Impaired Semen Quality Associated with Environmental DDT Exposure in Young Men Living in a Malaria Area in the Limpopo Province, South Africa

Journal of Andrology, Vol. 28, No. 3, May/June 2007 Copyright E American Society of Andrology

Impaired Semen Quality Associated With Environmental DDT Exposure in Young Men Living in a Malaria Area in the Limpopo Province, South Africa

NATALIE H. ANECK-HAHN,*{ GLORIA W. SCHULENBURG,{ MARIA S. BORNMAN,{ PAULINA FARIAS,§ AND CHRISTIAAN DE JAGER* From the *Environmental Health, School of Health Systems & Public Health, and Andrology, Department of Urology, University of Pretoria, Pretoria, South Africa, and `Department of Urology, University of Limpopo, Medunsa, South Africa; and §Instituto Nacional de Salud Publica, Cuernavaca, Morelos, Mexico.

ABSTRACT: The pesticide DDT [1,1,1-trichloro-2,2-bis(chlorodi- cross-frequency (BCF) was higher with a higher p,p9-DDT concen- phenyl)ethane] is 1 of the 12 persistent organic pollutants (POPs) tration (b 5 0.01, P 5 .000). There was also a statistically significant under negotiation at the Stockholm Convention to restrict or ban positive association between percent sperm with cytoplasmic their production and use because of their toxicity, resistance to droplets and p,p9-DDT concentration (b 5 0.0014, P 5 .014). The breakdown, bioaccumulation, and potential for being transported ejaculate volume (mean 1.9 6 1.33 mL) was lower than the normal over long distances. DDT has estrogenic potential, and the main range ($2.0 mL) according to WHO, and a significant decrease with metabolite, p,p9-dichlorodiphenyl-dichloroethylene (p,p9-DDE), is increasing p,p9-DDE values was seen for both square root– a potent antiandrogen. In response to mounting evidence on the transformed volume (b 520.0003; P 5 .024) and count (b 5 endocrine-disrupting influence of environmental chemicals on human 20.003; P 5 .04). Although there were no associations between health, this epidemiological study was initiated to test the hypothesis either p,p9-DDT or p,p9-DDE concentrations and the rest of the that nonoccupational exposure to DDT affects male reproductive seminal parameters, the incidence of teratozoospermia (99%; parameters. In a cross-sectional study, healthy male subjects (n 5 normal sperm ,15%) was high. Twenty-eight percent of the study 311) between 18 and 40 years (23 6 5) of age were recruited from 3 group presented with oligozoospermia (,20 6 106 sperm/mL), communities in an endemic malaria area in which DDT is sprayed which had a significant positive association with p,p9-DDE (odds annually. A semen analysis according to World Health Organization ratio [OR] 5 1.001, P 5 .03). There was a significant positive (WHO) standards was performed. The Hamilton Thorne Computer association between participants with asthenozoospermia (32%) and Assisted Sperm Analysis (CASA) system was simultaneously used p,p9-DDT (OR 1.003, P 5 .006) and p,p9-DDE (OR 1.001, P 5 .02). to determine additional sperm motility parameters. Blood plasma The results imply that nonoccupational exposure to DDT is samples were assayed for p,p9-DDT and metabolites as a measure associated with impaired seminal parameters in men. The high of exposure. The exposure levels were expressed as lipid-adjusted exposure levels of p,p9-DDT and p,p9-DDE are of concern because p,p9-DDT and p,p9-DDE values. The mean p,p9-DDT and p,p9-DDE these levels could have far-reaching implications for reproductive concentrations were 90.23 mg/g(6102.4) and 215.47 mg/g(6210.6), and general health. respectively. The multivariate linear regression analyses indicated Key words: Seminal parameters, organochlorine pesticides, p,p9- that mean CASA motility was lower with a higher p,p9-DDE DDT, p,p9-DDE, POPs, spermatozoa, CASA. concentration (b 520.02, P 5 .001) and the CASA parameter beat J Androl 2007;28:423–434

he Stockholm Convention resulted from a decision cals. The convention initially targeted 12 chemicals T made in 1995 by the United Nations Environment known as persistent organic pollutants (POPs), arguing Programme Governing Council (UNEP) to develop that those chemicals pose major and increasing threats a legally binding instrument to control certain chemi- to human health and the environment (UNEP, 1995). The Stockholm Convention on POPs became legally binding on May 17, 2004. The Convention is a global The National Research Foundation (NRF) and the Medical multilateral agreement with the aim of protecting Research Council (MRC), South Africa funded this study. Correspondence to: C de Jager, PhD, Environmental Health, School human and environmental health from the effects of of Health Systems & Public Health, University of Pretoria, PO Box 667, exposure to specific POPs. Restricting the use and Pretoria, 0001, South Africa (e-mail: [email protected]). production of these chemicals or banning them will, Received for publication September 29, 2006; accepted for when the measures of the convention are successfully publication December 18, 2006. The authors declare no financial or personal conflicts of interest that implemented, reduce the hazards posed by these might affect any aspect of this study. pollutants. Although South Africa ratified the Conven- DOI: 10.2164/jandrol.106.001701 tion on September 4, 2002 (Bouwman, 2004), it has

