Review of Scientific Methods and Findings for: Final Report of The Investigation of Unusual Illnesses Allegedly Produced by Endosulfan Exposure in Padre Village of Kasargod District (N. Kerala) Robert P. DeMott 1 , Thomas D. Gauthier 1 , Diane J. Mundt 2 1 - ENVIRON International Corporation, Tampa, Florida, USA 2 - ENVIRON International Corporation, Boston, Massachusetts, USA Corresponding Author Robert P. DeMott, Ph.D., Principal Toxicologist ENVIRON International 10150 Highland Manor Dr., Suite 440 Tampa, FL 33610 USA [email protected]Submitted -- March 27, 2012
28
Embed
Review of NIOH Study of Endosulfan in Kerala, India
Review of Scientific Methods and Findings for: Final Report of The Investigation of Unusual Illnesses Allegedly Produced by Endosulfan Exposure in Padre Village of Kasargod District (N. Kerala)
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Review of Scientific Methods and Findings for:Final Report of The Investigation of UnusualIllnesses Allegedly Produced by EndosulfanExposure in Padre Village of Kasargod District (N.Kerala)
Robert P. DeMott1, Thomas D. Gauthier1, Diane J. Mundt2
1 - ENVIRON International Corporation, Tampa, Florida, USA2 - ENVIRON International Corporation, Boston, Massachusetts, USA
Corresponding Author
Robert P. DeMott, Ph.D., Principal ToxicologistENVIRON International10150 Highland Manor Dr., Suite 440Tampa, FL 33610 [email protected]
Submitted -- March 27, 2012
Scientific Review –NIOH Endosulfan Study
1
Executive Summary
This report evaluates the scientific methods, conduct, interpretations and conclusion of a report
titled “Final Report of The Investigation of Unusual Illnesses Allegedly Produced by Endosulfan
Exposure in Padre Village of Kasargod District (N. Kerala)”conducted by the Indian National
Institute of Occupational Health (NIOH). This is a project report of a study regarding
environmental and human health conditions in an area near cashew plantations where there
has been a history of aerial application of the pesticide endosulfan. The report covers
environmental sampling of soil, water, sediment and other materials. It also includes a survey
used to obtain information about human health conditions and follow-up testing involving
physical examinations and blood sampling for endosulfan and hormonal analyses. The report
reaches broad and strong conclusions that human health conditions grouped under three
headings (neurobehavioral disorders, male reproductive system abnormalities, and congenital
malformations in females) are more prevalent in Padre village (the “study”population), near the
sprayed plantations, than in a reference village approximately 25 km distant where this
endosulfan application method was reportedly not used. The report goes so far as to conclude
that endosulfan is the “most probable”cause of the conditions documented in the report.
The evaluation of this report covers the study design, the analytical chemistry methods and the
relevance of the findings to endosulfan effects. First, a necessary limitation of the study design
employed to evaluate human health conditions, termed a survey study, is that it cannot reach
conclusions regarding the causes of effects noted. This method is generally accepted as
suitable solely for preliminary hypothesis generation. Based solely on the study design
selected, ascribing endosulfan as the most probable cause of the conditions recorded is not a
scientifically valid conclusion, even; without considering the quality of its conduct and findings.
When we do consider the analyses conducted and the interpretations, however, serious
limitations and inadequacies emerge. There were serious omissions and errors in the
computation and presentation of the analytical results for endosulfan in environmental samples.
In fact, there was not even confirmation that the measurements made were actually endosulfan.
Lack of confirmation, calibration and quality assurance/quality control information are such that
these results would commonly be rejected by environmental regulatory agencies and would not
be considered reliable for characterizing human health risks for any type of regulatory or judicial
action.
The survey method as conducted to obtain information regarding human health conditions does
not conform to the generally accepted requirements for epidemiological studies and is
insufficient for reaching a conclusion that endosulfan effects are observed in Padre village.
Substantial sources of potential bias were not prevented or considered in the analysis, and then
were not discussed in the report as possible sources of uncertainty and limitations. The
comparisons were frequently simplistic and, in some cases, presented in ways that served to
mask or overlook information that would have weakened the conclusion of adverse effects
occurring more prevalently in Padre village. The epidemiological results obtained and the
Scientific Review –NIOH Endosulfan Study
2
analyses conducted also to not meet the standards commonly required for studies used to
support regulatory or judicial action.
