Organophosphate poisoning has been associated with
chronic neurobehavioral dysfunction, but no epidemio-
logic data exist with regard to long-term consequences
from carbamate poisoning. This cross-sectional study eval-
uated the neurobehavioral performances of 81 banana
workers who, on average 27 months earlier, had received
medical attention not requiring hospitalization for mild
occupational poisoning by either an organophosphate or
a carbamate pesticide. These performances were com-
pared with those of 130 banana workers who had never
sought medical attention for pesticide poisoning. Poi-
soned subjects did less well than controls on tests measur-
ing psychomotor and visuomotor skills, language func-
tion, and affect, the differences being significant for
coding skills on the Digit-Symbol test and two tests of neu-
ropsychiatric symptoms. These deficits, in particular a
marked increase of neuropsychiatric symptoms, occurred
among the organophosphate-poisoned subjects, but small
deficits in performance were also seen in the carbamate-
poisoned subjects. The performances of the previously
poisoned subjects who had had contact with cholin-
esterase inhibitors within three months before testing
were particularly poor. These findings in workers with
mild poisoning are consistent with previous findings of
persistent damage to the central nervous system from
organophosphate poisoning. The possibility of persistent
neurobehavioral effects associated with poisonings by n-
methyl carbamate insecticides cannot be excluded. Work-
ers with histories of poisoning may be more susceptible to
neurobehavioral effects with subsequent exposures. Key
words: carbamate; cholinesterase inhibitor; Costa Rica;
cross-sectional study; developing country; epidemiology;
neurobehavioral effects; occupational exposure; organo-
phosphate; pesticides; poisoning.
INT J OCCUP ENVIRON HEALTH 2002;8:27–34
terase-inhibiting pesticides are also a public health
problem in some industrialized countries.3The possibility
of long-term neurotoxic effects from poisonings caused
by cholinesterase-inhibiting pesticides is a global concern.
Cholinesterase inhibitors bind to acetylcholin-
esterase, the enzyme that controls the transmission of
the nerve impulses at the cholinergic synapses through-
out the nervous system. This results in the accumulation
of acetylcholine at the neurojunctions and the appear-
ance of a cholinergic syndrome.4Organophosphates
bind irreversibly to acetylcholinesterase, and subsequent
“aging” of the phosphorylated enzyme may occur,
which makes the reactivation of cholinesterases depend-
ent on the synthesis of new enzymes. Chronic nervous
system effects have been observed among workers pre-
viously poisoned with organophosphate pesticides, in
case reports of poisoned patients5-7and in well-designed
Unlike organophosphates, carbamates occupy the
binding sites at the target enzymes in a reversible way
and without the “aging” reactions. Persistent neuro-
logic effects are therefore assumed to be unlikely.4
However, few carbamates have been submitted to
extensive neurotoxicity testing.11Some case reports
describe chronic neurologic effects from carbamate
poisonings,11-14but no results from analytic epidemio-
logic studies exist in the published literature.
In Costa Rica, both organophosphate and carbamate
nematocides (worm killers) are widely used on banana
plantations. In 1986, the banana plantation region of
majority of the pesticide poisonings in develop-
ing countries is caused by neurotoxic organo-
phosphates and carbamates.1,2These cholines-
Long-term Neurobehavioral Effects of
Mild Poisonings with Organophosphate and
n-Methyl Carbamate Pesticides
among Banana Workers
CATHARINA WESSELING, MD, PHD, MATTHEW KEIFER, MD, MPH,
ANDERS AHLBOM, PHD, ROB MCCONNELL, MD, JAI-DONG MOON, MD, PHD,
LINDA ROSENSTOCK, MD, MPH, CHRISTER HOGSTEDT, MD
Received from the Universidad Nacional, Heredia, Costa Rica
(CW); the University of Washington, Seattle, Washington (MK); the
Karolinska Institutet, Stockholm, Sweden (AA, CW); the University
of Southern California, Los Angeles, California (RM); Chonnam
National University Medical School, Kwangju, Korea (JDM); the Uni-
versity of California, Los Angeles, California (LR); and the National
Institute of Public Health, Stockholm, Sweden (CH) The study was
carried out at the Central American Institute for Studies on Toxic
Substances in Costa Rica. Supported by the Department for Research
Cooperation of the Swedish International Development Coopera-
tion Agency (SAREC/Sida); the United States Environmental Pro-
tection Agency (US-EPA); and the project Environment and Health
in the Central American Isthmus (MASICA–Costa Rica) of the Pan
American Health Organization (PAHO).