423 424 Journal of Andrology N May June 2007 applied for exemption as far as the use of DDT [1,1,1- endocrine-disruptive environmental pollutants (Turusov trichloro-2,2-bis(chlorodiphenyl)ethane] for malaria et al, 2002). Serum levels of p,p9-DDE are an integrated vector control is concerned. measure of internal dose, reflecting exposure from all POPs are organic compounds that, to a varying sources over the previous years (Hauser et al, 2003). degree, resist photolytic, biological, and chemical Reproductive disorders were among the first adverse degradation. These compounds are often halogenated effects linked to organochlorine exposure (Beard, 2005). and characterized by low water solubility and high lipid Reproductive abnormalities attributed to DDE expo- solubility. They are also semivolatile, enabling them to sure after a major pesticide spill in 1980 were found in move long distances in the atmosphere before deposition reptiles inhabiting Lake Apopka in Florida in the 1990s. occurs. POPs, which are noted for their persistence and The types of deformities found were ambiguous gonads bioaccumulative characteristics, include DDT, dieldrin, (ovotestes) in turtles and abnormal sex hormone levels, toxaphene, chlordane, and several industrial chemical poorly organized testes, and small penises in male products and byproducts, including polychlorinated alligators (Guillette et al, 1995). The Great Lakes fish biphenyls (PCBs), dioxins, and furans (UNEP, 2006). found to be contaminated with organochlorine com- DDT and similar stable chlorinated compounds can pounds such as PCBs, dioxins, DDT metabolites, and be transported via air, rivers (Rawn et al, 1999; Buehler dibenzofuran have exhibited reproductive and other et al, 2004), and ocean currents (Bidleman et al, 1995) endocrine abnormalities. Wildlife (birds, turtles, and over long distances and have been detected in the mammals) that have consumed these fish have also Antarctic and other areas, far from their production exhibited various abnormalities, which include impaired sites or regions of use (Bouwman, 2004). While DDT is reproduction, same-sex pairing, feminization, ambiguous targeted by the treaty, exemptions are available for genitalia, and reduced fertility (Colborn et al, 1993, 1996; countries that are still using DDT to combat malaria. Fry, 1995). In mice, the uterotrophic effect of DDT The treaty mobilizes much needed funding to help increased the weight of the uterus and the development of countries shift to safer alternatives for malaria control, a pseudouterus (Morozova et al, 1997). A permanent, which has drawn attention and resources to the ongoing functional male-to-female sex reversal following a single and long-ignored tragedy of malaria, particularly in exposure of eggs to o,p9-DDT was observed in medaka Africa. A small group of United States conservatives fish (Edmunds et al, 2000; Turosov et al, 2002). There is continues to push to re-establish DDT as a ‘‘safe’’ also evidence that DDT acts as a promoter of mammary chemical for use against malaria, despite a clear decision tumors in rats and that it can inhibit gap junctional by the international community that DDT should be intercellular communication (Snedeker, 2001). Other targeted for ultimate elimination (Pesticide Action evidence of hormone-disrupting effects of DDT and its Network North America). All these factors have made metabolites has included reproductive defects and egg- the continued use of DDT for malaria vector control in shell thinning in avian species (Fry, 1995; Snedeker, South Africa, Africa, and all the other countries that 2001). DDT or p,p9-DDE might alter sex hormone have applied for exemption a matter of global interest. metabolism, reducing available testosterone to tissues A number of reports have indicated that, in addition (Guillette et al, 1995). to being toxic, organochlorine pesticides, including In humans, a trend in decreasing human sperm count DDT and its metabolites, might act as endocrine might have occurred in several European regions during disruptors (Turusov et al, 2002). Endocrine-disrupting the last 50 years (Irvine, 1994; Auger et al, 1995; Toft et chemicals can be defined as compounds that influence al, 2004). The decrease in sperm count is paralleled by normal hormone functions, generally causing adverse a rise in the trend of testicular cancer and malformations effects (Godduhn and Duffy, 2003). Technical-grade of the male reproductive organs such as hypospadias DDT is a mixture of p,p9-DDT (,85%), o,p9-DDT and cryptorchidism (Toppari et al, 1996). Skakkebaek et (,15%), and o,o9-DDT (trace amounts), with both p,p9- al (2001) presented a hypothesis that poor semen DDT and o,p9-DDT having estrogenic activity. p,p9- quality, testicular cancer, cryptorchidism, and hypospa- Dichlorodiphenyl-dichloroethylene (p,p9-DDE), a persis- dias are all indicative of 1 underlying entity—testicular tent metabolite of p,p9-DDT, is a widespread environ- dysgenesis syndrome (TDS)—with an origin in fetal life. mental contaminant (Turusov et al, 2002). The p,p9- The cause of TDS is unclear, but owing to the rapid DDE isomer is antiandrogenic by inhibitive binding to temporal changes in symptoms over the last few androgen receptors (Rogan and Chen, 2005) and has decades, it is suspected to be at least partly linked to been shown to inhibit the action of testosterone (Kelce environmental and lifestyle factors. In addition, genetic et al, 1995; Danzo, 1997; Bhatia et al, 2005). The polymorphisms or aberrations might render some hypothesis has been advanced that p,p9-DDE interacts individuals particularly susceptible to potential environ- in an additive or multiplicative way with other mental disrupters (Bay et al, 2006). Because technical- Aneck-Hahn et al N DDT Exposure and Seminal Parameters in South Africa 425 grade DDT comprises estrogenic molecules and because research protocol July 11, 2002. An initial visit to the proposed its major metabolite is a potent antiandrogen, it has