERIMENTAL AND CONTROL
ANIMALS (CONTROL MINUS EXPERIMENTAL) IN NITRITE
LEVELS AT EACH TIME POINT STUDIES.
89
-------
between the methemoglobin values at each time interval after the injection
of nitrites. These are maximum at 15 and 30 minutes. After 60 minutes
the mean difference becomes small and insignificant. At 45 minutes the
mean difference is sizable, but not significant, probably due to the
smallness of the sample.
Figure XV-3 shows a plot of the mean differences in nitrite values
(control minus experimental) for each time interval following nitrite
injection. This curve resembles the methemoglobin curve in its biphasic
nature. There is a period of large difference followed by one of small,
insignificant difference. The highest difference in nitrite values occurs
earlier than the highest difference in methemoglobin values.
Thus, if one uses the mean difference between the control and the
experimental animals as an index of the protective effect of ascorbic
acid, the following may be said: 1) The main protective effect of ascorbic
acid appears in the first, and highest, portion of the methemoglobin forma-
tion curve, up to 60 minutes post nitrite injection. After that protection
seems to be lacking. The protective effect seen here was of the order of
magnitude of 6-7% methemoglobin less in the experiment than in the control
out of a total of 20% in the control. 2) This pattern seemed to parallel
the nitrite pattern, with maximum protective effect on the first and highest
portion of the curve, up to 20-25 minutes. After that the values between
control and experiment were very similar. It should be noted that the
difference in ascorbic acid was high between experimental and control,
at least up to 60 minutes post nitrite injection.
DISCUSSION
The mechanism of any protection given by ascorbic acid against nitrite
induced methemoglobinemia may be either direct or indirect. Direct reduc-
tion of methemoglobin by ascorbic acid certainly does occur, but is a
slower process than the enzymatic reduction(6). Indirect mechanisms in-
clude increase in reductase activity, blocking of nitrite binding sites,
and lowering of blood nitrite levels. From these data presented here
the question cannot be decided. However, there is suggestive evidence
that the ascorbic acid may act by lowering the peak nitrite levels.
There is a striking parallel between the nitrite and the methemoglobin
curves. That nitrites rise and fall sooner than the corresponding methemo-
globin curves is known in this laboratory from many previous experiments,
as well as the fact that the methemoglobin levels are directly propor-
tional to the corresponding nitrite levels. From the data here it
seems quite reasonable to postulate that the primary effect of ascorbic
acid is to lower the blood nitrite levels in the earliest, maximal
portion of the curve, the methemoglobin levels being consequent to this
action. In the latter portion of the curve, when the nitrite levels are
very similar for the control and the experimental groups, and the ascor-
bic acid levels are still high in the experimental animals, there is a
corresponding similarity between control and experimental methemoglobin
levels.
90
-------
As to why the nitrite peak is lower with ascorbic acid, one is
left with mere speculation. Three possibilities come to mind:
1) Nitrite and ascorbic acid combine directly in some sort of complex,
effectively removing the nitrite from either the absorptive peritoneal
surface or the blood stream. 2) Ascorbic acid prevents nitrite absorp-
tion in some other manner, possibly by peritoneal irritation. 3) Ascor-
bic acid, in some fashion, hastens nitrite detoxification and withdrawal
through enzymatic or non-enzymatic means.
Possible public health implications depend upon several considera-
tions. First of all, simply looking at the shape of the methemoglobin
curves, without regard to mechanisms, several facts are apparent.
1) There is partial protection against methemoglobinemia in the earliest
portion of the curve. There is about 1/3 less methemoglobin measured
as percent total hemoglobin in the experimental as compared with the
controls. 2) This protection is not complete even with ascorbic acid
blood levels quite high of the order of 5-6 mg%. 3) After one hour
the protective effect seems to be lost, and if anything, the methemoglo-
bin values in the experimental are slightly, though not significantly,
higher. The question naturally arises if in chronic exposure to high
nitrites or nitrates which convert to nitrites the methemoglobin levels
behave as if they were on the first or second portion of the curves.
Preliminary experiments in this laboratory indicate that high vitamin
C intake does lower the methemoglobin blood level in mice exposed to
high nitrite doses. However> a great deal of work remains to be done
in this area. If ascorbic acid, with chronic administration of high
nitrites, or nitrates, causes the methemoglobin levels to behave as if
they were on the second portion of the curve, it is apparent that admin-
istration of high doses of ascorbic acid will not completely prevent
methemoglobinemia, but may nevertheless provide a certain degree of
easily obtainable protection.
91
-------
REFERENCES
1. Faivre, M., Armond, J., and Faivre, J. Les methemoglobinemies
toxiques. Masson C., ed. Paris, 1970. p. 102.
2. Fitch, J., Waliker, G., and Baxter, E. Methemoglobinemia - a
cause of cyanosis in infants and children. Arch. Fed. 66:143-156,
1949.
3. Gruener, N., and Shuval, H.I. Health aspects of nitrites in
drinking water. In: Developments in Water Quality Research.
Proc. Jerusalem International Conf. on Water Quality and
Pollution Research. Jerusalem, 1969.
4. Pihlaja, T., Valmaki, I., and Yrjana, T. Effects of peroral
ascorbic acid on blood methemoglobin of newborn infants. Biol.
Neonat. 13:62-67, 1968.
5. Kociba, R.I., and Sleight, S.D. Nitrite toxicosis in ascorbic
acid deficient guinea pig. Tosic. and Appl. Pharm. 16:424-429,
1970.
6. Gibson, Q.H. The reduction of methemoglobin by ascorbic acid.
Biochem. J. 37:615, 1943.
92
-------
SECTION XVI
CHRONIC EXPOSURE OF RATS TO SODIUM NITRITE IN DRINKING WATER
Previous chronic studies by others have not shown significant MetHb
levels or other pathological changes in rats consuming water containing
nitrites in the range of 500-2,000 mg/l(l,2).
We have shown that rats being nocturnal animals consume 80% of their
water at night, and because of the rather rapid rate of the MetHb reduc-
tion, by 10:00 A.M. they may no longer show significant MetHb levels even
when consuming water high in nitrites. Our measurements showed peaks in
water consumption and MetHb level in the middle of the night. In design-
ing our chronic studies, we adjusted the "day" and "night" hours in the
animals' rooms so that blood examinations would be taken at about the
rats' "midnight," in order to detect the expected maximum daily concentra-
tion of MetHb.
Based on the information gained from our first chronic toxicity stud-
ies, in which 40 male rats divided into 5 groups, were exposed to nitrite,
for two years, ranging from 100-3,000 mg/1, we have designed a larger
comprehensive experiment to follow chronic toxicity of rats exposed to
sodium nitrite as well as sodium nitrate. The total number of rats in-
cluded in this study was 312. Four groups (II, III, IV, VI) exposed to
200; 1,000; 2,000; 3,000 mg/1 of NaN02 and one group (VI) exposed to
2,000 mg/1 of NaNO_ as well as a control group (I) of the same size were
included in this study. In each group, there are 52 male albino Hebrew
University Sabra rats. The following parameters were checked at regular
intervals:
1. Water consumption; 2. Body weight; 3. Methemoglobin(MetHb);
4. Hemoglobin (Hb); 5. Hematocrite; 6. MR; 7. Serum lactic
dehydrogenase; 8. Serum glutamic acid; 9. Blood glucose;
10. Nitrite in blood and 11. Diphosphoglyceric acid in red blood
cells.
Histological examinations were carried out on heart and lung taken
from rats of the six groups sacrificed at 3-month intervals.
A special animal room for chronic studies was prepared with controlled
light, temperature and ventilation. After weaning, animals were held in
the new conditions for one month for a period of adaptation prior to being
exposed to the various experimental conditions.
93
-------
ANALYTICAL METHODS
1. Methemoglobin determination:
This was done according to our method(1).
2. Hemoglobin was determined by the cyano-methemoglobin method(4).
3. Hematocrite was determined by the micromethod using a "Hawksley"
centrifuge.
4. Methemoglobin reductase was assayed as described in this report.
5. Glutamic acid levels in serum were determined according to
Bernt § Bergmeyer(S).
6. Blood glucose was determined with the toluidine method(6).
7. Nitrites in blood were tested according to our method(7).
8. 2,3 Diphosphoglyceric acid (DPG) was assayed enzymetically as
follows:
PRINCIPLE:
Phosphoglyceromutase
a) 2,3 Diphospho- > 3 phosphoglycerate
* 2-phosphoglycolic acid
Phosphoglycerate Kinase
b) ATP +3 phospho- ^ —> 1,3-diphosphoglycerate
glycerate + ADP+ H+
^ i i j- v t. Glyceraldehyde -P dehydrogenase
c) 1,3 dxphospho- -^Glyceraldehyde -3-P
glycerate + NADH ^ ^ + >
The phosphatase activity of muscle phosphoglycerate-mutase is en-
hanced by phosphoglycolic acid. Phosphoglycerate is formed from DPG
and is phosphorylated to 1,3 diphosphoglycerate (reaction b) which is
reduced to the aldehyde. Reduction of NADH absorbance is followed at
340 my. Standards of DPG were tested in each run.
Reagents
Buffer solution; contains 1.367 gm Imidazol, 1.017 gm MgCU,
0.651 gm Hydrazine sulfate, and 0.472 mg 2-Phosphoglycolate per
liter distilled H^O. Solution is adjusted to pH of 7.4, May be
refrigerated and Kept stored.
94
-------
2. Curvette mixture. Must be made daily. Does not store. Contains
for each 50 ml buffer solution, 30 mg Adenosine triphosphate (ATP)
38 mg reduced glutathione (GSH), 6 mg Diphosphopyridine nucleo-
tide reduced form (DPNH), 0.1 ml Glyceraldehyde-3-phosphate
dehydrogenase (Ga-3-PD) and 0.02 ml 3 Phosphoglyceratkinase (PGK).
3. Phosphoglycerate - Mutase (PGM) .
Procedure
1. Add 0.1 ml whole blood to 0.3 ml distilled water, mix well, and
take a sample for Hb determination.
2. Put the tube in boiled water for five minutes and centrifuge for
2 minutes at 15,000 RPM.
3. To spectrophotometer curvette, add 0.1 ml supernatant and 0.9
ml curvette mixture. Substitute 0.1 ml distilled water for
blank.
4. Read at 340 nm and do not start reaction until each curvette
has reached stability. Note this value as time 0.
5. To each curvette, add 0.005 ml PGM. Shake well.
6. Read until DPNH falls at the same rate as with water blank.
RESULTS
The average 24 hours liquid intake is presented in Table XVI-1. There
was essentially no change in the liquid intake expressed as mg/kg/day
throughout the experiment.