The effects reported to be linked to endosulfan are not biologically plausible as toxicological
responses to the low, background levels of endosulfan found. The congenital conditions
highlighted for male reproductive effects, in fact, are not even found in excess among the boys
from Padre village. And, the measurements characterized to reflect neurobehavioral disorders
are not generally accepted as direct markers of such effects. The interpretations stretch highly
uncertain and highly subjective tests into such serious terms as “disorders,”“abnormalities,”and
“malformations.”The results do not support the presence of elevated rates of such conditions in
Padre village. The analyses and interpretations do not meet generally accepted scientific
standards for establishing chemical exposures as an explanation for purported health effects.
Finally, there were significant and serious discrepancies in some cases between the raw data
obtained through a Right to Information request, the presentations made in the report, and the
subsequent presentations made in a journal article published from the study. Sample results
were excluded, transcription errors were made and numbers of subjects were changed in ways
that served to make the conclusions of the report and the journal article in particular stronger. In
conclusion, the NIOH Report cannot be used to draw a causal connection meeting generally
accepted scientific standards between endosulfan exposure and various reported symptoms
and outcomes because of the limitations of the design, the uncertainties and inadequacies of
the analyses, and the lack of concordance between the reported findings and actual adverse
health conditions.
Scientific Review –NIOH Endosulfan Study
3
1 Introduction
This report provides a critical review of the study prepared by the National Institute of
Occupational Health (NIOH), Indian Council of Medical Research, titled “Final Report of The
Investigation of Unusual Illnesses Allegedly Produced by Endosulfan Exposure in Padre Village
of Kasargod District (N. Kerala)”(hereafter referred to as the “NIOH Report”).
1.1 Background
The NIOH Report was prepared in response to several reports in the press of unusual diseases
in residents of small villages in the Kasargod district of Northern Kerala. The villages are
located below hilltop cashew plantations that have been treated for control of tea mosquitoes by
aerially spraying with endosulfan insecticide two to three times a year for over 20 years (NIOH,
2002; Saiyed, 2003).
At the request of the Indian Council of Medical Research (ICMR), a three-member team from
NIOH visited the area in August 2001 and recommended following up their visit with an
epidemiological study to investigate the prevalence of disease in school children from the
targeted population and a nearby control population. The field study was conducted from
September 24 to October 7, 2001 with the following objectives (NIOH Report –page 5):
To confirm the reported disease pattern in the exposed populations and evaluate the
magnitude of the problem by comparison with control populations through a well
designed epidemiological study.
To search for etiological factors if the exposed populations show abnormal disease
patterns and generate a hypothesis.
To confirm the presence of endosulfan residues in environmental and biological samples
and estimate their levels.
An initial draft (the First Report) was promised by December 2001 and a final version of the
report (the NIOH Report) was published in July 2002. The final NIOH Report included additional
analyses of drinking water and soil samples collected in June, 2002, after the initial report.
1.2 Information Sources
In completing our review we relied in part on information provided in:
The NIOH Report (NIOH, 2002);
An initial draft of the NIOH Report (hereafter referred to as the “First Report”)(NIOH,
undated); and
A companion paper published in Environmental Health Perspectives (Saiyed et al.,
2003) along with comments on the paper (Abraham, 2004; Indulkar, 2004) and the
author’s response (Saiyed, 2004) published in the same journal.
Scientific Review –NIOH Endosulfan Study
4
In addition, a limited amount of information was provided in response to the request by Mr. B.
Mallesham under the RTI Act 2005. The RTI response included 100 pages of material
including:
93 pages of Gas Chromatography-Electron Capture Detection (GC-ECD) output
corresponding to 2 soil samples analyzed on May 30, 2002 (Soil001.CHI and
Soil002.CHI); 2 water samples analyzed on December 13, 2001 (Water002.CHI and
Water003.CHI); 1 blank sample analyzed on November 21, 2002 (Blank001.CHI); and 4
standard samples analyzed on October 29, 2001 (VK013.CHI and VK014.CHI), October
31, 2001 (VK017.CHI) and December 6, 2001 (Std001.CHI).
1 laboratory notebook page (unsigned and undated) describing the extraction and clean-
up procedure for analysis of endosulfan in soils.
6 partially masked laboratory notebook pages (unsigned and undated) with limited
standard and sample information.