study area took place in October 2003. The project team been hypothesized that exposure to DDT is involved in approached the village Chiefs and Elders for permission to the increase in male reproductive tract anomalies address the community about the proposed study. Meetings (Guillette et al, 1995; de Jager et al, 2006). were held at all 3 villages to inform the residents about the study and the procedures that would be followed. A In Africa, indoor residual spraying of DDT has representative was selected from each village (Dididi, Tshiu- become part of the national Roll Back Malaria strategic lungoma, and Tshikhudini) to assist with the recruitment of plan in several countries (Hougard et al, 2002; Rogan and participants. The representative was also trained to assist with Chen, 2005). In South Africa, DDT is sprayed in the low- the study questionnaire. After being properly informed, any altitude parts of Limpopo Province, Mpumalanga man who volunteered to participate and met the inclusion Province, and KwaZulu Natal. Currently, of the approx- criteria was included in the study. imately 40 million people in South Africa, 10%,or The Tshilidzini Hospital near Thoyohandou was used as 4 million, live in a malaria risk area (Rogan and Chen, a central laboratory point. Samples were collected between 2005). This puts the inhabitants of the rural communities November 2003 and July 2005. The participants produced in these areas at risk of being exposed to high semen samples in specially provided rooms adjacent to the concentrations of DDT and DDE. The exposure occurs onsite laboratory. In addition to the semen samples, blood through inhalation (indoor air spraying of dwellings and samples were collected and all participants signed an informed consent form and completed a questionnaire. During this outdoors), dermal contact (soil and house dust), and period, 362 participants were recruited, of which 51 were ingestion of contaminated foods and water. In South unable to produce a semen sample or did not meet the Africa, information on the health effects of environmen- inclusion criteria. tal DDT exposure is not available. In light of the above discussion, this study aims to assess the effects of Questionnaire nonoccupational exposure to DDT and semen parame- The questionnaire included questions on general health ters in young healthy men in a rural area in the Limpopo history, DDT exposure source (whether houses were sprayed Province, South Africa, where DDT is still sprayed. with DDT for malaria control or not), diet, fertility history, and other potential spermatotoxic exposures. Exposures studied included physical agents (exposure to heat or radiation Materials and Methods and history of testicular trauma), biological agents (genitouri- nary tract infections, history of STDs, orchitis, and epididi- Study Design and Population mitis), and chemical agents (exposure to recreational and In a cross-sectional study design, the participants were occupational drugs, pollutants, other pesticides, or any other volunteer, nonoccupationally exposed Venda men. The par- chemical agent, as well as smoking and drinking habits). ticipants recruited were between 18 and 40 years old and had been living in the communities for at least a year. Participants Exposure Assessment were excluded if they presented with a history of testicular Blood samples were collected from each participant. The trauma, orchitis, urinary infection, sexually transmitted samples were centrifuged at 670 6 g for 10 minutes at room diseases, use of hormonal medication, or exposure to known temperature. Plasma was stored at 220uC on site and then gonadotoxins or had neuropsychiatric disorders. transferred to a 270uC freezer until analyzed. The Agricultural Research Council, Veterinary Institute, Residue Laboratory in Study Area Pretoria, South Africa, determined DDT and its metabolites The Limpopo Province is situated in the northeastern corner with the use of a Shimatzu GCMS-QP2010 (Shimatzu, Tokyo of South Africa and is divided into 6 districts, with the study Japan). Concentrations of DDT compounds in the plasma area lying within the Vhembe district. After consultation with were expressed on a lipid-adjusted basis (mg/g). The detection the regional Department of Health and Social Development, 3 limit for p,p9-DDT and p,p9-DDE was 0.02 mg/g lipid adjusted. rural communities, Dididi, Tshiulungoma, and Tshikhudini, Total cholesterol and triglycerides were determined by near Thoyohandou were selected from a malaria endemic area. enzymatic methods and the total plasma lipid concentration The housing in these communities consists of traditional mud was calculated according to the formula proposed by Rylander dwellings with thatch (grass) roofs or brick and cement houses. et al (2006). DDT is sprayed inside unpainted brick, cement, and mud houses annually, but not inside the painted houses. Semen Analyses Semen samples were obtained from 311 participants after the Recruitment and Sampling prescribed 3-day period of sexual abstinence. Semen specimens The Ethics Committee of the Faculty of Health Sciences, were produced by masturbation directly into a sterile wide- University of Pretoria (Reference 43/2003) and the Limpopo mouthed container. The semen sample was then incubated at Provincial Government’s Department of Health approved the 37uC until liquefied. Trained researchers performed semen 426 Journal of Andrology N May June 2007 analyses and additional andrological tests according to the binomial regression analysis. Information obtained from the standards and procedures of the World Health Organization questionnaire, as well as the