Table XVI-2 shows the change in body weight during the experiment. The
daily range of nitrites or nitrates which the experimental groups received
can be seen in Table XVI-3.
It can be seen from Table XVI-2 that only the rats that got the highest
nitrite quantities suffered from a slight retardation in growth and devel-
opment. They also showed some rejection of the drinking solution.
It is worth noting here that the highest nitrite groups received a daily
amount which is about twice the single dose LDrn of sodium nitrite for
adult rats (around 150 mg/kg). This exposure nevertheless did not cause
far-reaching physiological deviations. It appears that some form of nitrite
detoxification and compensatory mechanisms are strong enough to cope with
such levels of exposure when given in many small doses throughout the day.
Our other studies indicate that suckling rats are more sensitive.
95
-------
Table XVI-1. LIQUID INTAKE OF RATS DRINKING SODIUM NITRITE AND NITRATE SOLUTIONS
ml/kg B.W./24 hrs.
Group
II
vO
III
IV
VI
Time of
Exposure
Control
200 mg/1
NaN02
1000 mg/1
NaN02
2000 mg/1
NaN02
3000 mg/1
NaN02
2000 mg/1
NaN02
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
3
wee,ks
(13)
38.2
3.6
(13)
39.7
316
(13)
38.9
5.4
(13)
35.1
5.7
(13)
31.7
4.9
-
6
weeks
(13)
43.5
2.8
(13)
• 45.9
4.7
(13)
44.7
3.7
(13)
41.3
±3.5
(13)
35.2
±3.4
(13)
47.7
±2.4
12
weeks
(12)
40.6
±2.6
(11)
41.4
±3.2
"(12)
40.0
±1.6
(13)
35.2
±2.1
(13)
31.9
±4.6
(13)
43.7
±3.5
4
months
(12)
40.5
±3.5
(11)
43.0
±4.7
(11)
42.7
±4.9
(12)
38.2
±3.2
(12)
34.1
±4.6
(13)
44.0
±3.0
10
months
(9)
38.1
±4.7
0)
42.2
±5.6
(9)
37.9
±3.7
(9)
36.9
±5.8
(8)
32.9
±7.2
(10)
43.6
±5.8
16
months
(8)
37.1
±6.6
(8)
37.5
±4.0
(8)
34.4
±3.9
(8)
34.9
±5.6
(8)
30.0
±2.6
(8)
40.9
±5.9
-------
Table XVI-2 shows the change in body weight during the experiment.
The daily range of nitrites or nitrates which the experimental groups
received can be seen in Table XVI-3.
It can be seen from Table XVI-2 that only the rats that got the
highest nitrite quantities suffered from a slight retardation in growth
and development. They also showed some rejection of the drinking solu-
tion.
It is worth noting here that the highest nitrite groups received
a daily amount which is about twice the single dose LD5Q of sodium
nitrite for adult rats (around 150 mg/kg). This exposure nevertheless
did not cause far-reaching physiological deviations. It appears that
some form of nitrite detoxification and compensatory mechanisms are
strong enough to cope with such levels of exposure when given in many
small doses throughout the day. Our other studies indicate that suck-
ling rats are more sensitive.
Table XVI-4 represents the mean MetHb levels. Groups III, IV and
V were significantly raised throughout the study. Group III showed
what might be considered subclinical levels, while Groups IV and V
showed MetHb levels which would be considered in humans as clinically
significant.
Table XVI-5 represents steady-state levels of nitrites in blood.
There is a low level NO in the control blood which does not increase
in the lowest exposure g^roup (200 mg/1 of NaNO.) but rises proportionally
as the levels of exposure increase to the highest exposure group. It
is worth mentioning that a raised level of nitrites in blood was found
in Group VI which received only NaNO,. This group did not however
show raised MetHb.
Table XVI-6 represents the peaks of MetHb as caused by different
single dosages of sodium nitrite. There is a clear dose-response rela-
The kinetics of MetHb change after a single administration of NaN02
resulting in an increase phase of 20 min. to 90 min. which culminates
in a sharp peak and declines in a long recovery phase with complete
recovery after about five hours.
The kinetics of the recovery follow a first order reaction with a
of about 90 min. A characteristic recovery picture can be seen
:igure XVI-1. Methemoglobin reductase which is the main mechanism
of MetHb reduction was measured throughout the experiments (Table XVI-7).
The level of the enzyme activity shows fluctuations with age. when levels
are compared among-groups at each time of the determination there is a
definite decrease in the enzyme activity in the groups exposed to nitrites
after a lag period of one month of exposure and in recovery to to control
levels toward the end of the experiment (16 months).
97
-------
Table XVI-2. CHANGES IN BODY WEIGHT IN RATS DRINKING SODIUM NITRITE AND NITRATE
(grams)
Group
o>
II
III
IV
VI
Zero time
Time of (8 weeks
Exposure of age)
10 weeks
18 weeks
10 months 14 months
16 months
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
(52)
243.9
13.6
(52)
241.0
8.0
(52)
220.2
±2.4
(52)
220.0
±0.0
(38)
220.4
±1.4
-
(52)
308- >9
26.7
(52)
303.6
23.6
(52)
311.8
±26.2
(52)
313.4
±33.3
(36)
314.9
±28.5
-
(52)
411.2
34.4
(52)
442.7
36.9
(52)
447.9
±32.0
(52)
434.8
±44.9
(48)
407.5
±40.0
(52)
450.0
±35.7
(36)
526.6
60.5
(36)
502.9
43.2
(36)
536.3
±43.9
(36)
529.3
±56.8
(31)
485.3
±44.2
(40)
538.6
±51.76
. (35)
561.1
62.7
(36)
535.3
52.3
(34)
536.8
±54.8
(36)
557.9
±69.1
(29)
508.6
±40.9
(38)
573.9
±64.6
(29)
612.8
65.4
(29)
584.1
±64.6
(24)
602.5
±53.6
(26)
604.2
±77.1
(25)
544.8
±45.1
(32)
624.4
±74.8
-------
Table XVI-3. NITRITES AND NITRATE INTAKE IN CHRONICALLY EXPOSED RATS
VO
VO
Group II III IV V VI
Range of intake
throughout 16 months 15-30 75-150 150-300 200-400 150-300
of experiment
-------
Table XVI-4. METHEMOGLOBIN LEVELS OF RATS DRINKING SODIUM NITRITE AND NITRATE SOLUTION
(% MetHb)
Group
o
o
II
III
IV
VI
Time of
Exposure
(months)
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
0
(10)
0.37
0.34
(13)
0.68
±0.43
(14)
0.89
±0.70
(10)
0.61
±0.55
(14)
0.98
±0.40
0.5
(7)
0.54
0.48
(8)
0.79
±0.67
(6)
3.63
±2.57
(8)
0.61
±5.77
(9)
18.67
±7.72
1.0
(15)
0.76
0.66
(13)
0.59
±0.63
(16)
4.44
±3.23
(15)
11.43
±6.6
(16)
18.75
(52)
1.09
±0.70
2.0
(50)
0.66
0.43
(47)
0.78
±0.65
(47)
3.64
±2.89
(48)
9.64
±8.04
(48)
23.9
±11.57
(24)
0.59
±0.35
3.0
(16)
1.30
0.87
(15)
1.17
±0.68
(15)
3.44
±2.39
(16)
11.04
±7.6
(15)
22.91
±14.84
(52)
0.83
±0.28
4.0 •
(45)
0.68
0.68
(46)
0.96
±0.70
(46)
3.72
±2.65
(47)
8.51
±6.64
i
(42)
13.68
±8.63
(4)
0.63
±0.22
6.0
(42)
0.54
0.46
(44)
0.71
±0.45
(36)
1.78
±2.79
(37)
8.66
±5.79
(41)
12.84
±7.76
(22)
0.79
0.43
8.0
(34)
0.27
0.35
(36)
0.73
±0.32
(36)
1.91
±1.53
(35)
7.09
±5.68
(29)
9.70
±9.09
16.0
(26)
0.60
0.40
(26)
0.75
±0.66
(23)
2.00
±1.23
(23)
7.57
±4.97
(24)
19.18
±12.39
(30)
0.64
±0.33
-------
Table XVI-5. NITRITE LEVEL IN RAT BLOOD EXPOSED TO SODIUM
NITRITE AND NITRATE yg/ml
Group
I
II
III
IV
V
VI
n
8
20
28
26
24
19
Mean
0.13 ±
0.10
0.44
0.76
1.14
0.44
S.D.
0.28
0.08
0.42
0.67
0.72
0.15
Table XVI-6- METHEMOGLOBIN IN RATS AFTER A SINGLE DOSE
Dose
(mg/kg)
OF NaN02-I.P-
Time of max.
(min)
Max. MetHb
10
20
40
60
80
20
40
75
90
90
7.1
20.3
36.7
51.4
87.3 (died)
101
-------
Table XVI-7. METHEMOGLOBIN REDUCTASE ACTIVITY IN RATS EXPOSED
TO NaN02 AND NaNOj
(vmoles/min/mg Hb)
Group Time of Exposure (months)
0.5 1.0 2.0 4.0 16.0
II
III
IV
VI
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
n
X
S.D.
(8)
2.5
0.2
(8)
2.9
0.5
(8)
2.3
0.3
(8)
2.3
0.4
(9)
2.2
0.6
_
(15)
2.1
0.4
(16)
2.2
0.4
(15)
2.4
0.4
(15)
2.0
0.5
(16)
7.5
0.6
_
(50)
1.7
0.7
(51)
1.7
0.8
(48)
1.4
0.6
(46)
1.3
0.7
(48)
0.6
0.5
(23)
1.9
0.8
(48)
3.8
0.4
(47)
2.7
0.8
(48)
1.9
0.9
(48)
2.3
0.8
(41)
2.1
0.6
(7)
2.6
0.5
(24)
4.8
1.8
(29)
5.5
4.2
(23)
5.8
2.1
(23)
4.6
1.5
(24)
5.8
2.6
(31)
4.5
1.4
-------
50
30
CO
O
O
O
z
UJ
I
i—
QJ
20
10
0123456
TIME (hrs)
FIGURE XVI-1. METHEM06L0BJN REDUCTION RATE IN RAT.
103
-------
Diphosphoglyceric acid (PPG) levels
Concentrations of DPG in red blood cells are increased in conditions
of hypoxia, anemia, congenital heart disease and chronic lung diseases(9).
It was considered relevant to study the effect of induced methemoglo-
binemia on DPG levels.
Table XVI-8 shows that there is an increase in DPG level in the
higher nitrite groups.