Address correspondence and reprint requests to: Catharina Wes-
seling, Central American Institute for Studies on Toxic Substances,
Universidad Nacional, Apdo 86-3000 Heredia, Costa Rica; telephone:
(506) 277-3584; fax: (506) 277-3583; e-mail: <firstname.lastname@example.org>.
Costa Rica had one of the highest incidence rates of poi-
sonings reported in the world.15Since the 1990s,
improvements on banana plantations have diminished
the incidence of poisonings,16and those that occur are
usually fairly mild. This study was undertaken to deter-
mine whether poisoning with an organophosphate or a
carbamate compound is associated with long-term
effects on the central nervous system.
The performances of previously poisoned workers on a
battery of neurobehavioral tests were compared with
the performances of a control group of non-poisoned
workers, poisoning versus no poisoning being the
measure for exposed versus unexposed.
This study was conducted in two counties in the Atlantic
Region of Costa Rica. It was done on Costa Rican men,
aged 15–55 years at the time of testing, who had been
banana workers during at least one year and who did
not have a history of any disease or condition that might
induce nervous system damage or interfere with proper
testing (severe head trauma, severe alcohol or drug
abuse, hearing impairment, inability to “read” numbers,
history of epilepsy, or serious illness such as cancer).
Subjects who had had systemic occupational poison-
ings caused by cholinesterase-inhibiting pesticides
treated at any of the health centers in the study area at
least one year prior to testing were considered eligible for
inclusion in the poisoned (exposed) group. Poisoned sub-
jects were identified from compulsory occupational acci-
dent reports to the National Insurance Institute. Name,
personal identification number, home address, planta-
tion, and pesticide were extracted from the reports. The
poisoned workers were then traced by screening the pay-
rolls of all banana plantations in the study area during
May–June and again in August 1994. We were able to
locate 94 (58%) of the 162 poisoned workers identified
in the record search. Of the 82 eligible poisoned workers,
one refused to participate. The final exposed group con-
sisted of 81 poisoned workers (mean number of months
since poisoning = 27, range 12–43).
A control group of non-poisoned (unexposed) work-
ers was randomly drawn from the payrolls of participat-
ing banana plantations (one of the four large multina-
tional companies in the study area and practically all
other smaller companies). Subjects who had histories
of receiving medical treatment for pesticide poisoning
at any time of their lives were not eligible. Of the 144
eligible controls, ten refused to participate and four
were excluded because a plantation manager had arbi-
trarily substituted them for the randomly chosen work-
ers (response rate 90%). The final unexposed group
consisted of 130 non-poisoned workers. These controls
included field workers and employees in the packing
plants who had never had any contact with cholin-
esterase-inhibiting pesticides, as well as field workers
who worked with cholinesterase-inhibiting pesticides in
varying degrees. A subset of these workers reported
having experienced at any time in their life mild symp-
toms associated with pesticide exposures without need
for medical attention. Some of the results are reported
separately for this subgroup of controls.
Information about demographic variables and diseases
or conditions that could affect performance on the
tests was collected by means of a structured interview
(Table 1). A complete work history was taken. Indices
of lifetime exposures to cholinesterase-inhibiting pesti-
cides as well as to other pesticides were constructed
based on the number of days of exposure per year, the
number of years exposed, the fraction of the day
exposed per job title, and the use of protective equip-
ment. Near visual acuity (poor, mediocre, good) was
assessed with a short printed text and a chart with num-
bers. Confounders additional to those in Table 1 that
were considered in the analysis included medical con-
founders such as history of convulsions and other neu-
rologic or psychiatric disorders, malaria, chronic meta-
bolic and infectious disorders; current use of
medication; examiner; hours of sleep the night prior to
testing; and sense of well-being and liquor and caffeine
intakes on the day of testing.
