Incidence of Acute Pesticide Poisoning Among Cotton Growers in India 113or were hospitalized. On the contrary, farmers rarely stop working for more than a
day. This finding confirms the serious underestimation of statistics based on offi-
cial medical records (Keifer, 1996; Murray, 1994).
Low-income marginal farmers are more often subjected to severe poisoning
than landlords. Smallholders and landless people often apply pesticides through-
out the season as waged. Repeated exposure, in addition to malnutrition and other
diseases, might explain the higher vulnerability of these groups (WHO, 1990;
Repetto and Baliga, 1997). Indeed, pesticide toxicity and exposure time were pos-
itively correlated to the extent to which symptoms were experienced in this survey,
while formal education and landholding were negatively correlated to this measure
of ill health. Yet, only 29 per cent of the variation in symptom severity could be
explained by these factors.
However, more research is needed on factors contributing to the health of
people exposed to pesticides, in particular high-risk groups that are rarely included
in health surveillance on pesticides’ effects (Moses, 1993; Zham and Blair, 1993).
The survey aimed primarily to raise farmers’ awareness on the seriousness of
the poisoning occurring in the villages. It also aimed to quantify the problem by
direct farmers’ reporting. The method has some limitations. Murphy’s article
(Murphy et al, 2002) includes a detailed strength and weakness analysis of the
method. We have reported here only those aspects which are relevant to this sur-
vey. Since signs and symptoms of acute poisoning are non-specific, the health data
generated can be taken only as estimates. Whether the women over- or underre-
ported the true extent of the problem cannot be determined without biomarkers.
A gender bias related to the difference in reporting methods between women and
men could also have been introduced. Self-monitoring data would need to be
backed up by clinical data and blood sample analyses, such as cholinesterase depres-
sions. Another issue is that respondents belonging to the same village had close
interactions. This may have introduced a systematic bias yielding homogeneity of
reporting. Finally, the method cannot appreciate the chronic consequences of pro-
longed exposure to pesticides. Relevant in the case of women are the long-term
effects on the reproductive system that can lead to abortions, still births, neonatal
deaths and congenital defects (Restrepo, 1990; Taha and Gray, 1993; Zhang et al,
1992; Rojas et al, 2002). Nevertheless, a study conducted in India has shown that
female cotton workers experienced the same long-term consequences of exposure
to pesticides (Rupa et al, 1991). Our research concerned only adult respondents
(above 18 years) and no age factor on the severity of the poisoning was found.
However, the Pan American Health Organization estimated that between 10–20
per cent of all poisoning cases involve children. The cottonseed industry in India
employs thousands of female children from 7 to 14 years old to manually cross-
pollinate the plants. There is a need to investigate the impact on children exposed
to pesticides. The survey covered one cotton season and therefore the number of
records is limited. A second period of data collection is, however, scheduled for
2004 with the same respondents to estimate changes in farmers’ health induced by
the cotton IPM FFS.