Preliminary acute experiments done to study the kinetics of DPG
changes showed that 2-3 hrs after nitrite administration no change in DPG
could be seen.
Further experiments showed that after 24 hrs changes could be de-
tected and after 2 days DPG stabilized at the higher level and remained so
throughout the exposure time. After NaNO- was removed from the drinking
water the DPG concentration levels dropped but at a slower rate.
Glutamic acid in chronically treated rats
Glutamic acid was checked in blood of th'e experimental rats and
the results -&t& represented in" Table XVI-9.
Blood incubated in^ vitro with sodium nitrite caused a significant
decrease in glutamate metabolism. If this fact has any relevance to the
chronic phenomenon,it is still to be elucidated.
Pathology
Deaths of unknown origin of rats throughout the experiments are shown
in Table XVI-10. There were originally 52 rats in each group.
Histological examinations were done in the pretest on different
tissues but only the heart showed changes which called for further studies.
Up to 12 months of exposure no special differences between the control
and the experimental groups could be shown. At this age some change was
noted which was significantly obvious at 18 months (Table XVI-11).
While in most of the control animals the blood vessels showed some
degree of thickening and often even a marked hypertrophy and narrowing,
in the experimental groups the coronaries were thin and dilated (Figures
XVI-2 and XVI-3), their appearance not what is usually seen in animals of
advanced age. It is noteworthy that Group VI, which received sodium
nitrate, showed a similar prevelance of pathology as Group V?which was
exposed to sodium nitrite.
104
-------
Table XVI-8. DIPHOSPHOGLYCERATE LEVEL IN ERYTHROCYTES
OF RATS EXPOSED TO NaN02 AND NaN03
ymoles/gHb
Group n mean S.D.
I
II
III
IV
V
VI
11
7
8
12
7
7
12.7
11.6
14.7
16.8
17.9
12.6
1.6
3.6
1.8
4.3
4.3
2.4
Table XVI-9. GLUTAMATE LEVELS IN BLOOD OF RATS EXPOSED
CHRONICALLY TO SODIUM NITRITE AND NITRATE
(m ymoles/ml)
I II III IV V VI
n 21 25 21 22 23 t 31
mean 242.2 29.8 307.7 46.1 350.3 75.6 326.0 59.2 378.8 60.6 337.1 66.3
Table XVI-10. DEATHS OF UNKNOWN ORIGIN AMONG RATS
TREATED WITH SODIUM NITRITE AND NITRATE
Group I II III IV V VI
n 2 - 2 1 9 -
105
-------
Table XVI-11. THE EFFECT OF SODIUM NITRITE AND NITRATE ON CORONARY BLOOD
VESSELS OF RATS EXPOSED FOR 18 MONTHS
Sodium nitrite
Group
State of
Coronary
Blood vessel
Normal
Thin
Thick
Total*
I
Control
n
14
4
15
21
%
67
19
71
II
200
mg/1
n %
12 52
11 48
6 26
23
III
1000
mg/1
n %
3 19
15 94
4 25
16
IV
2000
mg/1
n %
4 21
13 63
8 42
19
V
3000
mg/1
n %
3 18
12 71
7 41
17
VI
2000
mg/1
n %
3 13
19 83
5 22
23
*Total number of animals. The same heart may show more than one state in the coronary
vessel.
-------
Figure XVT-2.
HEART, CONTROL, 18
MONTHS OLD RAT.
Figure XVI-3.
HEART, RAT, AFTER 18
MONTHS OF DRINKING
WATER CONTAINING 1000
ppm
107
-------
Animals were checked routinely for Hb, hemotacrit, blood glucose, and
serum LDH. No differences were detected between the various experimental
groups and the controls.
DISCUSSION
In spite of the fact that rather high dosages of sodium nitrite were
given to the rats in Groups III, IV and V, only in the last group could
cltaical symptoms be detected. Death rate in this Group (V) was higher than
in the control but not among the other groups. The daily intake of NaNCL
of Groups IV and V were higher than the single dose LD ' of NaNO but the
animal could handle these quantities without extreme deviations from their
normal physiology. This fact exhibits the efficacy of the elimination and
recovery mechanism toward nitrites. One such compensatory mechanism which
was described in this work is the increase in DPG in groups showing ni-
trite induced methemoglobinemia. This phenomenon of increased levels of
DPG have been described in different types of hypoxia(S). We can hypothe-
size that this phenomenon is a result of methemoglobinemia and the direct
effect of nitrites themselves. Such an hypothesis is strengthened by a
recent note(9) that increased levels of DPG were detected in a brother and
sister who suffered from congenital methemoglobinemia. The triggering
mechanism for this compensatory reaction is unknown as the regulatory
mechanism in other hypoxic states. It was found that oxyhemoglobin has
half of the affinity to DPG as deoxyhemoglobin(lO). It was suggested that
this binding reduces soluble DPG levels and causes an increase in glycoly-
sis and thus higher levels of DPG(ll). Such a mechanism can apply also to
MetHb. Relative affinity of DPG to MetHb and to other Hb species should
be compared. Binding of DPG to MetHb has already been shown(12).
Another possibility of regulating DPG by MetHb is by affecting the
glycolytic pathway rate. Higher consumption of NADH for reduction of
MetHb may stimulate glyceraldehylephosphate dehydrogenase to produce more
1.3 diphosphoglycerate and hence more DPG.
As can be seen from Table XVI-9 there is a correlation between glutamic
acid levels and sodium nitrite intake (and hence, levels of nitrite in
serum, Table XVI-5). Preliminary experiments carried out in_ vitro showed
that the presence of nitrite causes a reduction in the metabolism of
glutamic acid in red blood cells. Whether this observation had relevance
tc> the chronic changes and what are the enzymes affected is still to be
elucidated.
Glutamic acid is known to be a powerful excitator of neurons. There
is a possibility that nitrite will cause accumulation in brain glutamic
acid. If such a possibility will be verified it well may be an explanation
to the central effects of nitrites described in this report.
The pathological indications in coronary blood vessels of rats exposed
to sodium nitrite and nitrate has proved to be one of the few clear-cut
108
-------
effects of chronic exposure to these chemicals. It is particularly note-
worthy that the effect was just as preyelant in the rats exposed to sodium
nitrate as sodium nitrite although the former showed no signs of raised
MetHb throughout the experiment. That these pathological findings were
clearly detected in the group exposed to the lowest level of NaNC^ is of
particular concern since that group received only from 15-30 mg/kg/day.
Such an exposure level is not far from the exposure that could occur in
humans under certain, not too extreme, conditions. The significance of
this type of coronary pathology in human health is not clear. To what
extent these findings may be associated with the reported higher preve-
lance of heart disease in areas supplied by drinking water with high
nitrate concentrations(14) must be fully investigated.
109
-------
REFERENCES
1. Druckery, H., Steinhoff, D., Beuthner, H., Schneider, H., Klasner, P.,
Screening of Nitrite for Chronic Toxicity in Rats. Arzneimittel
Forsch. 13, 320-323, 1963.
2. Musil, J. The Effect of a Chronic Sodium Nitrite Intoxication on
Rats. Acta Biol. Med. German, 16, 388-394, 1966.
3. Hegesh, E., Gruener, N., Cohen, S., Buchkovsky, S., R. and Shuval,
H.I., A Sensitive Micromethod for the Determination of Methemoglobin
in Blood. Clin. Chim. Acta, 30, 679-682, 1970.
4. Drabkin, D.L. and Austin, J.H. Spectrophotometric Studies: Prepara-
tion from Washed Blood Cells; Nitric Oxide Hemoglobin and Sulfhemo-
globin, J. Biol. Chem., 112, 51-65, 1935.
5. Bernt, E. and Bergmeyer, H.U. Determination of L-glutamate in Method
of Enzymatic Analysis (H.U. Bergmeyer, ed.) pp. 384-388, Verlag
Chemic and Academic Press, Weinheim and New York.
6. AyvSrinen, A. and Nikkila, E.A. Specific Determination of Blood
Glucose with 0-taluidine, Clin. Chim. Acta, 7, 140, 1962.
7. Shechter, H., Gruener, N., and Shuval, H.I. A Micromethod for Deter-
mination of Nitrite in Blood. Anal. Chim. Acta, 60, 93-99, 1972.
8. Brewer, G.J. and Eaton, G.W. Erythrocytic Metabolism: Interaction
with Oxygen Transport. Science, 171, 1205-1211, 1971.
9. Kttbler et al. Congenital NADH-Dependent Methemoglobin Reductase
Deficiency in Combination with Glycerol Phosphate Dehydrogenase
Activity in Red Blood Cells. Blood, 40, 439, 1972.
10. Garby, L., Gerber, G.H., and de Verchier, C.H., Binding of 2-3-
Diphosphoglycerate and Adenosine Triphosphate to Human Hemoglobin
A, Eur. J. Biochem., 10, 110, 1969.
11. Lenfant et al. Effect of Altitude on 02 Binding by Hemoglobin and
on Organic Phosphate Levels. J. Clin. Inves. 42, 2652, 1968.
12. MacQuarrie, and Gibson, Q.H. The Effect of Organic Phosphates on
Methemoglobin. Fed. Proc. 30, 572, 1971.
13. Krnejevic, K. and Phillis, J.W. Intophosetic Studies of Neurons in
the Mammalian Cerebral Cortex. J. Physiol. (London) 165, 274,
14. Morton, W.E. Am. J. Pub. Health, 61:1371, 1971.
110
-------
SECTION XVII
THE REDUCTION OF METHEMOGLOBIN IN THE HUMAN ERYTHROCYTE
The main reduction pathway of MetHb in the erythrocyte is regarded as
NADH dependent(1). Several enzymatic preparations have been isolated
using mainly NADH for MetHb reduction(2-4). This report describes the
pattern of enzymatic activity which shows rather unusual features.
The source of MetHb and the enzymatic preparation was human outdated
blood. The washed erythrocytes were hemolyzed in water and the membrane
fraction was centrifuged at high speed. The supernatant fluid was dial-
yzed overnight against phosphate buffer (lOmM; pH 7.0). The dialyzed
solution was placed on a DEAE cellulose column (0-9 x 15 cm) on which the
enzyme was absorbed and the hemoglobin (Hb) eluted with phosphate buffer
(10 mM, pH 7.0) . The Hb solution was checked each time and found free of
MetHb reductase activity. The enzyme was eluted from the column with a
gradient of KCI solution (40-200mM).
