A cross-sectional study was performed, based on interviews and collection of semen and blood samples from 31 OP pesticide applicators (exposed group) and 31 non-exposed individuals. Inclusion criteria for subjects of the exposed group included history of work with pesticides and length of residence in Majes for at least two years prior to the study. The age of subjects ranged from 20 to 60 years old. Subjects in control group were included in the study if they had never worked as pesticide applicators and were not currently exposed to pesticides, either occupationally or non-occupationally.
Subjects in both groups were not sick at the time of the study; they had not taken any medication for at least three months prior to the study and had lived at least two years in Majes.
Regarding control group selection, some exposed subjects were asked to recruit a male friend without any exposure to OP or agricultural activity. Also, healthy men who responded to an announcement at the medical center, hotel, or municipality were invited to participate in the study. Subjects from the medical center were staff members. All subjects in the non-exposed group had lived at least two years in Majes.
The study was approved by the Institutional Review Board (IRB) of Universidad Peruana Cayetano Heredia, Lima-Peru. A signed informed consent was obtained from each study participant.
Majes is an agricultural area located in Caylloma, Arequipa. It is one of the main agricultural production areas in the Southern part of Peru. It is situated at 1420 m. above sea level. Its temperate climate makes agricultural production possible almost all year round. OP pesticides are used on a variety of crops including potatoes, alfalfa, onions, tomatoes, garlic, apples and grapes. Methamidophos is the most frequently used OP pesticide in the Majes valley .
Recruitment of study population
Pesticide applicators eligible for participating in the study were identified and recruited by agronomic engineers working in Majes Valley. From the universe of applicators in Majes (150 pesticide applicators), 64 accepted to participate in the study. Of these, 31 fulfilled the inclusion criteria, which were: i) To be working as a pesticide applicator for at least 2 years; ii) To have used pesticides within a week before the questionnaire application and semen sample analysis. Data from questionnaire, and blood and semen samples were collected one day after the last pesticide use reported.
The questionnaire was administered to each pesticide applicator to obtain information on sociodemographic characteristics; agricultural work practices, and knowledge and practice of safety guidelines for pesticide use.
Applicators were asked to define how frequently they use OP pesticides. Data related to the kind of pesticides used, protective measures practiced during application, and management of pesticides and clothes after pesticide application were also recorded.
Semen collection and Analysis
All participating subjects were asked to abstain from ejaculation for 3 to 5 days before semen collection. Information on date, time, spillage, occurrence of fever, and number of days since last ejaculation was recorded for each sample. The semen was collected by masturbation in a private room and analyzed on-site within one hour for both macroscopic and microscopic characteristics. Also, seminal fructose, true-corrected fructose, and zinc levels were determined .
Semen analysis was performed according to the protocol described in the WHO laboratory manual for the examination of human semen and sperm-cervical mucus interaction. Semen analysis included: liquefaction time, seminal volume, pH, sperm concentration, total sperm number (sperm concentration × seminal volume), sperm morphology, sperm motility, sperm viability and concentration of leukocyte determined by peroxidase staining using orthotoluidine blue method . A Neubauer improved bright-line hematocytometer (Marienfeld Laboratories, Germany) was used to determine sperm and leukocyte concentrations.
Time of liquefaction was measured as the time in minutes it took for the semen specimen to liquefy. Fresh ejaculated human semen is coagulated, however, due to proteolytic enzymes present in the prostatic secretions, the semen will liquefy. If liquefaction is not complete, the examiner may observe gel particles or mucous streaks.
According to the WHO, a semen sample liquefies within 15 minutes at room temperature however the normal range extends to 60 minutes 
Volume was also assessed macroscopically. The volume of ejaculate was measured in milliliters using a graduated cylinder. The normal or reference value for volume of ejaculate is considered to be 2 milliliters or more by WHO standards .
Morphology describes the shape of the sperm (including the head, neck, mid-piece and tail) and is expressed as the percentage of sperm that meet "normal" morphology criteria defined by WHO. Sperm morphology was evaluated microscopically. A smear from the fresh semen sample was stained with Giemsa in order to view and count the number of normal and abnormal spermatozoa until at least 200 consecutive spermatozoa were evaluated. Generally, the normal value for sperm morphology is at least 30%. No strict criteria were used for the assessment of sperm morphology.