participant’s p,p9-DDT and p,p9- (WHO, 1999), and quality control (QC) procedures were DDE levels, was compared between different categories by adhered to (European Society for Human Reproduction and analysis of variance or regression analyses. Bivariate analyses Embryology [ESHRE], 1998; WHO, 1999). with regression models were conducted between the different After liquefaction, the following seminal physical character- reproductive outcomes and questionnaire variables to de- istics were assessed: appearance, liquefaction, viscosity, ejac- termine the risk factors and to identify confounding factors. ulate volume, and semen pH (Mortimer, 1994). Sperm Multivariate models were examined to evaluate the effect of concentration was determined with a hemocytometer (WHO, DDE/DDT in the different reproductive outcomes. A saturat- 1999). Sperm motility was assessed manually on a wet ed multivariate model was produced for each dependent preparation according to the WHO (1999) motility classifica- variable (semen parameter), including all independent vari- tion. This classification uses the class a through d sperm ables with P # .15 in the bivariate analyses. A manual stepwise progression rating (where a indicates rapid progressive motility elimination was used until every variable in the multivariable and d, immotile sperm; Nordic Association of Andrology, model had P # .05 or was capable of altering the other European Society of Human Reproduction and Embryology coefficients by at least 10%. The p,p9-DDT and p,p9-DDE — Special Interest Group on Andrology, 2002; WHO, 1999). plasma levels were used as continuous as well as categorical The viable sperm were assessed by the eosin-nigrosin method variables in multivariate analysis. All final regression models (Mortimer, 1994). The presence of leukocytes, erythrocytes, were adjusted by age. Final model sensitivity to individual bacteria, and agglutinates was also noted. The presence of observations was done by plotting the residuals vs fitted immunoglobulin on the sperm surface was assessed with the values, leverage vs normalized residuals squared, and residuals IgG test (SperMar test) on all fresh samples with motility vs predicted values. The dfbetas statistic was also estimated. ,40%. Immunological infertility can be considered when 50% Models were tested without detected influential observations pffiffiffi or more of the motile sperm have IgG antibodies. Sperm with dfbetas w 2 n. morphology slides were stained by the Papanicolau method and scored according to the WHO (1999) classification. The morphology assessment was performed by the same technol- ogist in the Andrology laboratory at the University of Results Pretoria. This laboratory ensures that the technologists follow strict quality control (QC) and quality assurance (QA) The mean age of the participants was 23 6 4.7 years procedures. The Andrology laboratory also takes part in an (mean 6 SD). Participants were Venda men from a rural international external QC program with the European Society area and a low socioeconomic status who had never for Human Reproduction and Embryology, and all observa- been occupationally exposed to the pesticide DDT. A tions fall within 61 SD of the reference results. selection bias affecting results is not probable because participants were not aware of the study hypothesis. Computer-Assisted Sperm Analysis Because the study design controlled for sexual absti- Sperm motility was further evaluated with a Hamilton Thorne nence time and this variable was not significant, there sperm motion analyzer (HTM-IVOS, Version 12; Beverly, was no need to control for it in analysis. Other explored Mass) at 60 Hz. Twelve microliters of semen were placed into exposures did not prove to be sufficiently present or Leja slides (Leja, SC 20-01-C; Calicom Trading [PTY] Ltd, intense to cause sperm alterations. Johannesburg, South Africa) with a chamber depth of 20 mm. Thirty frames were captured for analysis; a minimum of 150 The mean serum concentration of p,p9-DDT was sperm were analyzed in duplicate at 37uC (Schrader et al, 1992; 529.67 6 617.7 mg/L, and the mean lipid-adjusted p,p9- Mortimer and Fraser, 1996). Samples having an estimated DDT concentration was 90.23 6 102.4 mg/g (Table 1). count of more than 40 6 106 spermatozoa/mL were diluted The p,p9-DDE level had a mean serum concentration of with cell-free seminal plasma from the same individual. The 1259.10 6 1297.0 mg/L. When expressed as a lipid- percentages of motile sperm, progressive motility, linear adjusted concentration, the mean p,p9-DDE concentra- velocity, and curvilinear velocity were measured. tion was 215.47 6 210.6 mg/g (Table 1). The source of p,p9-DDT or p,p9-DDE exposure was found to be Statistical Analyses statistically significantly higher between participants Exploratory data analysis was conducted on the final database whose houses were sprayed with DDT (n 5 249) (i.e., to detect missing or outlier values. Tabulation and graphical mud and thatch roof dwellings) when compared with univariate analysis was done to describe the distribution of those whose houses were not sprayed (n 5 48; p,p9-DDE each variable and identify the necessary transformation to P 5 .000; p,p9-DDT P 5 .000; Table 1). normalize variables. In each case, after exploring several transformations and the raw form, the set closer to normal The distribution of the semen parameters and their distribution was used in linear regression analysis. The age-adjusted regression associations with the serum lipid distribution of variables describing sperm morphology—head, p,p9-DDT and p,p9-DDE are shown in Table 2. Di- midpiece, and tail defects percentage—required negative agnostic tests on final regression models showed an Aneck-Hahn et al N DDT Exposure and Seminal Parameters in South Africa 427