It was not feasible to conduct the field work outside
a spraying season because different plantations spray
nematocides at different times. Thus, a substantial pro-
portion of the study population had been in contact
with cholinesterase-inhibiting pesticides within three
months of the examination (Table 2). None of these
workers reported symptoms of poisoning at the time of
testing. Red blood cell and plasma cholinesterase activ-
ities were determined with the Test-Mate organophos-
phate kit (EQM Inc., Cincinnati, Ohio),17which uses a
modified Ellman method and corrects for hemoglobin.
There was no difference between the mean tempera-
ture-adjusted red blood cell cholinesterase activity
levels in the poisoned and non-poisoned groups, or
between subsets with and without recent exposure. The
mean value of plasma cholinesterase levels of the work-
ers with recent contact with cholinesterase inhibitors
was lower than the mean for the workers without recent
contact (2.11 U/mL versus 2.26 U/mL blood, p = 0.01).
All of the 81 poisoned subjects reported that poison-
ings had not been severe enough to require hospital-
ization. This was confirmed by a review of medical
records for all but 20 subjects. Cholinesterase levels
28•Wesseling et al.
INT J OCCUP ENVIRON HEALTH
were seldom reported, and clinical descriptions in out-
patient records were not adequate to further assess
severity of the poisonings. However, the workers were
interviewed regarding the poisonings with a checklist
of 23 typical cholinergic symptoms.4,18The mean
number of symptoms was 13.7 (SD 3.9), with almost
identical numbers for the subjects with and without
review of their medical records. The most frequently
reported poisoning symptoms included lightheaded-
ness (94%), nausea (91%), general weakness (91%),
abdominal pain (89%), excessive sweating (83%), sali-
vation (79%), headache (78%), vomiting (75%),
blurred vision (75%), muscle twitching (67%), muscle
cramps (56%), and difficulty breathing (51%).
Seizures and involuntary loss of urine, which are symp-
toms of severe poisoning, were acknowledged by 3%
and 6%, respectively. Fifteen percent answered posi-
tively to “sneezing,” which had been included as a
symptom not likely to be related to poisoning in order
to be able to adjust for overreporting of poisoning.
Sources of information about the specific pesticides
responsible for the poisonings were the medical file,
the report to the National Insurance Institute, the
worker’s interview, and information provided by the
company about which nematocide had been used on
the day of the poisoning. When the sources did not
agree, the company’s information was used. Hospital
information was used in a documented epidemic of
carbofuran poisonings. In all but three cases, the poi-
sonings could be classified as an organophosphate (n =
54) or a carbamate (n = 24) (Table 2). Of the
organophosphate poisonings, 18 were caused by terbu-
fos, six by fenamifos, four by ethoprophos, two by diazi-
non, and one each by cadusaphos and chlorpyrifos.
Twenty-two pesticides were classified only as “organo-
phosphates.” Of the carbamate poisonings, 23 were
caused by carbofuran and one by oxamyl.
Previously trained technicians, who were blind to the
poisoning status of the subjects, examined all workers
between May and September 1994. Written informed
consent was obtained. The costs of travel, food, and lost
wages were reimbursed, but the workers did not receive
any additional incentive.
A test battery was assembled to assess the main
domains of neurobehavioral functions, including
VOL 8/NO 1, JAN/MAR 2002
Chronic Neurotoxicity from Pesticides• 29
TABLE 1 Characteristics of the Study Population
n = 81
n = 130Characteristic
Age, mean (SD)
Formal education, mean (SD)
Contact with cholinesterase inhibiting pesticides within
3 months before testing
Cumulative lifetime exposure to cholinesterase-inhibiting pesticidesa
Cumulative lifetime exposure to non–cholinesterase-inhibiting pesticidesa
Red blood cell cholinesterase level, mean (SD)
Plasma cholinesterase level, mean (SD)
Ever heavy drinkers
Time of day of testing
Worked with solvents > 3 months daily and continuously
Loss of consciousness > 1 hour
Poor or mediocre near visual acuity
28.3 years (7.6)
5.5 years (2.1)
29.0 years (10.0)
6.0 years (2.2)
29.6 U/mL (3.4)
2.18 U/ml (0.41)
29.6 U/ml (2.9)
2.22 U/ml (0.41)
aExposure index = SUM (number of days per year ? number of years ? fraction of the day per job title ? weight for
no use of protective equipment).
