vvEPA
United States
Environmental Protection
Agency
Health Effects Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S1-81-027 Apr. 1981
Project Summary
Health Effects of Swimming in
Lake Pontchartrain at
New Orleans
Virginia K. Ktsanes, Ann C. Anderson, and John E. Diem
This two-year study measured
health effects of swimming in marine
recreational waters in a subtropical
climate by testing the association be-
tween reported post-swim symptom
rates and density of bacterial indicator
organisms. The study replicated
previous investigations conducted at
beaches in temperate climates. Vari-
ation in water quality at the same site
provided bacterial exposure levels for
swimmers. Data were obtained from
approximately 5400 swimmers and
2300 nonswimmer controls over the
two years.
Indicators studied the first year were
E. co//, Klebsiella, Pseudomonas
aeruginosa, fecal coliforms, and
enterococci. The first four organisms
followed relatively the same concen-
tration pattern while there was an
inverse relationship between E. coli
and enterococci after rainfall. For
children under age 10 a significant
positive association was found be-
tween gastrointestinal symptom rates
and exposure to E. coli levels above
200/100 ml. No association was
noted for enterococci. For older per-
sons, age 10 and over, evidence of a
relationship between symptoms and
enterococcus density was strongly
suggested.
The second year study focused on
the relationship of E. coli and entero-
coccus levels with swimmer illness.
Bacteria counts were considerably
lower, especially E. coli levels. First
year results were partially substanti-
ated in that illness rates were again
higher for swimmers than nonswim-
mers and highest for young
swimmers. No association, however,
was found between either microbial
indicator and symptom rates for
children. For the older group there
appeared to be an interaction effect of
E. coli and enterococcus levels with
symptoms rates. Study of a small
sample of swimmers at a less polluted
site on the same lake (i.e., very low E.
coli and equivalent enterococcus
densities) corroborated the second
year findings by age.
In both years a relationship between
enterococcus density and gastroin-
testinal symptom rates in swimmers
over age nine established the impor-
tance of this indicator as a criterion of
swimming water quality. IMonconsist-
ent findings for children with respect
to E. coli do not rule out the impor-
tance of this indicator but suggest
perhaps a criterion level that could not
be adequately tested in the second
year.
This Project Summary was develop-
ed by EPA's Health Effects Research
Laboratory, Cincinnati. OH, to an-
nounce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
As part of the' national program to
develop health effects criteria for
-------
marine recreational waters, the U.S.
Environmental Protection Agency (EPA)
in 1973 through 1975 conducted pro-
spective epidemiological-microbiolog-
ical studies at bathing beaches at Coney
Island and The Rockaways, New York.
The overall purpose of the New York
beach studies was, in the first year, to
test methodology designed to overcome
limitations of past research to measure
health effects of swimming in recrea-
tional waters near cities, and in the
second year, to replicate the study for
corroboration of findings from the first
year's satisfactory methodological
experience. The combined results of the
two studies along with those of thethird
year were for the purpose of developing
a data base relating illness as measured
by symptomatology to some potential
microbial indicators of water quality. In
New York the health effects of swim-
ming at a "barely acceptable" (BA) site
were compared to those from a control
"relatively unpolluted" (RU) site. At
each site symptom rates for nonswim-
mers were used as the baseline
measure against which to compare
health effects attributable to water
exposure.
In general, the results showed that
the rates for broad categories of
symptoms (gastrointestinal, respiratory,
other, and disabling) were higher for
swimmers than nonswimmers at both
the BA and RU beaches The differen-
tials in rates (swimmers to nonswim-
mers) were higher at the BA site. The
only statistically significant difference,
however, was for combined gastroin-
testinal symptoms at the BA beach. This
difference appeared to be attributable
largely to higher reported rates of
Hispanic-American and young (under
age 20) swimmers at Coney Island (the
BA site) than their counterparts at
Rockaways (the RU site).
In attempts to demonstrate a relation-
ship between indicator density and
symptomatology, Escherichia coli and
fecal streptococci appeared to be the
most promising among several indica-
tors that were examined.
A third phase of the national program
was to replicate the study procedure at
some subtropical site for the purpose of
verifying or expanding the relationship
between microbial indicators and
reported symtomatology obtained in the
New York study. Further, it was hoped
that a mathematical model could be
developed and tested on data obtained
from these geographically distinct but
representative locations.
Following the pretesting of conditions
at beaches near two southern cities in
1976, a popular site on Lake Pontchar-
tram at New Orleans was selected for
study in Summer 1977. Pretests
indicated that replication of the epide-
miolog.'cal procedures of the New York
study was feasible in terms of size of the
beach-going population, itssociodemo-
graphic mix, and response rate. New
Orleans did not provide, however, two
distinctive bathing sites with respect to
pollution level. The lake at this site was
noted for almost day-to-day variation in
fecal streptococci and fecal coliform
densities, having been tested for a
number of years by the City Health
Department. The variation appears to be
associated primarily with rainfall, an
association that is so consistent that the
Health Department abandoned an
earlier practice of posting the beach
"closed" on days when counts were
high in favor of a permanently posted
warning that "the water. . . .is subject
to intermittent pollution, especially after
heavy rainfall."
The association between fecal coli-
form count and rainfall is attributed to
the city's low-lying position, the high
water table, and soil-settling charac-
teristics. In Orleans Parish (county) the
sewerage system is below river level
and ordinarily pumped into the
Mississippi River south and east of the
city. Because of differential settling, the
pipes become cracked and broken.