Scientific Review –NIOH Endosulfan Study
5
2 Comments on Analytical Data Quality
The NIOH Report presents the results of analytical chemistry testing used to measure
endosulfan concentrations from samples of soil, sediment, water and human blood. The report
uses these results to make comparisons between the environmental conditions in the Padre
village area, hypothesized to be impacted by aerial applications of endosulfan, and a reference
area (Meenja Panchayath village) approximately 25 km away where this type of application was
reportedly not used. The report uses the results from blood samples to make comparisons
between children from the study versus reference areas. These analytical data are, thus, the
critical foundation upon which the comparisons attempting to link endosulfan to the observations
in Padre village depend.
In our review we identified a number of issues in the way the analytical chemistry tests were
conducted and interpreted that affected the quality and reliability of the analytical data. These
issues were of a nature and extent that preclude the analytical results being considered valid
according to the standards required by governmental regulatory agencies. The results are also
unreliable to the extent that they do not meet the generally accepted standards for use in
guiding scientific interpretations of environmental or public health conditions. These issues
include:
Selected data have been excluded from the NIOH Report without explanation.
Information obtained from NIOH under the Right to Information Act reveals
inconsistencies between the raw data compared to the summary results presented in the
NIOH Report.
The reporting of endosulfan in water and soil at concentrations well below the minimum
detection limit of 1 to 3 ppb and the large numbers of “peaks”reported in the blank and
standard reference samples suggests that random peaks due to electronic noise were
routinely misinterpreted as endosulfan peaks.
Chromatographic peak identification is based solely on retention time –no mass
spectrometry confirmation was performed on any of the study samples because
endosulfan concentrations were too low to confirm by GC/MS.
No calibration data are presented and there is no explanation as to how concentrations
in the samples were calculated from instrument readings.
Results from QA/QC samples and analyses required for the analytical method are not
provided, a condition which frequently results in regulatory agencies rejecting and not
relying upon analytical data.
2.1 Excluded Sample Results
Table 1 of the NIOH Report (page 14) provides levels of endosulfan in water samples collected
in 2001. Data are presented for six samples (3 well, 1 suranga, 1 stream and 1 pond);
presumably, though not specified to be from the study area. Concentrations ranged from
- -endosulfan) and the
degradation product (endosulfan sulfate). While Table 1 contains data for six samples,
Scientific Review –NIOH Endosulfan Study
6
Annexure 6 of the NIOH Report (page 82) notes that “a total of seven samples were collected
from the exposed area (Village Vaninagar Padre) and three samples from the reference area
(Miyapadavu, Meenja Gram Panchayat).”Thus, it appears that selected data, including
apparently all of the results from the reference area, have been excluded from the NIOH Report
without explanation. Excluding the results from the reference area precludes any comparison
between the reference and study areas and any valid scientific representation whether the
conditions differ between the areas.
Similarly, Table 1-A of the NIOH Report (page 15) provides summary data for endosulfan
residues in soil samples collected in 2001. The table indicates that eight samples were
collected from the study area and 2 samples were collected from the reference area. However,
Annexure 6 (page 82) states that “eight soil samples were collected in polyethylene bags from
the exposed area (Village Vaninagar Padre) and three samples from the control area
(Miyapadavu, Meenja Gram Panchayat).”Once again, selected data are omitted from the NIOH
Report without explanation. The fact that only sample results from the reference area were
omitted (i.e., the study and reference areas were apparently treated differently) impairs the
scientifically interpretations that can be reached.
In addition, the description provided in the NIOH Report suggests that the soil samples were
likely not collected at locations useful for comparisons between groups of people from the study
village compared to the reference area. The soil sampling locations described for the study
area were not in the village itself, where the children reside, but from the cashew plantations
750 meters or more upslope from the village (page 12). The specific location relative to the
village is not described for the reference area, but there was reportedly no endosulfan spraying
in this area anyway. Comparisons between soil results from the plantations, in the one case,
versus the village, in the other case, are not reflective of the potential exposure differences for
school children.
2.2 Raw Data from RTI Release Not Matching Results Presented in Report
Raw data obtained from NIOH under the Right to Information Act, 2005 indicate that the
summary statistics for endosulfan levels in soil (mean and standard deviation) presented in
Table 4 on page 18 of the NIOH Report contain numerous calculation errors leading to
inaccurate conclusions as itemized below:
-endosulfan concentration reported for the study area (0.274 ppb) in Table 4
-endosulfan concentration measured for the reference area. The
-endosulfan concentration reported for the study area should be reported as
0.222 according to the raw data. Note that if the values from the raw data were
presented in the table, the soil sample results from the study area (0.222 ppb) would be
shown to be lower than the results from the reference area (0.274 ppb). Currently, the
report refers to this table to substantiate the statement that “the levels were higher in
study area as compared to reference area (page 16).”The raw data contradict this
conclusion.