memory, attention, psychomotor and visuomotor abili-
ties, language, and affect (Table 3), based on the results
from previous research on persistent neurotoxic effects
from organophosphate poisonings.8-10We used the
World Health Organization Neurobehavioral Core Test
Battery (WHO-NCTB),19which has produced good
results in different cultural contexts, except for the Pro-
file of Mood States.20Instead, we used the Swedish
Questionnaire-16,21validated in Nicaragua,22and the
Brief Symptom Inventory (BSI).23In view of the many
relatively uneducated subjects in our study population,
we administered the BSI (a self-administered test) orally
with the help of a visual scoring scale. To assess the neu-
robehavioral domains more comprehensively, this core
battery was complemented with other standardized,
internationally well-known tests, as listed in Table
3.19,21–27The Wechsler vocabulary test was included as a
“hold” test to assess premorbid intelligence.28All tests
had been used in an earlier, similar study of more severe
hospitalized poisoning cases in Nicaragua.9
Potential associations between poisoning and neurobe-
havioral deficits were assessed by means of multiple
linear regression analyses. Confounding was assessed
by determining for each outcome whether the crude
difference of means changed meaningfully following
the addition of potential confounders to a univariate
model with poisoning status as the explanatory vari-
able, or to a model including poisoning status, age, and
education.29Age, education, cumulative lifetime expo-
sure to cholinesterase inhibitors, alcohol intake, sol-
vent exposure, loss of consciousness, examiner, time of
the day of testing, and plasma cholinesterase level at
the time of examination were entered as covariates in
all the models. In addition to adjustment for plasma
cholinesterase activity in the multivariate analysis, con-
founding by recent exposure to cholinesterase-inhibit-
ing pesticides was controlled by stratification into sub-
sets with and without contact with cholinesterase
inhibitors within the preceding three months. Seven
subjects whose near visual acuity was poor were elimi-
nated from the Benton, Pursuit aiming, Digit vigilance,
Digit-symbol and Trails-A tests.
Unstandardized and standardized coefficients along
with their 95% confidence intervals were estimated for
all poisoned workers, for organophosphate- and carba-
mate-poisoned workers, and for the subgroups with
and without recent contact with cholinesterase
inhibitors (see Table 2). In addition, we performed
analyses comparing the performances of the poisoned
subjects with those of the subset of controls restricted
to people who had never experienced any symptom
attributable to pesticide exposure. The analyses were
conducted with the BMDP statistical software.30
30• Wesseling et al.
INT J OCCUP ENVIRON HEALTH
TABLE 2 Numbers of Subjects in the Exposure Categories and Subgroups
Number of Subjects
Contact with Cholinesterase
Inhibitors within 3 Months
NoExposure Category YesTotal
Poisoned with medical attention
Poisoned by organophosphates
Poisoned by n-methyl carbamates
Poisoned by nonspecified cholinesterase inhibitor
Never any symptoms from pesticide exposures
Ever an episode of pesticide-related symptoms without
TABLE 3 Characteristics of the Study Population
Benton visual retention
Rey verbal learning
Pursuit aiming II
Simple reaction time
The poisoned group as a whole performed less well
than did the non-poisoned comparison group on 13 of
the 14 neurobehavioral tests (Table 4), but the differ-
ences on most tests were small. The largest deficits
among the poisoned subjects were observed for tests of
psychomotor and visuomotor functions, language, and
affect. The poisoned subjects performed significantly
less well on the Digit-symbol test (beta = –2.7, 95% CI
–5.3, –0.1) and, in addition, reported significantly
more neuropsychiatric symptoms than did the non-poi-
soned workers on the Questionnaire-16 (beta = 1.7,
95% CI 0.5, 3.0) as well as the BSI (beta = 15.7, 95% CI
Figure 1a shows the adjusted standardized coeffi-
cients after stratification into poisonings by organo-
phosphate and carbamate pesticides. The deficits
occurred among both the organophosphate- and the
carbamate-poisoned subjects, but more clearly among
the organophosphate-poisoned subjects. Neuropsychi-
atric symptoms on the Questionnaire-16 and the BSI
were significantly increased among the organophos-
phate-poisoned subjects, but the excesses among the
carbamate-poisoned subjects were small and not statis-
tically significant. Figure 1b shows the results after strat-
ification into subsets with and without handling of
cholinesterase-inhibiting pesticides within three
months before testing. Deficits for the poisoned sub-
jects were observed in both subsets, but were more
apparent in the subset of workers who had been in
recent contact with cholinesterase-inhibiting pesti-
cides. However, the excess reporting of neuropsychi-
atric symptoms by poisoned workers was similar for
those with and without recent exposure to cholines-
The analyses restricted to the subgroup of the refer-
ence workers who had never experienced poisoning
symptoms in connection with pesticides yielded results
similar to those for all controls in the models. (Results
are not shown.)