Viability measures the percentage of sperm that are alive because it is possible that sperm may be alive but not moving. Sperm viability was assessed using an Eosin dye. Under the microscope, the examiner was able to differentiate the live (unstained) sperm from the dead (stained) sperm and calculate the percentage of viable sperm. The reference value for sperm viability is 75% . Sperm motility was also analyzed microscopically. Motility describes the percentage of sperm that are moving.
Using an unstained sample of fresh semen, the number of motile sperm was counted until a total of 200 spermatozoa were assessed. The procedure was conducted twice to improve accuracy of the measurement. Motility is classified in four categories:
If sperm has a rapid and linear movement,
If sperm has a slow or sluggish linear or non-linear movement,
If it has a non-progressive motility,
According to the WHO, 50% or greater motility grades a and b or 25% or greater sperms grade a are considered normal . Sperm concentration or density, sometimes referred to as the "count," is a measurement of millions of sperm per milliliter. By looking at a diluted semen sample under the microscope, the examiner counted the number of spermatozoa in a defined field of view. The procedure was repeated to ensure accuracy and a conversion factor was used to calculate the concentration. The normal sperm concentration value is at least 20 million spermatozoa per milliliter 
Knowing the sperm concentration and volume, the total number of sperm in the ejaculation was calculated (sperm concentration × volume = total number of sperm). Total number of sperm is expressed in millions of spermatozoa per ejaculation. The reference value for total sperm number is at least 40 million spermatozoa per ejaculation .
Additionally, the concentration of zinc in seminal plasma was determined by spectrophotometric method using a commercial kit (RANDOX Laboratories, United Kingdom). The proteins in the sample were precipitated with trichloroacetic acid, the supernatant mixed with a water-soluble pyridylazo dye and the absorbance measured at 560 nm. Also, fructose was measured in semen using a spectrophotometric method.
True-corrected fructose is calculated by multiplying the log of motile sperm concentration by the seminal fructose concentration . Although true-corrected fructose is not included in the WHO manual, previous studies have demonstrated that true-corrected fructose is a better marker of seminal vesicle function .
Fructose is a compound secreted by the seminal vesicles. Measurement of seminal fructose used universally as a marker of the seminal vesicle function is not an appropriate approach due to its inverse relationship with the sperm count. The true corrected fructose defined as [log. motile sperm concentration] multiplied by [seminal fructose concentration] has been shown to be a better marker of the seminal vesicle function .
Zinc is a marker of prostate function . Low zinc concentration associated to high pH is associated to low prostatic secretion .
Leukocytes in semen were measured with peroxidase stain using ortholuidine blue as suggested by WHO. The percentage of peroxidase positive neutrophiles was recorded. Leukocytospermia is diagnosed if semen sample has more than 1 million leucocytes peroxidase positive . All seminal analyses were performed in duplicate by the same investigator who did not know about the characteristics of the subjects.
Blood collection and reproductive hormones assay
Venous blood samples were obtained after a 12-h overnight fast. Blood was centrifuged at 1000 g, and serum was collected after centrifugation and kept frozen until assayed for serum testosterone (T), estradiol (E2), follicle stimulating hormone (FSH), and luteinizing hormone (LH) concentrations (Diagnostic Products Co., Los Angeles, CA, USA).
T and E2 concentrations were determined by radioimmunoassay (RIA) using 125I-labeled testosterone and 125I-labeled estradiol, respectively, as radioactive markers (Diagnostic Products Co., Los Angeles, CA, USA).
The assays were performed using commercially available kits (Diagnostic Products Co., Los Angeles, CA, USA). All samples were run in the same assay. The within-assay variation was 5.5% for testosterone and 6.4% for estradiol. The level of detection of the testosterone and estradiol assays was 0.04 ng/ml and 8.0 pg/ml respectively.
Serum FSH and LH levels were measured by immunoradiometric assay (IRMA) in solid phase using commercially available kits (Diagnostic Products Co, Los Angeles, CA, USA). Within-assay variation was 2.9% for FSH and 1.3% for LH. The sensitivity of the assays was 0.06 mIU/ml for FSH, and 0.15 mIU/ml for LH.
Urine collection, storage and organophosphate determination
One day after applying OP pesticides, each worker was provided with one polyethylene urine collection bottle and instructed to collect a urine sample. For this purpose, the first avoid in the morning was collected. All the collected urine samples were immediately placed inside a plastic container with ice and transported to the medical center for freezing at -20°C. The time between urine collection and freezing was 10–15 minutes.