MetHb_ reductase assay I. The reduction of MetHb to oxyHb was fol-
io wed~aT~577nm~at~2513C^ A Unicam sp 1800 with a bandwidth of Inm was
used. The incubation mixture contained 50 ymoles of tris buffer (pH 7.4)
or 50 pmoles of citrate buffer (pH 4.7); 30 m vimoles of EDTA; 120 m pmoles
of NADH; 50-250 yg of enzyme preparation. Different concentrations of
MetHb together with potassium ferrocvanide were present in the assay in
the molecular ratio of 4:1 (Fe(CN) ~ :MetHb). The mol wt reading of
MetHb was 66.000. Total volume was 0.6 ml.
MetHb reductase assay II. The reaction mixture was as described in
Assay I. Samples were withdrawn up to 15 min. in 2 1/2 min. intervals
from the incubation mixture. The percentage of MetHb was determined by
our method(5). The percentage values were converted into chemical concen-
trations on the base of total Hb value determined by the cyanomethemoglo-
bin method(6). Enzymatic activity was calculated according to the method
of Lee and Wilson(7). Protein was determined according to the method of
Lowry et al.(8).
Sigmoid curves are obtained by plotting enzyme activity vs. MetHb
concentration at low and high pH (Figures XVII-1 and XVII-2). The Hill
coefficient -n(9) was in the range of 2 to 2.2 in different experiments
at both high and low pH. The S~ 5 was about 50 pM at the acidic pH, and
about 500 viM at the neutral one.*
Hebesh and Avron reported low activity at pH 7.4 compared with
"optimal" pH 5.0(3). This was true only at rather low concentrations
111
-------
300
250
200
c
7 150
•0
6 100
50
z^ v r
J» .^* 0 Qi ... i ,„.„; ,
J&^ 100 500 1000
»-• s
1 . 1 i 1 . 1 i 1 | i i 1 i i I i
i
100
200 300
400 500
MHb (pM
'700 900 1100 1300
Figure XVII-1 EFFECT OF MHb CONCENTRATION ON METHEMOGLOBIN REDUCTASE
ACTIVITY. Hill plot of the results is shown in the in-
sert.
(o-o-o in the presence of buffer citrate pH 4.7)
(•-•-• in the presence of buffer tris pH 7.4)
112
-------
1000 2000 3000 4000
MHb (iiM)
Figure XVII-2.
Effect of the MHb concentration on methemoglobin reductase
activity. Hill plot of the results is shown in the insert.
The reaction mixtures were of the composition described in
Assay II. Buffer tris (pH 7.4) was used.
0.800
0700
0600
.5 0.500
E
O 0.400
o
0.300
0.200
0.100
I
100 200
MHb
300
400
Figure XVII-3.
Changes in enzyme activity after exposure to MHb. Preincubated
mixtures were as described in Assay I, using buffer citrate (pH
4.7). Control assays (not illustrated) preincubated for 3 hrs.
without the substrate show similar activities as those assayed
Immediately.
113
-------
of MetHb. The maximum activity obtained at high MetHb concentration
is higher at pH 7.4 than at 4.7.
When the enzyme was further purified on a sephadex G 200 column
(eluante- 10 mM phosphate buffer pH 7.0) the same peak activities
were obtained at pH 7.4 and 4.7.
If we estimate the intraerythrocyte Hb concentration at about
4-5 mM, the acidic maximum activity occurs when MetHb level is about
2% of the total Hb, while at pH 7.4 the maximum activity is obtained
at about 50% of the total intracellular Hb. This phenomenon is com-
patible with our findings(10) that recovery of MetHb in vivo increases
with MetHb concentration.
An interesting feature exhibited by the curve is the inhibitory
effect of MetHb at high concentrations. This effect appears at both
high and low pH but at different concentrations (Figures XVII-1 and
XVII-2). Inhibition is peculiar to MetHb as the other substrates
(Hb, NAD and NADH) do not affect the enzyme in a concentration range
at physiological levels. Metmyoglobin.with a structure similar to
that of MetHb but a quarter of its mol weight,showed the same charac-
teristic sigmoid curve (n = 3) and the inhibitory effect at high con-
centration. Incubation of hemolysates in the presence of different
concentrations of MetHb can convert the enzyme after the exposure to
its substrate to a type more sensitive to MetHb (Figure XVII-3),
both as activator and inhibitor of the system.
Sigmoid curves are common in enzymes following cooperative kine-
tics (9). Other interpretations that do not involve allosteric
mechanisms are also possible(ll).
However, our preliminary findings support the interpretation that
the kinetics of the enzymatic reaction exhibits a real cooperative
mechanism and is not due to the presence of an inhibitor or of im-
proper substrate concentrations. The main findings are as follows:
1. Different enzymatic preparations, such as hemolysate and
various purified fractions, exhibit sigmoidal curves.
2. MetHb reductase activity in the presence of different sub-
strates (MetHb or metmyglobin), from different sources and
after various treatments, follows an allosteric behavior.
3. Dialysis of the enzyme yields an identical kinetic curve as
the control with a cooperative factor (n) of 2.2 in both
cases.
4. Controlled preincubation of the enzyme with urea in various
concentrations up to 3M did not change the cooperative
factor (n).
114
-------
Figure XVII-4.
The inhibitory effect of several unions on the methemo-
globin reductase activity. The determinations were done
in the presence of buffer citrate pH 4.7.
115
-------
5. After fractionalion of the enzyme with amonium sulfate, the
main activity was present in a wide range varying from 55-90%
of saturation.
6. Controlled pre-incubation of the enzyme with trypsin increased
the cooperative factor.
7. Dilution of the enzyme preparation decreased the cooperative
factor.
In our laboratory, preliminary purification of the enzyme led to
a protein with a mol weight of 66.000 (determined by gel filtration),
double the figure reported recently by Kuma a'nd Inomata(12). These
two preparations may be the monomer and dimer types of the same
enzyme. When the effect of ionic strength on enzyme activity was
tested it was found that concentrations of Na+ and K+ up to 200 mM
had no effect. The effect of anions on enzyme activity was also tested
and it was found that COOH", I~, F~, N03~ had no effect but CN", N3~,
and N02~ inhibit the enzyme either by direct reaction with the enzyme or
by changing the substrate to a complex (NC^-MetHb) which is reduced more
slowly than MetHb itself. (See Figure XVII-4). Further evidence is
necessary on the structure and mechanism of the enzyme. Such information
will advance our knowledge on the MetHb reduction which seems to operate
through a sophisticated mechanism.
116
-------
REFERENCES
1. Scott, E.M. in Hereditary Disorders of Erythrocyte Metabolism
(E. Beutler, ed.) pp. 102-113, Brune and Stratton, New York, 1968.
2. Scott, E.M. and McGraw, J.C., Purification and Properties of Di-
phosphopyridien Nucleotide Diaphorase of Human Erythrocytes,
J. Biol. Chem. 237, 249-252, 1962.
3. Hegesh, E. and Avron, M., Enzymatic Reduction of Ferrihemoglobin
Purification of a Ferrihemglobin Reductase from Human Erythrocytes,
Biochem. Biophys. Acta 146, 397-408, 1967.
4. Sugita, Y., Seuchi, N. and Yoneyama, Y., Purification of Reduced
Pyridine - Neucleotide Dehydrogenase from Human Erythrocytes and
Methemoglobin Reduction by Enzyme, J. Biol. Chem. 246, 6072-6078,
1971.
5. Hegesh, E., Gruener, N., Cohen, S., Bochkovsky, R., and Shuval, H.I.
Sensitive Micromethod for Determination of Methemoglobin in Blood,
Clin. Chem. Acta 30, 679-682, 1970.
6. Drabkin, D.L. and Austin, J.H., Preparations from Washed Blood Cells;
Nitric Oxide Hemoglobin and Sulfhemoglobin, J. Biol. Chem. 112, 51-65,
1935.
7. Lee, H. J. and Wilson, I.E., Enzymic Parameters: Measurement of
Vand Km, Biochem. Biophys. Acta 242, 519-522, 1971.
8. Lowry, O.H., Rosebrought, N.J., Farr, A.L. and Randall, R. J.,
Protein Measurement With the Folin Phenol Reagent, J. Biol. Chem.
193, 265-275, 1951.
9. Atkinson, D.E., Regulation of Enzyme Activity, Ann. Rev. Biochem.,
35, 85-124, 1966.
10. Gruener, N. and Shuval, H.I., in Environmental Quality and Safety
(F. Coulston and F. Korte, eds.), Vol II pp 216-226, Georg Theim
Verlag and Academic Press, Stuttgart and New York.
11. Ferdinand, W., Interpretation of Non-Hyperbolic Rate Curves for
2 - Substrate Enzymes - A possible Mechanism for Phosphofructokinase,
Biochem. J., 98, 278-283, 1966.
12. Kuma, F. and Inomata, The Purification and Molecular Properties of
Reduced Nicotinamide Adenine Dinucleotide - Dependent Methemoglobin
Reductase, J. Biol. Chem., 247, 556-561, 1972.
117
-------
SECTION XVIII
CHANGES IN THE MOTOR ACTIVITY OF MICE GIVEN
SODIUM NITRITE DRINKING SOLUTION
INTRODUCTION
Induced methemoglobinemia in mice treated with sodium nitrite drink-
ing solution causes lowering of oxygen carrying capacity of the blood.
Furthermore, apart from methemoglobin production, sodium nitrite toxicity
in itself interferes with several other bodily physiological mechanisms.
It is hypothesized that low levels of chronic subclinical methemo-
globinemia would cause major activity deviations in mice treated with
sodium nitrite drinking solution.
In this study C^jbl/^j mouse was chosen for its known high activity,
as well as for its homogeneous genetic constitution which would reduce
the individual biological differences in the inbred mice.
The activity measured in the barrier activity box is oriented toward
the environment, i.e., toward different components of the barrier box
which are either arrived at by crossing the barrier, by peeking, or by
grooming. The activity thereby measured is distinguishable from such
undirected activity as is measured in an activity wheel. In the activity
box the activity measured is more free responding than it would be in the
activity wheel. In an effort to detect a correlation between the possible
effects of subclinical methemoglobinemia and motor activity changes the
present behavio ral study was undertaken.
In the present study it was intended to investigate behavioral
changes due to chronic treatment of NaN02 in Cgybl/gj mice using devia-
tions in the motor activity as the main parameter.
MATERIALS AND METHODS
A population comprised of 71 male C^bl/^j mice of 50 +_ 5 days old
was used in this study. Animals were divided into five groups with each
group comprised of 15 mice. The five groups were as follows:
1. Group A - Control group, received only tap water.
2. Group B - Experimental group, received 2000 mg/1 NaN02 in its
drinking water.