Workers previously poisoned with cholinesterase-
inhibiting pesticides tended to perform less well on
psychomotor and visuomotor tests as compared with a
non-poisoned control group. These, mainly nonsignifi-
cant, differences were observed in the organophos-
phate-poisoned subgroup, but small deficits in per-
formance were also observed for the carbamate-
(mostly carbofuran-) poisoned subjects. A marked
increase in neuropsychiatric symptoms was observed in
the organophosphate-poisoned workers. Neurobehav-
ioral deficits of previously poisoned workers, as com-
pared with non-poisoned workers, were more apparent
within the subset who had worked with cholinesterase
inhibitors within the three months before testing.
VOL 8/NO 1, JAN/MAR 2002
Chronic Neurotoxicity from Pesticides• 31
TABLE 4 Neurobehavioral Performances of Banana Plantation Workers Previously Poisoned with a Cholinesterase-
inhibiting Pesticide, as Compared with Non-poisoned Workers
Mean (SD) Mean (SD)
Neurobehavioral Testn = 81 n = 130
Benton visual retention (no. correct)
Rey verbal learning (no. after
Digit vigilance (s)
Digit span (forward)
Santana dexterity (both hands)
Pursuit aiming II (no. correct dots)
Finger tapping (both hands)
Mean simple reaction time (ms)
Questionnaire—16 (no. yes
Brief symptom inventory
6.3 (2.2)6.9 (2.1) –0.5* –0.4–1.0, 0.2–
9.1 (4.2)6.5 (3.8)2.6‡ 1.7‡ 0.5, 3.0–
45.8 (25.3)26.8 (40.7 19.0‡ 15.7‡5.5, 25.8–
aThe poisoned subjects performed worse on all tests, except digit vigilance, where no difference was observed.
bAdjusted for age, education, long-term exposure to cholinesterase inhibitors, alcohol intake, solvent exposure, loss of conscious-
ness, examiner, time of day of testing, and plasma cholinesterase levels.
cCI: Confidence interval.
d(–) The poisoned subjects performed worse; (+) the poisoned subjects performed better.
*p < 0.10; †p < 0.05; ‡p < 0.01
32• Wesseling et al.
INT J OCCUP ENVIRON HEALTH
Figure 1—Neurobehavioral performances of previously poisoned banana plantation workers as compared with those of non-poisoned workers, stratified (A) by type of
cholinesterase inhibitor and (B) by recent contact with cholinesterase inhibitors. Standardized regression coefficients with 95% confidence intervals adjusted for age, education,
long-term exposure to cholinesterase inhibitors, alcohol intake, solvents, loss of consciousness, examiner, time of day of testing, and plasma cholinesterase levels.
Overall, these results are consistent with growing lit-
erature that suggests that poisoning with organophos-
phate insecticides causes persistent damage to the nerv-
ous system.5-14,31The diffuse profile of deficits that we
observed is in accordance with the findings of other
studies.8-10Also, in these three previous studies signifi-
cant deficits on the Digit-symbol test were observed.