Table 1. Exposure data indicating the p,p9-dichlorodiphenyl-dichloroethylene (p,p9-DDE) and p,p9-1,1,1-trichloro-2,2- bis(chlorodiphenyl)ethane (p,p9-DDT) serum levels (n 5 303) Mean Houses Sprayed (6SD)3

Metabolite Mean (6SD) Median Minimum* Maximum No (n 5 48) Yes (n 5 249) p,p9-DDE (mg/L) 1259.10 (1297.0) 697.0 ND 6621.0 529.7 (658) 1409.8 (1339) p,p9-DDE (mg/g), lipid adjusted 215.47 (210.6) 134.0 ND 997.0 99.5 (123) 239.0 (215)4 p,p9-DDT (mg/L) 529.67 (617.7) 249.0 ND 2644.0 167.0 (339) 602.4 (630) p,p9-DDT (mg/g), lipid-adjusted 90.23 (102.4) 46.0 ND 519.0 30.5 (58) 101.9 (104)4 * ND indicates nondetectable (detection limit 5 0.02 mg/g). 3 Yes indicates that a participant’s house was sprayed with DDT within the last year. Six participants did not know. 4 P 5 .000. influential observation for one subject that, by itself, significant associations. Volume and count P values consistently and considerably altered the statistical with and without the subject in the final models changed significance, but not the coefficients. Although p,p9- from .05 to .02 and from .1 to .04, respectively. DDT and p,p9-DDE and semen results were measured Parameters showing a significant positive association correctly in this case, the subject was excluded because with continuous p,p9-DDT levels were the round cells he was not representative of the studied population. (beta 5 0.0013, P 5 .000) and the cytoplasmic droplets Excluding him only changed previously borderline (beta 5 0.0014, P 5 .014). The significant negative

Table 2. Distribution of seminal parameters and age-adjusted regression associations* with serum lipid p,p9-1,1,1-trichloro- 2,2-bis(chlorodiphenyl)ethane (p,p9-DDT) and p,p9-dichlorodiphenyl-dichloroethylene (p,p9-DDE) p,p9-DDE p,p9-DDT

Parameter n Mean (6SD) Median Beta 95% CI Beta 95% CI Semen Volume (mL)3 303 1.88 (1.3) 1.5 20.0003 20.0006, 20.00004 20.0005 20.001, 0.00004 Total count (mL/ejaculate)3 295 101.6 (159.3) 59 20.003 20.006, 20.0002 20.001 20.007, 0.005 Sperm concentration (106/mL)3 296 51.76 (48.2) 39 20.0003 20.0020, 0.0014 0.0022 20.0014, 0.0057 pH4 300 7.46 (0.3) 7.5 0.0033 20.0226, 0.0291 0.0195 20.0337, 0.0727 Progressive motility (%) (sum of grades a + b)§ 298 48.14 (21.1) 55 20.2807 21.099, 0.5379 20.8514 22.533, 0.8302 Motility (%) (sum of grades a + b + c)4 299 50.1 (15.8) 57 21.56 263.90, 60.74 227.63 2155.8, 100.5 Immotility (%) (grade d)3 299 49.26 (20.7) 43 0.0006 20.0002, 0.0013 0.0013 20.0003, 0.0028 Viability (%)§ 267 54.13 (21.8) 59 20.6571 21.756, 0.4417 21.7258 23.993, 0.5415 Normal morphology (%) 282 4.13 (2.70) 4 0.00006 20.0003, 0.0004 0.0002 20.0006, 0.0009 Head defects (%) 282 95.13 (3.3) 96 20.00009 20.00007, 0.00005 20.00002 20.0001, 0.0001 Neck/midpiece defects (%) 282 15.01 (5.4) 15 20.00009 20.0003, 0.0001 20.0004 20.0008, 0.00004 Tail defects (%) 282 12.92 (7.7) 11 20.0002 20.0005, 0.0001 20.0007 20.001, 20.00004 Round cells (106/mL)3 291 1.13 (1.3) 1 0.0005 0.0002, 0.0008 0.0013 0.0007, 0.0019 Cytoplasmic droplets (%)3 282 11.47 (6.5) 10 0.0005 20.0008, 0.0010 0.0013 0.0002, 0.0024 5 Teratozoospermic index 282 1.39 (0.1) 1.4 8.94 6 1026 20.00003, 0.00005 0.00003 20.00005, 0.00014 CASA Average path velocity (VAP, mm/s)" 241 36.51 (15.2) 36.0 20.0065 20.0154, 0.0024 20.0166 20.0354, 0.0022 Straight-line velocity (VSL, mm/s)3 241 26.98 (13.0) 26.5 20.0006 20.0014, 0.0001 20.0018 20.0034, 20.0002 Amplitude of lateral head displacement (ALH)3 241 2.54 (0.8) 2.5 20.0001 20.0003, 0.00003 20.0003 20.0007, 20.00003 Beat cross-frequency (BCF)" 239 28.68 (6.3) 29.0 0.0064 0.0028, 0.0100 0.0138 0.0062, 0.0213 Mean motility (%)" 240 48.53 (18.6) 52.0 20.0175 20.0283, 20.0068 20.0494 20.0720, 20.0268 * Dependent variables were transformed when required to normalize their distribution for linear regression analysis. Negative binomial regression was used for morphology parameters because of their distribution. CI indicates confidence interval. 3 Square root transformation. 4 Cubed transformation. § Squared transformation. 5 Reciprocal transformation. " No transformation (raw). 428 Journal of Andrology N May June 2007

Table 3. Oligozoospermia, asthenozoospermia, and teratozoospermia distribution and their age-adjusted association to lipid- adjusted p,p9-dichlorodiphenyl-dichloroethylene (p,p9-DDE ) and p,p9-1,1,1-trichloro-2,2-bis(chlorodiphenyl)ethane (p,p9-DDT )

Classification (%) 95% Confidence Parameter n No Yes Odds Ratio P Interval Oligozoospermia (,20 million sperm/mL) 295 72 28 DDE, 1.001 .03 (1.0001, 1.0025) Asthenozoospermia (,50% grades a + b motility) 285 68 32 DDT, 1.003 .006 (1.0007, 1.0055) Teratozoospermia (,15% normal morphology) 291 0.5 99.5 * * * * Not computable because of variable’s distribution. associations with continuous p,p9-DDT levels were sprayed houses than in men from nonsprayed houses volume (square root–transformed, beta 520.0003, P (102.0 mg/g p,p9-DDT and 239.0 mg/g p,p9-DDE vs 5 .024) and count (square root–transformed, beta 5 31.0 mg/g p,p9-DDT and 100.0 mg/g p,p9-DDE, P 5 20.003; P 5 .04). Semen volume and sperm count (both .0000) highlighted an important route of exposure, square root–transformed) were also significantly re- which could be an indication that indoor residual duced when the first 3 quartiles of lipid-adjusted p,p9- spraying contributes to increased exposure to DDT. DDE (0–345 mg/g) were compared with the fourth The mean lipid-adjusted p,p9-DDT levels (90.23 6 quartile (346–997 mg/g): volume beta 520.16, P 5 102.4 mg/g) and p,p9-DDE levels (215.47 6 