118
-------
3. Group C - Experimental group, received 1500 mg/1 NaNCL in
its drinking water.
4. Group D - Experimental group, received 1000 mg/1 NaNCL in
its drinking water.
5. Group E - Experimental group, received 100 mg/1 NaN02 in
its drinking water.
All the animals were kept in the animal room where dark and light
hours as well as room temperature were controlled automatically. Room
temperature varied between 19-22°C and the relative humidity varied
between 45%-60%. Dark hours started at 10 A.M. and ended at 10 P.M.
Water and nitrite drinking solution per 24 hours were measured and no
rejection of sodium nitrite drinking solution was noticed. The mean
daily intake of water and sodium nitrite drinking solution is shown in
Table XVIII-1. All the experimental groups were treated with sodium
nitrite drinking solution for three weeks.
The behavioral tests were carried out between 2-4 P.M., the time
of highest intake of drinking solution. Our studies on chronic exposure
of rats to sodium nitrite in drinking water (this report) showed that
rats, being nocturnal animals, consume 80% of their water at night, and
that by 10 A.M. they may no longer show significant MetHb levels even
when consuming water high in nitrites. The measurements showed a peak
in water consumption and MetHb level in the middle of the night. In
designing this experiment the "day" and "night" hours in the animal room
were adjusted so that blood examinations be taken at about the rats'
"midnight", so as to detect the expected maximum daily MetHb concen-
tration.
A barrier activity box(l) was used for the motor activity measure-
ments. The activity box was divided into four quarters by barriers
(Figure XVIII-1). After introducing a mouse to the first quarter, imme-
diately the following parameters for motor activity measurements were
observed and scored: Grooming, Peeking, and Jumping.
Activity as measures:
1. By completely crossing the barrier (Jumping);
2. Peeking over the barrier but returning to its previous enclo-
sure. (Peeking)
3. Licking face with tongue or brushing with foreliabs. (Grooming)
Each observation lasted for five minutes, and the aforementioned
parameters were scored, after which time tail blood was taken for Hb
and MetHb determinations.
119
-------
Table XVIII-1. RELATIONSHIP BETWEEN WATER INTAKE IN MICE AND EXPOSURE TO NiaN02
MEAN DAILY INTAKE OF NaN02mg/kg
*>
Nitrite Concentration of Water mg/1
ED C B
100 1,000 1,500 2,000
H-»
g mg/1 mg/1 mg/1 mg/1
Nitrite Intake mg/kg
8.8 88.8 133.3 177.7
*Mean body weight = 45 gm
Mean water intake =4.0 ml/day
-------
Barriers
»•••
12.7
4..
¥
o
3
1
•»**•• XJ-»
6
o
3
/I
* /
&£
ft
L
y/.
31^ —
i
<
\
L
c
E
o
O
D
•••",,r
Light bulb
50- w-
•Entrance
door
H—30. cm
Figure Xni.I-1. BARRIER ACTIVITY BOX
121
-------
RESULTS
The mean Hb of the control group as well as that of the experimental
groups showed to be 17.74g% with S.D. of 0.46. The mean MetHb percent in the
control group was 0.5% and the mean activity was 40.57/_ min. In mice in
Group B (2000 mg/1 of NaNO?) the mean MetHb percent was 10.14 with a mean
activity of 27.0/5 min.
The results of the tests are tabulated in Table XVIII-2. Table
XVIII-2 shows comparison of motor activity and the MetHb percent between
the control group and experimental groups.
In Group E, 100 mg/1 NaNCL does not seem to produce any change in
blood MetHb level nor any change in motor activity. In Group D, 1000
mg/1 NaNO- changes produced in blood MetHb level and motor activity were
insignificant. In Group C 1500 mg/1 NaNO~, changes in blood methemoglobin
begin to be noticeable but with no significant changes in the motor activity.
The pool of the obtained data was arranged in an array according to
MetHb in order to show the relationship between MetHb and motor activity
(Table XVIII-3).
As it can be seen in Table XVIII-3, 58% of the population fall within
the range of normal methemoglobin and 15% fall within the border line,
where slight increase in MetHb begins to show diminuation in motor activity.
The remaining 27% showed a marked decrease in activity and an increase
in the MetHb with a level of significance P=0.01. The coefficient of
correlation "r" calculated from the pool of the data is -0.6480. From
the obtained pool of the data the (Figure XVIII-2] coefficient of correla-
tion was calculated for; MetHb and motor activity and in this way a regres-
sion line was obtained which shows an inverse correlation between the two
sets of variables.
DISCUSSION
It is noted from the results that 100 and 1000 mg/1 NaN02 did not
raise significantly blood methemoglobin levels nor did they cnange the
motor activity of the mice. In mice Groups c and B, 1500, and 2000 mg/1
NaN02 produced significant changes both in blood methemoglobin level as
well as effect in lowering of motor activity of mice. This phenomenon
was also observed in rats treated with sodium nitrite drinking solution.
The regression line shows an inverse correlation between the two
sets of variables with P=0.002. However, when an attempt was made to
correlate the motor activity and the calculated mean sodium nitrite
dose mg/Kg B.W.,only a poor correlation between these two variables
was indicated. The most probable explanation of this is that the true
sodium nitrite intake between the animals varied greatly.
122
-------
CO
Table XVIII-2. COMPARISON OF MEAN VARIATIONS IN MOTOR ACTIVITY
AND METHEMOGLOBIN PERCENT BETWEEN THE CONTROL GROUP AND EXPERIMENTAL GROUPS
A
Cont. group
n
Mean
S.D.
Max.
Min.
Act.
E
100 mg/1
14
MetHb
' %
0.59
0.48
2.3
0.1
Act.
5 min.
40.57
4.0
47.0
34.0
MetHb
0.67
0.22
1.2
0.1
= Activity = grooming +
15
Act.
5 min.
44.70
8.18
57.0
33.0
peeking +
D
1000
C
mg/1 1500
mg/1
11 16
MetHb
1.27
0.69
2.8
0.8
j umping
Act . MetHb
5 min. %
40.27 2.71
8.5 1.85
48.0 6.1
29.0 0.2
Act.
5 min.
41.75
6.37
56.0
33.0
B
2000 mg/1
15 a
MetHb Act .
% 5 min.
10.14 27.0
4.47 7.27
18.1 35.0
3.1 16.0
Zn=71
5 min.
-------
p = - 0.6480
68 1O 12
MctHb%
14 16 18 2O
Figure XVIII-2. RELATIONSHIP BETWEEN MOTOR ACTIVITY AND METHEMOGLOBIN LEVEL IN MICE
-------
10
Table XVIII-3. RELATION BETWEEN METHEMOGLOBIN AND MOTOR ACTIVITY
IN MICE CONSUMING NaN02 IN DRINKING WATER
MetHb X Activity n
%
0.0 1.9 43.8 41.0
2.0 3.9 40.0
4.0 7.9 35.7
8.0 13.6 27.5
11.0
10.0
^n m
9.0
r * -0.6480 Z n = 71.0
p = level of significance.
-------
FIG. 3
MEAN DIFFERENCE BETWEEN EXPERIMENTAL
AND CONTROL ANIMALS (CONTROL MINUS
EXPERIMENTAL) IN NITRITE LEVELS AT
EACH TIME POINT STUDIES
15 30 65
TIME (min)
60
* statistical significance p^O-05
shown are means + 1SD
Figure XVIII-3. MEAN DIFFERENCE BETWEEN EXPERIMENTAL AND CONTROL
ANIMALS (CONTROL MINUS EXPERIMENTAL) IN NITRITE
LEVELS AT EACH TIME POINT STUDIES
126
-------
The threshold of lowering of motor-activity in C_^/6 mice starts
at less than 2% MetHb. This is the upper limit of metnembglobin which
is found in normal infants (this report).
There are a number of factors which may as well participate in the
lowering of motor activity of mice which should be taken into considera-
tion. Sodium nitrite can cause the conversion of myoglobin into metmyo-
globin which interferes with muscle activity by lowering of the oxygen
capacity of the muscular tissues. Nitrites are also known to be anti-
cholinergic substance. Nitrites can also act directly on the central
nervous system.
In the study of Petukov & Ivanov(2) the mean MetHb of children aged
12-14 years exposed to high levels of nitrates in drinking water (105
ppm as NO ) is reported to be 5.3% which indicates that low levels of
chronic subclinical methemoglobinemia may eventually lead to bodily
physiological changes. This level is reported to bring about some
changes in paychophysiological reactions of these children.
It is as yet premature to extrapolate the results of these studies
as to the significance of sodium nitrite on human health and behavior.
However, since sodium nitrite is such a widely used food additive and
nitrates are so commonly found in water and certain vegetables, it is
essential that the implications of these environmental exposures be
carefully followed up and studied.
REFERENCES
1. Guttman, R., Lieblich, J., and Naftali, G. Sequential Behaviour
in a Novel Situation in Two Inbred Strains of Mice and Their
Hybrids. Life Sciences, Vol. 8, Part II, pp. 893-899, 1969.
2. Petukov, N.I. and Ivanov, A.V. Investigation of Certain Psycho-
physiological Reactions in Children Suffering from Methemoglobinemia
due to Nitrate in Water. Hyg. San. U.S.S.R. 35, 29-31, 1970.
127
-------
SECTION XIX
THE EFFECT OF NITRITES ON ISOLATION-INDUCED AGGRESSION IN MICE
The interest in the effect of environmental agents on behavior is
on the increase as a growing segment of the population is exposed to
such agents. Nitrates and nitrites are frequently ingested in drinking
water, natural food such as vegetables, and in processed foods when
food additives have been utilized. One of the complications following
excessive nitrate and nitrite intake is methemoglobinemia. This disease
is characterized by decrease in the oxygen transport capacity of the
blood. Nitrite which is the direct toxic agent can oxidize hemoglobin
to methemoglobin which is biologically inactive. Nitrates can be reduced
to nitrites mainly by microbiological activity in the food or in diges-
tive systems. Infants in the first months of life are especially
susceptible to methemoglobinemia from these salts(1).
The question of the health implications at the subclinical level of
exposure to nitrates and nitrites is of considerable importance, because
detection of symptoms at such levels is difficult and thus no preventive
actions are likely to be taken.
We have already shown(2) that C5_B1 mice that have been given sodium
nitrite in their drinking water, showed an increase in methemoglobin
level. The average was 1.2% at a nitrite concentration of 1 g/1 (control
mean level * 0.6%) and 10.1% of methemoglobin at 2 g/1 of nitrites.
Levels of 10% or higher of methemoglobin are considered clinically signi-
ficant. In the same report we have shown that nitrites reduced the motor
activity of the mice but only at the highest concentration (2 g/1).