This study was designed to be comparable to one of
these studies, of Nicaraguan agricultural workers who
had been hospitalized because of organophosphate poi-
soning.9Our study of less severely poisoned patients
(not requiring hospitalization) demonstrated an overall
pattern of neurobehavioral effects similar to that in the
Nicaraguan study, but the deficits were smaller. Steen-
land et al. also found evidence of increasing neuropsy-
chological impairment with increasing severity of
This is the first epidemiologic study that examined
neurologic effects from carbamate poisoning, though
the number of cases (24) was small. The deficits were
small and the confidence intervals wide, but chronic
neurotoxic effects, suggested by previous case reports,11,14
cannot be excluded. Further epidemiologic and exper-
imental data are needed, especially for carbofuran,
which accounted for all but one of the carbamate poi-
sonings in our data.
The deficits in the performances of poisoned sub-
jects in the subset with contact with cholinesterase
inhibitors within the three months before testing were
considerably larger than were those in the subset with-
out recent exposure to cholinesterase-inhibiting pesti-
cides (see Figure 1b). Within this subset with recent
contact with cholinesterase inhibitors, the previously
poisoned workers had a lower mean value of plasma
cholinesterase activity than did the controls (2.05
U/mL versus 2.20 U/mL, p = 0.12), which may indicate
higher recent exposures among the poisoned subjects.
However, the results were adjusted for plasma
cholinesterase levels and, in addition, no difference
was found for red blood cell cholinesterase activity,
which is a better marker of acute effects on the nervous
system.4Therefore, even if there were an additional
effect from recent exposure, the observed deficits
cannot be explained by recent exposure.
Several possible sources of bias could have influ-
enced the results. For 25% of the poisoned workers,
the medical record could not be obtained to confirm
the clinical diagnosis of poisoning with a cholinesterase
inhibitor. However, the diagnosis was confirmed for all
61 workers for whom medical records were located.
Exclusion from the analyses of unconfirmed poison-
ings did not produce appreciable changes in the effect
estimates. Medical doctors in banana plantation areas
are unlikely to misdiagnose pesticide poisoning, which
is common and has characteristic clinical features.
Therefore, misdiagnosis of pesticide poisoning is an
unlikely source of bias.
Banana plantation workers are a distinctive social
group, unlike other agricultural or industrial worker
populations. The random selection of the controls from
the same population guaranteed identical socioeco-
nomic and occupational backgrounds, which indirectly
controls for a number of important potential con-
founders. As a result, however, pesticide exposures in
the comparison group were substantial. Some controls
had experienced pesticide-related symptoms, albeit not
requiring medical attention. The mean number of
symptoms was considerably smaller than among the poi-
soned who had sought medical attention (7 versus 14),
and the symptoms were mostly nonspecific (headache,
nausea, lightheadedness, weakness, vomiting), but an
overlap of this subgroup of comparison workers with
the poisoned group is conceivable. However, when
restricting the analysis to the subset of controls never
having experienced symptoms, the differences between
poisoned and non-poisoned workers did not increase,
as would be anticipated if such bias existed.
Forty-two percent of the identified cases of poison-
ings could not be located. There were no objective
data for comparison of located and non-located work-
ers, which might have made it possible to evaluate the
direction of this potential selection bias. It is likely that
most of those persons still working on banana planta-
tions or living in the region were located. However, the
most affected workers may have been more likely to
leave the banana region due to potential intolerance
of previously poisoned subjects to subsequent pesti-
cide exposures.6,7,32This might have led to the exami-
nation of a healthy subset of the previously poisoned
group, minimizing differences between poisoned
workers and controls.
More non-poisoned than poisoned workers had
received some high school education. In addition, the
lower scores on the vocabulary test for the poisoned
group could indicate lower premorbid intellectual
aptitude of the poisoned subjects.28However, it has
been shown that vocabulary test scores are sensitive to
solvent-induced central nervous system effects.33,34
Also, two of the three major epi-studies of organophos-
phate poisonings found considerable deficits of the
poisoned subjects on the Wechsler vocabulary test,8,9
while the third did not include a language test.10In
any case, adding the vocabulary test score to the model
in the multivariate analyses did not substantially
change the coefficients of the poisoning status after
controlling for education.