210.6 mg/g) .01; sperm count beta 521.16, P 5 .04. The CASA in the nonoccupationally exposed population of this parameters that had a statistically significant negative study can be considered very high when compared with association with p,p9-DDT were straight-line velocity another study in the same province. Dalvie et al (VSL; beta 520.002, P 5 .03), amplitude of lateral (2004a,c) assessed the reproductive effects of long-term head displacement (ALH; beta 520.0003, P 5 .03), DDT exposure in malaria vector-control workers (n 5 and beat cross-frequency (BCF; 20.01, P 5 .000). The 47) in Limpopo Province. The mean lipid-adjusted p,p9- CASA mean motility (cubed) had a significant negative DDT levels in that study were 26.1 6 13.7 mg/g and the association with p,p9-DDE (beta 520.02, P 5 .001; mean p,p9-DDE levels were 65.0 6 48.8 mg/g. Those Table 2). In comparing the first 3 p,p9-DDE quartiles levels are almost 3.5 times lower than those found in this against the fourth, the coefficient for cubed motility was study, which is to be expected because the malaria found to be 28.79 (P 5 .001). vector-control workers wear protective clothing and The participants’ semen characteristics were classified take the necessary safety precautions when working with according to dichotomous abnormal semen categories DDT. In a similar nonoccupational exposure study done (WHO, 1999), and these were expressed as percentages (Table 3). Of all the participants, 28% were classified with oligozoospermia, 99.5% with teratozoospermia, and 32% with asthenozoospermia. The distribution and crude regression of the dichotomous abnormal semen categories indicated by the odds ratio (OR) showed that those participants with oligozoospermia were significantly associated with p,p9-DDE (OR 1.001, P 5 .03; Table 3). The distribution of oligozoospermia was statistically significantly associated with the lipid-adjusted p,p9-DDE percentile concentrations, as is shown in the Figure. The distribution of asthenozoospermia was also significantly associated with the lipid-adjusted p,p9-DDT percentile concentrations (OR 1.003, P 5 .006) and p,p9-DDE (OR 1.001, P 5 .02). The graph shows that the higher the p,p9- DDE concentration, the greater the incidence of oligozoospermia and asthenozoospermia (Figure).

The distribution of oligozoospermia and asthenozoospermia associated with dose-dependent, lipid-adjusted p,p9-DDE percentile con- Discussion centrations (1 indicates the 25th percentile; 2, the 50th percentile; 3, the 75th percentile; and 4, the 100th percentile). The graph shows that the higher the p,p9-DDE concentration, the greater the incidence The finding that both p,p9-DDT and p,p9-DDE values of oligozoospermia (OR 1.001, P 5 .03) and asthenozoospermia were statistically significantly higher in men living in (OR 1.001, P 5 .02). Aneck-Hahn et al N DDT Exposure and Seminal Parameters in South Africa 429 in 2000–01 by de Jager et al (2006) in Chiapas, Mexico, DDE was analyzed by quartiles, results showed that the the p,p9-DDE level (45 6 32 mg/g) was found to be expected semen volume of the most exposed men (p,p9- almost 5 times lower than in this study; however, this DDE 5 346–997 mg/g) would be approximately was after DDT had been phased out by 2000. It is 1.38 mL. Pant et al (2004) showed that p,p9-DDE and known that p,p9-DDE concentration in lipids is used as p,p9-DDD (p,p9-DDD 5 1,1 dichloro-2,2 bis (p-chlor- a surrogate for chronic exposure to technical DDT, ophenyl)ethane) were higher in the semen of infertile a mixture that comprises estrogenic compounds such as men compared with fertile men. The levels of gamma- o,p9-DDT and p,p9-DDT and the androgen antagonist glutamyl transpeptidase and acid phosphatase activity p,p9-DDE (de Jager et al, 2006). The high p,p9-DDT were also lower in infertile men, whereas the high level in this study indicates current acute exposure to fructose level observed could suggest ‘‘non-utilization of DDT and the high p,p9-DDE levels indicate chronic the enzyme by sperms due to some biochemical defects’’ long-term exposure. These levels are much higher than (Pant et al, 2004:213), although this was not studied. It the levels in the above-mentioned studies that found would have been in the interest of the study to assess reproductive effects because of DDT exposure (Ayotte the accessory gland markers such as fructose and a- et al, 2001; Dalvie et al, 2004a,c; de Jager et al, 2006). glucosidase, but unfortunately, because of the low The levels in this study are supported by lower semen volume, this could not be done. volume, total sperm count, progressive motility, and Although the mean total sperm count was within the viability with higher levels of p,p9-DDT. Higher levels of WHO (1999) reference range, there was a significant p,p9-DDE also resulted in lower semen volume, total negative association with p,p9-DDT and p,p9-DDE. sperm count, progressive motility, and viability. Analyzing p,p9-DDE by quartiles showed that the To assess reproductive function, a basic semen expected sperm count in the highest quartile (p,p9- analysis was carried out and CASA motility parameters DDE 5 346–997 mg/g) would be approximately were evaluated as well. This included semen volume, 56.3 million. The participants were divided into oligo- pH, and viscosity. The epididymal epithelium is and normozoospermic categories according to WHO androgen dependent and has both absorptive and (1999). The crude regression associations showed secretory functions. The epididymal plasma in which a statistically significant positive dose-dependent asso- the sperm are suspended within the epididymis is also ciation between participants