>
>
In the following section, we will show that nitrites cause a signi-
ficant increase in the aggression of mice.
METHOD
Ten Ct._-Bl female mice were mated with males of the same strain.
Each couple was placed in a separate cage. Five couples were given a
sodium nitrite solution (1 g/1 in tap water) for their drinking water
and the other five couples drank tap water. At birth, the adult males
were removed from the cages and the newborne mice were kept with their
mothers for the next 21 days. The mothers who drank NaNO- during ges-
tation continued to receive the same solution through the nursing period.
The young were weighed twice a week through the nursing period.
At weaning (after 21 days) twelve young males from each group were
randomly chosen and each was placed in a separate cage and isolated
for eight weeks. The nitrite group continued to get the nitrites at
the same level (1 g/1) throughout the isolation period, and for the
first five weeks of the behavioral tests.
128
-------
At the end of the isolation period the mice were tested for isolated-
induced aggression by a modified method of Banerjee(3). Each confrontation
between two mice was allowed to take place for ten minutes in a square
fighting cage, and behavior was rated according to a predetermined aggres-
sion scale with maximum score of 18.
Each animal was exposed to another animal once a week, six times, four
of the six sessions were intra-group confrontations and two were extra group
confrontations. There was no set order for these confrontations because the
order might influence the results of the confrontations.
After five weeks the experimental animals were taken off the nitrite
solution and were given tap water to drink. After a break of two weeks the
tests were resumed and the first four sessions were repeated. During the
latter period all the animals drank tap water.
RESULTS AND DISCUSSION
The results of the confrontation sessions are presented in Table XIX-1
and XIX-2. The cumulative score for each group was averaged per the number
of the sessions. Table XIX-1 shows that there was an increase in the
aggressive behavior of the treated group particularly where this group met
the controls. At these sessions, the controls had a significantly low
score. Returning the experimental group to regular tap water was accom-
panied by a decrease in aggressions to the level of the control group which
did not change during this time. The simultaneous increase in the mean
score of the water group when exposed to the previous nitrite group shows
that the score for the individual in each session is not independent of its
opponent.
Table XIX-1. THE EFFECT OF SODIUM NITRITE ON AGGRESSION IN MICE
Number of Mean
Confronta- Aggression
Group tions Score S.D. p*
1. H20 - intra 44 9.01 3.14
2. H20 - extra 22 5.68 3.55 <0.01
3. NaNO, (lg/1)
-intra 38 10.88 4.68 <0.05
4. NaNO (lg/1)
-extra 22 12.95 3.66 <0.01
*The differences in the mean aggression score between Group 1 and Groups 2,
3, and 4, are analyzed statistically according to the "t" test.
"Intra" means sessions in which mice exposed to others belong to the same
experimental group. "Extra" means sessions in which mice exposed to others
belong to the other experiment group.
129
-------
Table XIX-2. AGGRESSION OF MICE AFTER RETURNING TO TAP WATER
Number of Mean
Confronta- Aggression S.D. p*
Group tions Score
1. HO - intra 32 8.97 5.00
2. H20 - extra 15 8.77 5.00 N.S.
3. NaNO,, (lg/l)-intra 38 9.43 5.31 N.S.
' i. . v
4. NaN02 (lg/1)-extra 15 9.50 5.20 N.S.
*The differences in the mean aggression score between Group 1 and Groups 2,
3, and 4, are analyzed statistically according to the "t" test.
"Intra" means sessions in which mice exposed to others belong to the same
experimental group. "Extra" means sessions in which mice exposed to others
belong to the other experimental group.
Several reports have recently described the different phenomena caused
by nitrites or nitrates on the central nervous system(4,5,6). However,
no suggestion has been given regarding the' site(s) or mode of action of
these ions in the tissue. This study hints that at least the behavioral
effect is. reversible and disappears after a short break. Elucidation of
the mechanism of this aggression phenomenon might help in the evaluation
of the risks of exposure to nitrates and nitrites.
130
-------
REFERENCES
1. Gruener, N., and Shuval, H.I. Health Aspects of Nitrate in Drink-
ing Water, in Water Quality Research, Ann Arbor, Humphrey Science
Publishers, Ann Arbor-London, pp 89-105 1970.
2. Behroozi, K., Guttman, R., Gruener, N. and Suuval, H.I. Changes in
Motor Activity of Mice Given Sodium Nitrite in Drinking Solution,
Israel J. Med. Sci. 8, 1007, 1972.
3. Banerjee, U. An Inquiry into Genesis of Aggression in Mice Induced
by Isolation, Behavior XL, 86-89, 1971.
4. Robinson, S., Behroozi, K., Gruener, N., and Shuval, H.I. The
Effect of Chronic Exposure to Sodium Nitrite on the Electroencepholo-
gram of Rats, Environmental Research 5, 4, 1972.
5.' Petuknov, N.I. and Ivanov, A.V. Investigation of Certain Psycho-
physiological Reactions in Children Suffering from Methemoglobinemia
Due to Nitrate in Water, Hyg. San. 55:29-51, 1970.
6. Henderson, W.R., and Raskin, N.H. "Hot-Dog," Headache: Individual
Susceptibility to Nitrite. Lancet, December, 1162, 1972.
131
-------
SECTION XX
THE EFFECT OF CHRONIC EXPOSURE TO SODIUM NITRITE
ON THE ELECTRO ENCEPHALOGRAM OF RATS
INTRODUCTION
In the course of our analysis of the possible side effects of raised
MetHb levels we considered the influence of reduced oxygen supply on
sensitive organs such as the brain as a potential line of investigation.
As yet, little attention has been paid to the toxicological effects of
nitrites on the central nervous system. Petukov and Ivanov(l) have
recently reported on psychophysiological changes in school children in
regions of high nitrates concentrations in drinking water.
The objectives of the present study were to investigate the long-
term chronic effects of NaN02 administered in drinking water on the CNS ,
and changes which it would cause in the brain electrical activity as indi-
cated by EEC deviations. Two sets of experiments were carried out: the
first by Dr. S. Robinson and Mr. K. Behroozi and the second by Dr. M.
Guttnick and Mr. M. Dalit.
FIRST EXPERIMENT
METHODS AND MATERIALS
Sixteen male albino rats of the Hebrew University's "Sabra" strain
weighing initially 200110 gm each were used. Four monopolar ball-pointed
silver electrodes were sfereotactically placed on the dura mater of the
cortex in the left and right anterior and left and right posterior areas(2)
The electrodes were fixed in place with dental cement and later joined to
a connector, cushioned on the skull. Following the implantation of the
electrodes, the animals were allowed to recuperate for two to three weeks
before the recordings' were made.
Animals were divided into four groups, each group comprised of four
rats. Each rat was placed in a separate cage. All groups received plain
tap water for a period of two weeks during which time recordings from all
of them were taken every three days in order to serve as controls of their
own for later comparisons. The three experimental groups received sodium
nitrite in their drinking water for a period of two months, after which
time the NaN02 was removed from their drinking water and the animals re-
ceived plain tap water for an additional period up to four and a half
months. The four groups were as follows:
Group A - Control Group received only plain tap water
Group B - Experimental Group received: 2000 mg/1 NaN02
Group C - Experimental Group received: 300 mg/1 NaN02
Group D - Experimental Group received: 100 mg/1 NaN02
132
-------
All the recordings were made under identical conditions and time.
The room temperature varied between 19° and 22° C. During the record-
ings the animals were kept in a round metallic cage and permitted to
move about freely while recordings were made and animals' behavior
could be observed. Following connection of the EEC adaptor to the
socket, the animals were allowed a period of 15-20 minutes for adjust-
ment. An Alvar 8 channel electroencephalograph was used in this study.
After each recording blood samples were taken from the tail of the
rats for hemoglobin and methemoglobin determination. The evaluation of
EEC records was made by visual methods and wave group counting.
RESULTS
Different concentrations of NaN02 in the drinking water appear
to reveal characteristic changes in the EEG pattern as well as methemo-
globin formation and behavioral changes.
Group A
Group A showed no changes in the brain electrical activity or in
the behavior during the three and a half months period of follow-up.
There were also no changes in the MetHb percent. The mean MetHb per-
cent was 0.5% and varied between 0.0 and 0.8% which is normal in
rats (See Figure XX-1).
Group B
This group was exposed to 2000 mg/1 NaN02 drinking solution which
is equivalent to a daily dose of 280 mg/kg body weight of the rat per
twenty-four hours. What appear to be changes in the brain electrical
activity were noted from the fourth day of treatment. There appeared
diffused spikes and sharp waves and the frequency of the background
electrical activity gradually increased from x.f.8.4 c/s to x.f.11.3
c/s* (See Table XX-1), achieving maximum of frequency after two weeks
of treatment. After two weeks, general paroxysmal outbursts of high
amplitude, up to 240 nv sharp waves became apparent mixed with diffused
theta waves (See Figures XX-2 and XX-3).
During the two months of chronic exposure to NaN02 the EEG records
were of similar pattern. MetHb percent ranges varied between 6.7 and
30.8% with average 12.16%. All the rats in this group were generally
sedate when compared with the control group as well as with their own
behavior before the treatment.
*(5.f. c/s = mean frequency cycles/second)
133
-------
CONTROL 2 WEEKS
15 WEEKS
R.R.L.P.
RA.R.P.
LA-.L.R
w^ ,YV%Y >V%^^
^
Figure XX-2. CHANGE IN BACKGROUND ACTIVITY AFTER TWO AND A HALF WEEKS
OF NaN02 DRINKING SHOWING ACCELERATION OF THE RHYTHM,
APPEARANCE OF SPIKES AND SHARP WAVES MIXED WITH DIFFUSED
THETA AND DELTA WAVES. (Group B)
134
-------
Table XX-1. RELATIONSHIP BETWEEN MEAN FREQUENCY OF BACKGROUND
BRAIN ELECTRICAL ACTIVITY, AMPLITUDE RANGES AND
METHEMOGLOBINEMIA IN RATS CONSUMING DIFFERENT CONCENTRATIONS OF NaNO,
i
Group A Group B Group C Group D
Taj> Water
EEG xfc/s 8.4
Prior to Exposure
S.D. 1.48
EEG Amp. Range yv
Prior to exposure 30-60
Y MetHb percent 0 . 5
Prior to Exposure
S.D. 0.24
EEG Xf -c/s
During Exposure
S.D.
EEG Amp. Range yv
During Exposure
)C MetHb percent
During Exposure
S.D.
EEG xf c/s
Return to tap water
S.D.