In practice, the cleverest workers often are selected
for dangerous tasks such as nematocide spraying and,
thus, may be more likely to get poisoned. To evaluate
this possibility, we compared the performances of con-
trols with recent exposure to cholinesterase inhibitors
(who were mostly nematocide applicators) with those
of controls without recent exposure to cholinesterase
inhibitors. The recently exposed did considerably
VOL 8/NO 1, JAN/MAR 2002
Chronic Neurotoxicity from Pesticides• 33
better on all cognitive tests. Thus, even though the con-
trols were somewhat better educated, it seems unlikely
that they had greater cognitive ability, because only rel-
atively few controls were nematocide sprayers versus a
majority of the poisoned subjects.
The size of our total study population was sufficient
to detect a 10% difference in performances with good
precision on most of the tests,35but the numbers of
exposed subjects in the subgroups became small. In
spite of the wide confidence intervals, the point esti-
mates of the effects were, in general, consistent over
the various categories (see Figure 1), which supports
the validity of the observed results.
In conclusion, this study lends additional support to
previous findings that organophosphate poisonings
have adverse chronic effects on a broad spectrum of
nervous system functions. We observed such effects
among workers with only mild poisonings. We could
not exclude the possibility of persistent neurobehav-
ioral effects associated with poisonings by n-methyl car-
bamate insecticides. Workers who have histories of poi-
soning may be more susceptible to neurobehavioral
effects with subsequent exposures.
The authors are indebted to the Social Security of Costa Rica (Caja
Costarricense del Seguro Social) and to the National Insurance
Institute (Instituto Nacional de Seguros). They thank the banana
companies and, in particular, the workers who participated in the
study. Without the help of César Jarquín of the Department of Toxic
Substances of the Ministry of Health, it would not have been possi-
ble to conduct this study.
1. World Health Organization/United Nations Environmental
Program (WHO/UNEP). Public Health Impact of Pesticides
Used in Agriculture. Geneva, Switzerland: WHO, 1990.
Wesseling C, McConnell R, Partanen T, Hogstedt C. Pesticide
use in developing countries: occupational health impact and
research needs. Int J Health Services. 1997;27:273-308.
Ciecielski S, Loomis DP, Rupp Mims S, Auer A. Pesticide expo-
sures, cholinesterase depression, and symptoms among North
Carolina migrant farmworkers. Am J Public Health. 1994;84:
Ballantyne B, Marrs TC (eds). Clinical and Experimental Toxi-
cology of Organophosphates and Carbamates. Oxford, U.K.:
Rosenstock L, Daniell W, Barnhart S, Schwartz D, Demers PA.
Chronc neuropsychological sequelae of occupational exposure
to organophosphate insecticides. Am J Ind Med. 1990;18:321-5.
Gershon S, Shaw FH. Psychiatric sequelae of chronic exposure
to organophosphorous insecticides. Lancet. 1961;1:1371-4.
Tabershaw IR, Cooper C. Sequelae of acute organic phosphate
poisoning. J Occup Med. 1966;1:5-20.
Savage EP, Keefe TJ, Mounce LM, Heaton RK, Lewis JA, Burcar
PJ. Chronic neurological sequelae of acute organophosphate
pesticide poisoning. Arch Environ Health. 1988;43:38-45.
Rosenstock L, Keifer M, Daniell WE, McConnell R, Claypoole K.
Chronic central nervous system effects of acute organophos-
phate pesticide intoxication. Lancet. 1991;338:223-6.
Steenland K, Jenkins B, Ames RG, O’Malley M, Chrislip D,
Russo J. Chronic neurological sequelae of organophosphate
pesticide poisoning. Am J Public Health. 1994;84:731-6.
11. Ecobichon DJ, Joy RM (eds). Pesticides and Neurological Dis-
eases. Boca Raton, FL: CRC Press, 1994.
Umehara F, Izumo S, Osame M. Polyneuropathy induced by m-
tolyl methyl carbamate intoxication. J Neurol. 1991;238:47-8.
Dickoff DJ, Gerber O, Turovsky Z. Delayed neuropathy after
ingestion of carbamate insecticide. Neurology. 1987;37:1229-31.