with oligozoospermia secreted by the epididymal epithelium. It is a complex (28%) and p,p9-DDE concentration (OR 1.001, P 5 fluid that changes along the length of the epididymis. .03). This indicates that participants with high concen- The spermatozoa experience a series of sophisticated trations of p,p9-DDE are at risk of presenting with microenvironments that regulate their maturation oligozoospermia. In support, the distribution of oligo- (Mortimer, 1994). Under physiological conditions, the zoospermia shows that the higher the p,p9-DDE various components of the ejaculate originate in concentration, the greater the incidence of oligozoos- different parts of the male reproductive tract and are permia. This trend is similar to the findings of a study by emitted in a definite order (Mann and Lutwak-Mann, Rozati et al (2002), who suggested that there was 1982). The vesicular fluid is the last fraction of semen a significant deterioration in semen parameters, in- ejaculated and contributes up to 70% of the ejaculate cluding sperm count in infertile men with PCBs in their volume, whereas the prostate contributes the other seminal plasma, compared with the control group. This 30% (Mortimer, 1994). It has been hypothesized that trend is also in agreement with a report citing an inverse toxicants or their metabolites can act directly on correlation of PCBs and sperm motility in men with accessory glands by altering the quality or quantity of oligozoospermia (Bush et al, 1986). their secretions and that this could influence semen There were negative associations between the pro- volume (Mann and Lutwak-Mann, 1982; Pant et al, gressive motility, total motility, and viability and the p,p9- 2004). Because both the prostate and seminal vesicles DDT and p,p9-DDE concentrations. Some animal data are also androgen-dependent organs (Mann and Lut- suggest that p,p9-DDE might be hormonally active and wak-Mann, 1982), the antiandrogen properties of p,p9- therefore would adversely affect semen parameters. The DDE could have an influence on the functions of the compounds that readily pass the blood-testis barrier organs. This could account for the mean semen volume might directly affect spermatogenesis (Hauser et al, (1.88 6 1.3 mL), which was slightly lower than the 2003). Effects at the mitotic or meiotic level could lead WHO (1999) reference value and similar to the Chiapas to decreased sperm production, whereas the targeting of study (mean volume 5 1.84 mL; de Jager et al, 2006). the postmeiotic processes and epididymal sperm matura- The study showed a strong and significant negative tion might lead to impaired sperm motility (Hauser et al, association between p,p9-DDE and volume; a similar 2003). Despite a mean motility within the WHO normal association was found in the Chiapas study. When p,p9- range (WHO, 1999), 32% (n 5 285) of the participants 430 Journal of Andrology N May June 2007 presented with asthenozoospermia (.50% a + bor.25% This study population had a high percentage of round grade a). Although there was no significant association cells (1.13 6 1.3 6 106/mL), which showed a statistically between sperm motility and p,p9-DDE, the distribution of significant positive association with both p,p9-DDT asthenozoospermia shows that the higher the p,p9-DDE (beta 5 0.0013, P 5 .000) and p,p9-DDE (beta 5 concentration the greater the incidence of asthenozoos- 0.0005, P 5 .000). Indications are that round cells occur permia. The multivariate logistic regressions OR showed frequently in infertile patients and are associated with a statistically significant positive dose-dependent associ- poor semen quality (Arata de Bellabarba et al, 2000). A ation between participants with asthenozoospermia and study carried out in Austria showed a significant p,p9-DDT concentrations (OR 1.003, P 5 .006) and p,p9- increase in round cells in the semen of smokers DDE concentrations (OR 1.001, P 5 .02). This result was compared with nonsmokers (Trummer et al, 2002). similar to a finding in a study by Hauser et al (2003), In this study, smoking was taken into account as which showed that PCB-138, which is also an organo- a possible confounder and did not affect the outcome of chlorine compound, was inversely associated with sperm the data. motility and morphology. During spermatogenesis, spermatids are transformed The mean CASA motility (48.53% 6 18.6%) com- into sperm by different processes, including condensa- pared well with the manual mean motility (50.1% 6 tion and structural shaping of the cell nucleus and the 15.8%). CASA is being used in reproductive toxicology formation of the flagellum. Disruption at this stage of because some of the motility parameters are sensitive to development can cause impairment of sperm condensa- toxins (ESHRE, 1998). The significant association with tion, motility, and morphology (Parvinen, 1998; de p,p9-DDE and the cubed CASA mean motility (beta 5 Jager et al, 2006). A significant proportion of the sperm 20.02, P 5 .001) in this study compares well to a study in each sample might be morphologically abnormal, but by Hirano et al (2001), which compared the CASA if the proportion is above 5%, it could account for parameters of ‘‘good’’ (fertilization rate . 