After collection was completed, all samples were shipped frozen to Pacific Toxicology Laboratories (Los Angeles, California U.S.A) where the following metabolites of organophosphates were measured: Dimethylphosphate (DMP), Dimethylthiophosphate (DMTP), Dimethyldithiophosphate (DMDTP), Diethylphosphate (DEP), Diethylthiophosphate (DETP) and Diethyldithiophosphate (DEDTP). All urine samples were stored at -20°C until extraction. Urine pH was not adjusted prior to freezing.-20°C
The standards of DMP (100% purity), DMTP (99% purity), DMDTP (98% purity), DEP (98.3% purity), DETP (99% purity), and DEDTP (99% purity) were obtained from (Cerilliant Corporation, TX, USA).
For extraction, freeze-dried urine samples were treated with a benzyltolytriazine reagent (Sigma-Aldrich Inc., Steinheim, Germany) to produce benzyl derivatives of alkylphosphate metabolites. A saturated salt solution was added to the tubes and the benzyl derivatives were extracted with cyclohexane (Sigma-Aldrich Inc., Steinheim, Germany) and analyzed by gas chromatography with flame photometric detection .
Likewise, the quality control was made in-house by spiking normal urine sample. We run 2 levels of in-house made urine controls. The assay was run with a reagent water blank and urine blank. The recovery rate ranged from 80 to 120% of expected value. R2 (Coefficient of determination) correlation data of methylated and ethylated samples were: (DMP = 0.99), (DMTP = 0.99), (DMDTP = 0.99), (DEP = 0.99), (DETP = 0.99), and (DEDTP = 0.99).
The limit of detection was 5 μg/l for DMP, DEP, DETP and DMTP, and 10 μg/l for DEDTP and DMDTP. Creatinine was also measured in the urine samples by a colorimetric method (Creatinine Procedure No 555; Sigma Diagnostics, St Louis, Mo). Its measurement was used to adjust results of OP metabolites (ug/gram creatinine) to avoid the variable dilution caused by the different hydration states of the sample donor.
Data recorded in the questionnaires, semen and serum samples were introduced in an Excel database. Statistical analysis was performed using the statistical package STATA (version 8.0) for personal computer (Stata Corporation, TX, USA). Descriptive data were presented as mean ± standard deviation (SD), as well as frequencies. The percentage of subjects with detected OP metabolites in urine (percentage of samples above detection limit for each analyte) was also calculated. Many urine samples had concentrations below detection limits of some metabolites.
For the dialkylphosphates metabolites, the samples below the respective limit of detection (LOD) were assigned to have concentrations equal to one-half the LOD for statistical analyses as used by others in a previous report .
The influence of the variables was evaluated for single alkylphosphates and for the sum of dimethyl (DMP + DMTP + DMDTP), diethyl (DEP + DETP + DEDTP). We named these sums methylated, and ethylated OP metabolites, respectively. To calculate these sums, analytes below the detection limit were counted as a value half the detection limit.
Statistical analysis of the samples was then carried out, including a value half the detection limit for non detectable analytes. We used the Kolmogorov-Smirnov test to check the distribution of samples for the six alkylphosphates; we found a positive asymmetric distribution, which became normal after log transformation. Parametric analysis (multiple regressions) was therefore used for subsequent comparisons. Statistical significance was set at P < 0.05.
Homogeneity of variances was assessed using the Barlett test. Variables that were not normally distributed (corrected fructose, seminal zinc, sperm/ml, sperm/ejaculum, motility grade a, a+b, seminal leukocytes, serum levels of E2, FSH, LH, T/E2 and T/LH) were transformed. Student's t test was used to compare the mean between groups.
Multivariable regression analyses were performed to explore the relationship between the following dependent variables: OP metabolites levels (methylated and ethylated), time of exposure (hours worked as a pesticide applicator) and toxicity (related to the severity degree of OP pesticide used) with respect to each of the semen parameters or hormone levels as dependent variables. Each multivariate regression analyses were controlled for age and alcohol consumption. One problem with multiple comparisons is that the greater the number of tests, the higher the likelihood of falsely rejecting the null hypothesis. For this, we have used the Holm's test for correcting multiple comparisons.