Branch RA, Jacqz E. Subacute neurotoxicity following long-term
exposure to carbaryl. Am J Med. 1986;80:741-5.
Wesseling C, Castillo L, Elinder CG. Pesticide poisonings in
Costa Rica. Scand J Work Environ Health. 1993;19:227-35.
Wesseling C, Hogstedt C, Fernandez P, Ahlbom A. Time trends
of occupational pesticide-related injuries in Costa Rica, 1982–
1992. Int J Occup Environ Health. 2001;7:1-6.
McConnell R, Cedillo L, Keifer M, Palomo MR. Monitoring
organophosphate insecticide exposed workers for cholin-
esterase depression: new technology for office or field use. J
Occup Med. 1992;34:34-7.
Namba T, Nolte CT, Jackrel J, Grob D. Poisoning due to
organophosphate insecticides, acute and chronic manifesta-
tions. Am J Med. 1971;50:475-92.
World Health Organization (WHO). Office of Occupational
Health. Operational Guide for the WHO Neurobehavioral Test
Battery. Geneva, Switzerland: WHO, 1986.
Anger WK, Cassito MG, Lian Y, et al. Comparison of perform-
ance from three continents on the WHO-recommended neuro-
behavioral core test battery. Environ Res. 1993;62:125-47.
Hogstedt C, Andersson K, Hane M. A questionnaire approach
to the monitoring of early disturbances in central nervous func-
tions. In: Aitio A, Riihimäki V, Vainio H (eds). Biological Moni-
toring and Surveillance of Workers Exposed to Chemicals.
Washington, DC: Hemisphere Publishing Corporation, 1984.
Amador R, Lundberg I, Escalona E. Development of a ques-
tionnaire in Spanish on neurotoxic symptoms. Am J Ind Med.
Derogatis LR, Spencer PM. The Brief Symptom Inventory:
Administration, Scoring and Procedures Manual—I. Baltimore,
MD: Johns Hopkins University School of Medicine, 1982.
Rey A. L’examen Clinique en psychologie. Paris, France: Presses
Universitaires de France, 1964.
Lewis R, Kapke T. The Lafayette Clinic repeatable neuropsy-
chological test battery. Presented at the Annual Meeting of the
Southeastern Psychological Association, Hollywood, FL, 1977.
Reitan RM, Wolfson D. The Halsteadt–Reitan Neuropsychologi-
cal Test Battery: Theory and Clinical Interpretation. Tucson,
AZ: Neuropsychology Press, 1985.
Wechsler D. Manual for Wechsler Adult Intelligence Scale—
Revised. New York: Psychological Corporation, 1981.
White RF, Proctor SP. Research and clinical criteria for develop-
ment of neurobehavioral test batteries. J Occup Med. 1992;
Kleinbaum DG, Kupper LL, Muller KE. Applied Regression
Analysis and Other Multivariable Methods. Belmont, CA: Dux-
bury Press, 1988.
Dixon WJ (ed). BMDP Statistical Software Manual. Berkeley,
CA: University of California Press, 1992.
London L, Nell V, Thompson ML, Myers JE. Effects of long-
term organophosphate exposures on neurological symptoms,
vibration sense and tremor among South African farm workers.
Scand J Environ Health. 1998;24:18-29.
Dille JR, Smith PW. Central nervous system effects of chronic
exposure to organophosphate insecticides. Aerospace Med.
1964; 35, May:475-8.
Michelsén H, Lundberg I. Neuropsychological verbal tests may
lack “hold” properties in occupational studies of neurotoxic
effects. Occup Environ Med. 1996;53:478-83.
Viaene M, Veulemans H, Masschelein R. Experience with a
vocabulary test for workers previously and still exposed to
styrene. Scand J Work Environ Health. 1998;24:308-11.
Agency for Toxic Substances and Disease Registry (ATSDR).
Adult Environmental Neurobehavioral Test Battery (AENTB),
Examiner Training Manual. Atlanta, GA: Public Health Service,
34• Wesseling et al.
INT J OCCUP ENVIRON HEALTH