EEG. Amp. yv
Return to tap water
*
X" MetHb percent 0.49
Return to tap water
S.D.
NaN00 2000 mg/1
8.4
1.48
30^60
0.5
0.24
11.3
1.11
50-240
12.2
5.89
9.2
0.52
50-240
0.70
0.33
135
NaN00 300 mg/1
8.4
1.48
30-60
0.5
0.24
6.5
0.42
50-200
3.0
0.94
8.0
1.47
50-170
0.47
0.17
NaNO,, 100 mg/1
8.4
1.48
30-60
0.5
0.24
6.1
0.73
30-170
1.1
0.47
7.0
0.91
30-80
0.42
0.10
-------
2^2 WEEKS NO2 20QO.. mg/lit
RA.L.A. V^^/VAV/^AV^W^^
RP-LP. vv--H*A*<"V*-^
RA-RP. .^>^V^^^^A^Mv^^^^^/|V• x^/vyA^
*} j*
LA-LP'. ^ /v'^ ,'.;^y/y/MvV^^/NVr'V\i AV^ JH!*V'
. , % - ^ \r
snj^TlScc. ,
Figure XX-3. APPEARANCE OF GENERAL PAROXYSMAL OUTBURST OF HIGH VOLTAGE,
THETA, DELTA AND SHARP WAVES. (Group B).
RrtWatcr 2 WEEKS
^^-^^
li'ny^V^V^VAiKii^
50 JIv.Tlsee. ,
Figure XX-4. BACKGROUND ACTIVITY STILL RAPID WITH GENERAL PAROXYSMAL
OUTBURSTS OF HIGH VOLTAGE DELTA, THETA AND SHARP WAVES.
(Group B).
136
-------
During the electrical outburst their behavior suddenly changed and the
rats remained motionless. After the end of the outburst their motor activity-
returned to that of their previous pattern. During the electrical outbursts,
there were no clinical convulsions or any collapse and the animals were
thought to be awake. To check this a sharp noise was made during a few such
motionless periods. The animals were alert to the noise but the outburst
continued unabated.
Two of the four rats in the group showed pronounced cyanosis character-
istics of methemoglobinemia with MetHb percent up to 30% and they seemed more
sedate than the rest. In these rats the frequency of the occurrence of
paroxysmal outbursts was greater and also of longer duration.
During the four and a half months of follow-up after withdrawal of NaNC>2
the EEG pattern was similar to that during the NaN02 treatment, i.e., the
outbursts continued to appear and the background activity had a x.f.9.2 c/s
and while the MetHb percent returned to normal and was between 0.3% and 0.7%
(See Figure XX-4).
Group C
This group was exposed to 300 mg/1 NaNC>2 drinking solution from which was
received an equivalent daily dose of 42 mg/kg body weight of the rats per
twenty-four hours. Here also changes appeared after four days of NaN02 treat-
ment and the EEG recordings showed spikes and sharp waves, similar to Group B.
The background activity appeared to be slower than that of their own control.
Here too, after two weeks of treatment, there appeared general paroxysmal
outbursts comprising slow and sharp waves with predominant slow waves of theta
and delta bands. The frequency of the occurrence of these outbursts was less
than in Group B, but the amplitudes were similar fSee Figure XX-5).
During the follow-up of three and a .half months after the withdrawal of
NaN(>2, the EEG background activity had a x.f.8.0 c/s. General outbursts still
continued to appear. (See Figure 6).
The MetHb percent during the treatment ranged between 1.7 and 4.5% and
after the withdrawal of NaN02, it ranged between 0.3 and 0.7%. During the
paroxysmal outbursts the animals in Group C behaved similarly to those in
Group B.
Group D
This group was exposed to 100 mg/1 NaN02 drinking solution from which
was received an equivalent daily dose of 14 mg/kg body weight of the rat per
twenty-four hours. Here too, what appear to be slight changes in the EEG
were noted after four days, and they seemed to be similar to those described
in Group C, except that the frequency of the occurrence of outbursts was
lower.
137
-------
30Omg/fit.
RP-L.P.
R.A-R.R
L.A.L.R
-». ••»v*-*WA-*VXVjf'--*>/\rt*/
-------
During the return to plain tap water for a period of two and a
half months, slow background activity continued x.f. c/s 7.0. The
general outbursts disappeared after two weeks but spikes and sharp
waves were still recorded. The MetHb percent ranged between 0.8 and
2.4% (x MegHb percent 1.1) and after the withdrawal of NaN02 the
MetHb percent ranged between 0.3 and 0.6%.(See Figure XX-8).
In general, their behavior was similar to Group A, but during the
general outbursts they were similar to Group B.
DISCUSSION
The results indicate that NaN02 may have some effect on the EEG of
the rats treated chronically with this substance. In the unfolding of
the recordings obtained, it is noted that increased levels of NaNG*2 in
drinking water appear to lead to:
1) more accentuated changes in the EEG;
2) increase in MetHb percent;
3) decrease in general motor activity as observed visually.
Changes appeared in the background EEG at the three concentration
levels. In rats exposed to 2000 mg/1 NaN02, the background brain elec-
trical activity appeared to be faster than the control group as well as
the other two experimental groups, while at 300 mg/1 and 100 mg/1 NaN02,
the background electrical activity became slightly slower than the con-
trol group. Spikes and sharp waves appeared in the EEG of all the rats
in the experimental groups from the fourth day of the treatment and
continued so during the whole length of the experiment as well as after
their return to plain water. It might be hypothesized that the des-
cribed EEG changes might be due to brain anoxia caused by degenerative
vascular changes in the brain. Huper and Landsburg(3) report on the
brain vascular degenerative changes and vacolation in midbrain and
brainstem with erythroltetranitrates and NaN02 and relate the afore-
mentioned phenomena to stagnant hypoxemia and hyperemia caused by
vasodilatory effects of NaN02-
Another possible cause of hypoxia might be due to nitrite inducing
methemoglobinemia which results in the lowering of oxygen carrying
capacity of the blood. Garbuz(4) in a brief communication has reported
the irreversible acceleration of the EEG rhythm in rabbits acutely
treated with NaN02 and related this phenomenon to anoxic hypoxia which
is caused by methemoglobinemia.
In our present investigation features which might suggest brain
hypoxia expressed by the appearance of slowed background brain electri-
cal activity in EEG were seen mainly at low concentrations of NaNC<2 in
the drinking water of the rats with the mean of 1.1% MetHb in Group D
and 3.0% MetHb in Group C. Such levels are normal or nearly normal for
139
-------
2 WEE KS 1OO mg/|lt NO2
RA_LA.v
RP_UR
J^^r^^^j^
'v^/^'w^
R A RP
L.A.L.P.
V^^VA^^^^^u^vV^V^v^V^M^V^S^^
. Ilatc.
Figure XX-8. SLOW BACKGROUND ACTIVITY STILL CONTINUES WITH DIFFUSED SPIKES
AND SHARP WAVES. (Group D).
140
-------
rats and hypoxia would not be anticipated. It is noted that at the
highest NaNC»2 dose of 2000 mg/1, when high MetHb percent levels were
detected and anoxia might logically be anticipated, the EEC showed
rapid background activity not typical for hypoxic conditions.
Our findings raise the possibility that the disturbed EEC in our
experimental groups after exposure to NaN02 may also be due to toxic
effects of the NaN02 itself on the CNS, even at relatively low levels
of exposure. During the paroxysmal outbursts it is noted that the out-
bursts are diffused, symmetrical, synchronized and uniform. This would
seem to suggest the outbursts are of center-encephalic origin. It is
seen that NaN02 at different concentrations has different characteris-
tic effects on the EEG. This might be due to biphasic effects of NaN02
on the EEG of the rats. Such phenomenon is known from the effects of
other drugs such as pentobarbital .
The observed changes in the brain electrical activity of the rats
appeared to be irreversible in that paroxysmal outbursts in Groups B and
C, and spikes in Group D continue to appear even after the removal of
from the drinking water of the rats.
In these experiments interpretation of the EEG charts were carried
out but the researchers themselves who were aware of the nitrite doses
administered to the rats.
SECOND EXPERIMENT
On the completion of the first small-scale study, it was decided
to continue this line of investigation but due to changes in personnel
the first efforts were devoted to a recheck of the findings of the
first experiment; while the first study was carried out at the Talbieh
Psychiatric Hospital using their facilities and equipment under the
supervision of Dr. S. Robinson, the second series was carried out in
the Environmental Health Laboratory using a Grass electroencephalograph
on loan from the Department of Neurology of the Hebrew University-
Hadassah Hospital. The work here was supervised by Dr. Michael Guttnick,
a neurobiologist from the Department of Zoology. In the second study,
rats were exposed to the following levels of nitrites in drinking water
with four animals per group:
a) 50 mg/1
b) 100 mg/1
C) 300 mg/1
d) 2,000 mg/1
As in the first experiment, Hebrew University Sabra male albino
rats weighing between 180-220^gms had electrodes implanted on the dura
mater of the cortex. AftejKthe implantation of the electrodes, the
animals were given a two-week recovery period. On the completion of the
141
-------
recovery period, control recordings were made on each individual rat in
order to determine whether the implanted electrodes were satisfactory
and whether the rats' EEC recordings were within what might be expected
to be normal, predictable ranges for such animals. Four pre-exposure
recordings were made at intervals of three to four days. Each record-
ing session was of thirty minutes duration. On the completion of the
two-week control period, the animals were exposed to various concentra-
tions of nitrites in their drinking water. The concentration of the
nitrites in the drinking water was checked regularly.
Three or four days after exposure to nitrites in drinking water
was initiated, EEG recordings were made. These recordings were carried
out at regular intervals according to a carefully standardized protocol
during a period of a month and in such a way that each animal had between
six and eight recordings made. After each recording session, a sample
of tail-blood was taken in order to determine the level of MetHb. The
recordings were made on freely-moving animals held in either grounded
metal cages or in a cage made of Perspex. During the complete period
of the recording, the animal was observed for unusual behavior patterns
which were noted on an appropriate point on the EEG charts. Environ-
mental conditions during each recording session were kept as uniform
as possible with constant temperature and light conditions.
RESULTS
The results of the EEG recordings after the exposure to various
doses of nitrites were compared with each animal's own pre-exposure EEG
recording. No obvious differences could be observed in the frequency
or in the appearance of unusual wave-forms or outbursts. The readings
of the EEG recordings were made by three trained observers independently,
two of whom were not informed of the exposure patterns of the animals
whose recordings they were evaluating. No behavioral differences were
noted during the recordings or at other periods, regardless of the
nitrite dose to which the animals were exposed. The methemoglobin level
of the rats were similar to those exposed to the same nitrite dose in
the first experiment.