During heavy rains as the ground
becomes saturated, there is infiltration
from sewage to drainage lines To pre-
vent flooding the excess water must be
pumped out through canals. One such
canal is the Orleans Outfall Canal
through which water is pumped out into
Lake Pontchartram at a point approxi-
mately 1200 yards from Levee Beach,
the site chosen for study From contigu-
ous Jefferson Parish, however, sewage
is discharged into the Lake at several
points, the nearest being about 2 5
miles from the beach The flow from that
point is in the general direction of the
beach
Given the noted variation in microbial
levels, plus climatological data for the
past few years showing periods of dry
and wet weather during each summer,
it appeared reasonable to assume that
in any given year the one site would
provide both "experimental" and
"control" conditions for measuring
health effects associated with swim-
ming in subtropical recreational waters
under different levels of pollution
For the New Orleans study, the
number of indicators was limited to five'
E. coli, Klebsiella, Pseudomonas aeru-
ginosa, fecal coliforms, and enterococci
Testing conditions the first year
proved satisfactory. Promising tech-
niques for measuring individual
exposure to water conditions defined by
indicator density were developed, and
results with respect to swimmer-
nonswimmer differential symptom
rates and relationships between
symptoms and indicator levels tended to
corroborate the findings of the New York
study.
The study was repeated the following
year, summer 1978, using the same
procedural design, at the same site, but
limiting the indicators to E. coli,
Klebsiella, and enterococci A major
unanticipated change in water condi-
tions was observed the second year,
however. There was considerably less
variation in and lower levels of indicator
densities, and the relationship of the
indicators differed from one year to the
next, particularly E. coli and enterococci.
This effect was apparently attibutable to
"natural" causes since no changes had
occurred in sewage disposal or storm
water runoff provisions for the city. The
month prior to study, however, ten \
inches of ram had fallen on the city in
the matter of a few hours which may
have had a cleansing effect.
In the second year, also, the attempt
was made to study a second site on the
opposite side of the lake known to be
less polluted
Conclusions and
Recommendations
The results of this study measuring
health effects of swimming in relation to
specific indicator organisms present in
the subtropical water of Lake Pontchar-
train are as follows:
Rates of swimmers were generally
higher than rates for nonswimmers for
a variety of symptoms, including gastro-
intestinal, respiratory, eye-ear-nose,
and other For gastrointestinal
symptoms the swimmer rates were con-
sistently significantly higher. Diarrhea
was the most frequently reported
specific symptom among swimmers.
Children under age 10 who swam
showed the highest gastrointestinal
symptom rate (12 percent) and the
greatest swimmer-nonswimmer differ-
ence (six percent). The symptom rate for
-------
swimmers age 10 and over was approxi-
mately eight percent and the swimmer-
nonswimmer difference three percent
In the second year, statistically signif-
icant rate differences were found
between swimmers and nonswimmers
in the older but not the younger age
group for symptoms other than gastro-
intestinal. The rates for older and
younger swimmers were essentially the
same, but the rates for young non-
swimmers were higher than the rates
for older nonswimmers, resulting in a
considerably smaller swimmer-non-
swimmer difference for the young. The
disparity in these findings by age
suggests that the etiological agents for
these symptoms are not necessarily
water-related. The consistent signifi-
cant findings both years for gastroin-
testinal symptoms regardless of age,
however, is evidence for the argument
of the presence of microbial agents in
Lake Pontchartrain that are associated
with these types of complaints.
Investigation of the relationship
between microbial indicators and
symptomatology strongly suggests dif-
ferent importance of two indicators £".
coli and enterococci, by age. A
significant association between E. coli
density above 200/ml. and gastroin-
testinal symptom rates for children was
found the first year This, however, was
not substantiated the second year,
perhaps because E. coli densities only
slightly above this level were found on
only three of the 12 trial days The
absence of an enterococcus effect
either year for children is viewed as
important for further substantiation A
positive relationship between entero-
coccus density and gastrointestinal
symptoms for the older age group
appeared in the findings both years,
either independently or in interaction
with E. coli.
Lake Pontchartrain is probably more
unique than typical of a marine recrea-
tional bathing site in the subtropical
region, hence generalizations of the
findings to other such sites may not be
appropriate. However, analysis of the
relationships between indicators and
symptomatology is not necessarily
specific to a situation. The day-to-day
and often hour-to-hour variation in
water quality at this site presented a
challenge in measurement of exposure.
Although a number of methods were
attempted, the principal technique, in
terms of indicator density by day, time,
and place of swimming for each individ-
ual, has not been used in other studies.
This lack of uniformity in exposure
measurement hampers precise
comparison of results from studies done
elsewhere, especially with respect to
suggested thresholds of exposure. Con-
clusions from this and other studies are
consistent with respect to the
importance off. co//and enterococci as
agent indicators and the greater vulner-
ability of children. The differential
importance of each indicator by age
found in this study warrants further
investigation.
Virginia K. Ktsanes, Ann C. Anderson, and John E. Diem are with Tulane University
School of Public Health and Tropical Medicine, New Orleans. LA 70112.
Victor J. Cabelli is the EPA Project Officer (see below).
The complete report, entitled "Health Effects of Swimming in Lake Pontchartrain at
New Orleans," (Order No. PB 81-178 741; Cost: $6.50, subject to change} will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, OH 45268
i US GOVERNMENT PRINTING OFFICE 1981 757-012/7058
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
ir s o o u o 3 ? y
(I 5 L'
IT,
------- |