Scientific Review –NIOH Endosulfan Study
7
Accordingly, total endosulfan concentrations in the top layer of soil, which is the layer
where the greatest potential exposures would occur, are also reported inaccurately.
According to the raw data provided by NIOH, the mean total endosulfan concentration in
the upper soil layer in the study area was 0.242 ppb compared to a mean concentration
of 0.303 ppb in the reference area. Again, the concentration found in the study area for
total endosulfan is lower, not higher than that found in the reference area, similarly
contradicting the study conclusion quoted above.
For the middle layer of soil, the raw data provided by NIOH show that the mean
-endosulfan was 0.104 ppb in the study area compared to 0.184 in the
reference area. Contrary to the study conclusion, this difference is not statistically
significant (p = 0.15) at the 95% confidence level, and the study area is, again, lower
than the reference area. Table 4 has these values switched and incorrectly reports a
value of 0.089 ppb instead of 0.104 ppb as derived from the raw data.
In total, 16 out of 24 summary statistics reported in Table 4 of the NIOH Report (66% of
the values) differed from summary statistics calculated from the raw data sheets
obtained from NIOH under the Right to Information Act, 2005 (see Table 1).
Additionally, simple editing and oversight review of the report during its production
should have revealed the presence of errors and need for careful evaluation. There is
no need for access to the raw data to catch that obviously erroneous information is
presented in this table. For example, total endosulfan levels are reported to be less than
-endosulfan levels in some cases, which is impossible.
Table 1. Differences Between Summary Statistics Reported in Table 4 of NIOHReport and Summary Statistics Calculated from Raw Data
Soil
Layer
Area Mean ±SD Difference
(Yes/No)
Constituent Report Table 4 Raw Data
Top Study -endosulfan 0.274 ±0.161 0.222 ±0.133 Yes
Arrebola, F.J., J.L. Martinez Vidal and A. Fernandez-Gutierrez. 2001. Analysis of endosulfanand its metabolites in human serum using gas chromatography-tandem mass spectrometry.Journal of Chromatographic Science. 39(5): 177-182.
Cerrillo, I., Granada, A., Lopez-Espinosa, M-J., Olmos, B., Jimenez, M., Cano, A., Olea, N., andM.F. Olea-Serrano. 2005. Endosulfan and its metabolites in fertile women, placenta, cordblood, and human milk. Environmental Research 98:233-239.
Goyal, R.K. and H.G. Koshia. 2011. Report of the Committee to Evaluate the Safety Aspects ofEndosulfan. Report submitted to Department of Health & Family Welfare, Government ofGujarat, Gandhinagar, Gujarat, India. March 15.
Gupta, P.K. and R.C. Gupta. 1979. Pharmacology, toxicology and degradation of endosulfan.A review. Toxicology 13: 115-130.
Mathur, H.B., Agarwal, H.C., Johnson, S., and N. Saikia. 2005. Analysis of Pesticide Residuesin Blood Samples from Villages of Punjab. Report of the Centre for Science andEnvironment, Pollution Monitoring Laboratory. New Dehi. March 2005.
National Institute of Occupational Health (NIOH). 2002. Final Report of The Investigation ofUnusual Illnesses Allegedly Produced by Endosulfan Exposure in Padre Village of KasargodDistrict (N. Kerala). Report submitted to the Honorable National Human Rights Commissionby Indian Council of Medical Research. Ahmedabad-380016. July 22.
National Institute of Occupational Health (NIOH). Undated. Report of The Investigation ofUnusual Illnesses Allegedly Produced by Endosulfan Exposure in Padre Village of KasargodDistrict (N. Kerala) (First Report). Report prepared by Indian Council of Medical Research.Ahmedabad-380016.
National Institute of Occupational Health (NIOH). 2010. September 24 Letter Report from P.C.Yadav (NIOH) to B. Mallesham re: Information Provided under the RTI Act, 2005.
Saiyed, H., A. Dewan, V. Bhatnager, U. Shenoy, R. Shenoy, H. Rajmohan, K. Patel, R.Kashyap, P. Kulkarni, B. Rajan and B. Lakkad. 2003. Effect of endosulfan on malereproductive development. Environmental Health Perspectives 111(16): 1958-1962.