CONCLUSION
In the attempt to repeat the first experiment, no electrical brain-
wave activity changes were detected between pre-exposure controls and
post-exposure recordings, even with nitrite doses as high as 2,000 mg/1,
as were found in the first study. Since the repeat-experiment also used
only a very limited number of animals, it would be difficult to conclude
that these findings alone totally negate the previous findings, despite
the fact that the second study was carried out under more carefully con-
trolled conditions and the recordings were evaluated independently by
142
-------
trained observers who were not aware of the treatment given to the
animals. At most, we can say that these conflicting findings leave
us in a position where we can say little at this time as to the validity
of the findings of suspected neurological changes resulting from exposure
to nitrites in drinking water. Our own conflicting results and the
findings of Russian researchers(4) in this area nevertheless leave this
question open for further study.
REFERENCES
1. Petuknov, N.E. and Ivanov, A.V. Investigation of Certain Psycho-
physiological Reactions in Children Suffering from Methemoglo-
binemia Due to Nitrates in Water, Hyg. San. 35, 29-31, 1970.
2. Nir, I., Behroozi, K., Assael, M., and Ivriani, I. Changes in
the Electrical Activity of the Brain Following Pincalectomy,
Neuroendocrinology, 4:122-127, 1969.
3. Hueper, W.C. and Landsberg, J.W. Experimental Studies in Cardio-
vascular Pathology, I. Pathologic Changes in the Organs of Rats
Produced by Chronic Nitrite Poisoning, Arch, of Pathology, Vol. 29,
pp. 633-648, 1940.
4. Garbuz, A.M. Functional State of the Central Nervous System in
"Symptomless" Methemoglobinemia, Communication in Brief, Gig. Sanit,
36(1), pp. 101-102, U.S.S.R. 1971.
143
-------
APPENDIX I
SURVEY ON METHEMOGLOBINEMIA
QUESTIONNAIRE
Questionnaire No.:
Identity Card No.:
Date of completing form:
A. DETAILS OF INFANT
1. Surname
2. First Names
3. Sex
4. Hospital where born
5. Date of birth in kg.
6. Home address of parents
7. Name of Clinic
8. Age in months
9. Present weight
B. DETAILS OF PARENTS
10. First Name - Father: Mother:
11. Country of Birth - Father: Mother:
12. Country of Birth of Father's Parents: Father: Mother:
13. Religion
C. GENERAL
14. Was the baby away from the above-mentioned address last week?
Yes No
15. If so, how many days ago did he return home?
16. How many days was he away from home?
Where was he?
144
-------
D. SICKNESS AND TREATMENT
17. Does the baby have diarrhea at present? Yes No
18. If so, for how many days?
19. If not, did he have diarrhea during the last month? Yes No,
20. Was the diarrhea heavy or mild? Heavy Mild
21. Was a clinical diagnosis made? Yes No
If so, what was the diagnosis?
22. Other illnesses over the last month:
23. Did he ever receive iron? Yes No
If so, when?
E. NUTRITION (in the past 24 hours)
24. Mother's Milk Yes No.
25. Milk Powder Yes No.
from when?
26. Pasteurized Milk Yes No.
from when?
27. Sterilized Milk Yes No.
from when?
28. Fresh Cow's Milk Yes No.
from when?
29. Goat's Milk Yes No.
Type of Milk Powder (brand name)
Additional comments on nutrition:
30. Was supplementary water given? Yes No
If so, in what form was the additional water given?
(e.g., water added to milk powder, or sweetened water as
liquid supplement)
(daily portions)
Other Additions to Nutrition
31. Soup • " Yes No.
32. Porridge Yes No.
33. Sausage Yes No.
145
-------
34. Spinach Yes No.
35. Vitamin C Yes No.
36. Tomatoes Yes No.
37. Citrus Yes No,
Clinical symptoms if any ,
146
-------
APPENDIX II
List of Publications
1. Gruener, N. and Shuval, H.I., "Health Aspects of Nitrates in Drink-
ing Water" in Developments in Water Quality Research, Ann Arbor-
Humphrey Science Publishers, Ann Arbor - London, pp. 89-106, 1970.
2. Hegesh, E., Gruener, N., Cohen, S., Buchovsky, R., and Shuval, H.I.
"A Sensitive Micro-Method for Determination of Methemoglobin in
Blood," Clin. Chim. Acta, 30, 579, 1970.
3. Shechter, H., Gruener, N. and Shuval, H.I., "A Micro-Method for
the Determination of Nitrite in Blood," Anal. Chim. Acta 60, 93,
1972.
4. Shuval, H.I. and Gruener, N., "Epidemiological and lexicological
Aspects of Nitrates and Nitrites in the Environment," Am. J. Pub.
Health, 62, 1045, 1972.
5. Behroozi, K., Robinson, S., Gruener, N. and Shuval, H.I., "The
Effect of Chronic Exposure to Sodium Nitrite on the Electroencepha-
logram of Rats," Environ. Research 5,4, 1972.
6. Robinson, S., Behroozi, K., Gruener, N. and Shuval, H.I., "Changes
in the Electrical Activity of the Brain of Rats fed with Sodium
Nitrite," Abstract: Proc. Symposium on Environmental Physiology,
Israel J. Med. Sci., 8:1006, 1972.
7. Gruener, N. and Shuval, H.I., "Studies on the Toxicology of Nitrites,"
in the Environmental Toxicology and Safety, Ed. F.G. Coulston and
F. Korte, Academic Press, London-New York, Vol. II, 1972.
8. Gruener, N., Shuval, H.I., Behroozi, K., Cohen, S. and Schechter, H.,
"Methemoglobinemia Induced by Transplacental Passage of Nitrites in Rats"
Abstract: Isr.J. Agr. Res. 1972. Also.Bull. Env. Cont. Toxicol. 9,44,1973.
9. Shuval, H.I. and Gruener, N., "The Association Between Nitrates in
Drinking Water and Infant Methemoglobin in Several Communities in
Israel," Abstract: Isr. J. Agr. Res., 1972.
10. Behroozi, K., Guttman, R., Gruener, N. and Shuval, H.I., "Changes
in the Motor Activity of Mice Given Sodium Nitrite in Drinking
Solution," Abstract: Proc. Symposium on Environmental Physiology,
Beersheba, Israel J. Med. Sci. 8:10007, 1972.
11. Gruener, N. and Shuval, H.I., "Health Aspects of the Exposure to
Nitrates in the Environment," Harefuah, July 1973.
147
-------
12. Gruener, N., Yarom, R. and Shuval, H.I., "Histological Changes in
Rat Heart Arteries After Chronic Administration of Nitrates and
Nitrites," Abstract: Proc. European Symp. on Impact of Ecological
Factors on Peripheral Vascular Disease, September, 1973.
13. Shuval, H.I. and Gruener, N., "The Effects on Man and Animals of
Ingesting Nitrates and Nitrites in Water and Food" in "Effects of
Agricultural Isotope Studies," Proc. Joint FAO/Joint FAO/IEAE Panel
Experts, Vienna, 117-130, 1974.
14. Gruener, N. and Cohen, S., "The Reduction of Methemoglobin in the
Human Erythrocyte," Physiol. Chem. 5 Physics 5:575, 1973.
15. Gruener; N., "The Effect of Nitrites on Isolation-Induced Agression
in Mice," Pnarmacol. Biochem. Behavior 2:267- 1974.
16. Gruener, N., Scheuermann, D. and Shuval, H.I., "Survey of Liquid
Intake in Infants," Environ. Physiol. Biochem. 4, 1974.
17. Gruener, N. and Shuval, H.I., "Evaluation of the Health Effects of
Nitrates in Water" in Recent Advances in the Assessment of the
Health Effects of Environmental Pollution, 1974. Commission on
European Communities, Luxembourg, 2:1067-1072, 1975.
18. Gruener, N. and Toepleiz, R., "The Effect of Changes in Nitrate
Concentration in Drinking Water on Methemoglobin Levels in Infants,"
Int. J. Environ. Studies. 7(3) 1975,(RECD 1976) 161-163.
19. Gruener, N., "Ontogenetic Development of NADH Dependent Methemoglobin
Reductase in Erythrocytes of Different Species," Comparative Biochem.
§ Physiol. J. Toxicol. Environ. Health., 1(5):787-791, 1976.
20. Shechter, H. & Gruener, N., "An Evaluation of the Ion Selective
Electrode for Determination of Nitrate in Highly Mineralized Drinking
Water," J. Am. Water Works. Assn. 68:543, 1976.
148
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/1-77-030
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Health Effects of Nitrates in Water
5. REPORT DATE
June 1977 issuing date
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Hillel I. Shuval
Nachman Gruener
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Environmental Health Laboratory
Hebrew University
Hadassah Medical School
Jerusalem, Israel
10. PROGRAM ELEMENT NO.
1BA614
11. CONTRACT/GRANT NO.
06-012-3
12. SPONSORING AGENCY NAME AND ADDRESS
Health Effects Research Laboratory -. Cin., OH
Office of Research and Development
U. S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. TYPE OF REPORT AND PERIOD COVERED
Final Renort P.T.. 480
=ni
INC
14. SPONSORING AGENCY CODE
EPA/600/10
Prnj
15. SUPPLEMENTARY NOTES
16. ABSTRACT
multi faceted study of the health effects of nitrate in drinking water
using epidemio logical and toxicological techniques is reported on. Several analyti-
cal procedures were developed to allow the research to be conducted.
The results of the epidemiological studies indicate that infants consuming ap-
preciable amounts of water high in nitrates in the form of powdered milk formula
show significantly raised methemoglobin levels. This is also true for infants con-
suming tap water having a nitrate concentration ranging from 45-55 milligrams per
liter .
Laboratory studies of acute and chronic exposure used nitrites and found in-
dication that nitrites can pass the rats placenta and cause raised methemoglobin
levels in the fetus; that pregnant rats are particularly sensitive to exposure to
nitrites, that pups born to dams exposed to nitrites during gestation show poor
growth and development; and that rats exposed to sodium nitrate as well as sodium
nitrite in their drinking water for 18 months show distinct deviations in heart
blood vessels even at the level of 200 mg/liter of NaNO^. Behavioral effects were
noted in mice exposed to high concentrations of nitrite in drinking water.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
cos AT I Field/Group
Nitrites
Potable water
Hemoglobins
Water supply
Drinking Water Standards
Health Effects
13B
18. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
161
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
149
* U.S. GOVERNMENT PRINTING OfHCE: 1977-757-0 56 /64ZO
------- |