50%) and impaired fertility (Menkveld et al, 1990). WHO (1999) ‘‘poor’’ (fertilization rate # 50%) fertilization groups. states that below 15% normal forms, the fertilization The poor fertilization group was found to be 48.9% 6 rates in vitro will be reduced. The study investigating 22.1% as opposed to the good fertilization group (59.9% exposure to DDT in malaria vector-control workers had 6 16.5%; Hirano et al, 2001). This indicates that the a mean normal morphology score of 2.5% 6 1.8%,with fecundity of this exposed Limpopo population might be 84% of the morphology scores being below the WHO compromised. CASA parameters showing a significant (1992) and Tygerberg strict criteria. The Tygerberg strict negative association with p,p9-DDT were the VSL (beta criteria could be said to be inappropriately strict for 520.002, P 5 .03), ALH (beta 520.0003, P 5 .03), epidemiological settings in which the aim is to detect and BCF (beta 520.01, P 5 .000). The VSL (26.98 mm/ more subtle effects (Dalvie et al, 2004b). A significantly s) and ALH (2.54 mm) values in this study are similar to high proportion of the participants in this study those in a study by Guo et al (2000), which found that presented with teratozoospermia (99.5%), and the mean the VSL (25.4 mm/s) and ALH (2.9 mm) were lower in percent normal morphology was 4.13% 6 2.70%, which a PCB-exposed population than in an unexposed is well below the WHO (1999) reference range of 15% population (VSL 5 33.0 mm/s, ALH 5 3.3 mm). The normal forms. Cytotoxic effects, such as the production BCF is useful in determining changes in the flagellar of superoxide anion and activation of various intracel- beat pattern (Mortimer, 2000). The negative association lular signal transduction pathways, might explain the with p,p9-DDT indicates that higher levels of DDT significant decrease in normal morphology (Rozati et al, cause an increase in the flagellar beat pattern with an 2002). adverse effect on sperm motility. In the study by Guo et During spermiation, residual cytoplasm is shed from al (2000), the BCF was lower (17.4 Hz) for participants the neck of the mature spermatid and a small residual prenatally exposed to PCBs and dibenzofurans but was cytoplasmic droplet remains attached to the testicular comparable to this study (BCF 5 29 Hz). These findings sperm, which is lost through epididymal transit and were similar to those in animals, in which in utero sperm maturation (Hess et al, 2001). The mean exposure to similar toxic levels of these chemicals prevalence of cytoplasmic droplets is 2.2% (Belsey et reduced daily sperm production and increased the al, 1980; Mortimer, 1994), whereas in this study, it was percent abnormal sperm (Faqui et al, 1998; Guo et al, 11.5%. Fisher et al (1998) demonstrated that neonatal 2000). The findings of this study indicate that DDT exposure of rats to diethystilboestrol (DES) caused exposure could have a negative effect on sperm motility. permanent distention of the rete testis and efferent Sperm motility is commonly believed to be one of the ducts, with loss of epithelial height through adulthood. most important characteristics correlated with fertility Sharpe (1998) argues that, owing to the loss of the apical (Eimers et al, 1994; Hirano et al, 2001). portion of the cell, the endocytotic apparatus might be Aneck-Hahn et al N DDT Exposure and Seminal Parameters in South Africa 431 dysfunctional. This could imply that, similar to DES, Conclusions prenatal DDT (estrogenic) exposure might have affected the development of these epididymal cells and sub- The data on seminal parameters from 311 participants sequently could contribute to the high number of makes this study one of the largest to look at the effects cytoplasmic droplets. of DDT exposure. The study found evidence that Not only are cytoplasmic droplets associated with indicated that nonoccupational exposure to p,p9-DDT immature spermatozoa, their presence is correlated with and its metabolite p,p9-DDE has an effect on seminal oxidative damage (Mortimer, 1994; Gergely et al, 1999; parameters of young men living in the Limpopo Chantler and Abraham-Peskir, 2004). A study by Province. The mean age of the participants was Gomez et al (1996) correlated a specific morphological 23 years, which is young when compared with similar defect of human sperm with reactive oxygen species studies. This is of concern because it has been found that (ROS). The residual cytoplasm present in the midpiece semen volume, motility, and morphology decrease with of the human sperm revealed a significant correlation age (Kidd et al, 2001). It is impossible to know what the between excess residual cytoplasm in the midpiece and effects will be on the fertility potential of this population the enhanced generation of ROS. The study by Aziz et al in 5 to 10 years with the continued use of DDT for (2004) supported this finding and additionally showed malaria vector control. This data and the Mexico data a positive correlation of ROS with percent sperm with (de Jager et al, 2006), together with the well-documented cytoplasmic droplets (8%) and tail defects (12%). This effects of DDT on animals (Toppari et al, 1996), should study showed a statistically significant positive associ- provide sufficient evidence to elicit concern about the ation between cytoplasmic droplets (11%) and p,p9- effect of this pesticide and its metabolite p,p9-DDE on DDT (beta 5 0.0014, P 5 .014); although there was no human health. association with the tail defects (13%), these percentages were similar to those in the study by Aziz et al (2004). Long-term exposure to small amounts of organochlo- Oxidative stress or ROS generation can be induced in rine contaminants leads to the accumulation of consid- the testes by exposure to common xenobiotics, such as erable burdens in animal and human tissue (de Jager et nonylphenol and dioxin (Aitken et al, 2004; Chitra and al, 2006). The young are the most vulnerable. It is not Mathur, 2004). Subchronic exposure to DDT is necessarily the amount of DDT to which the mother is associated with an increase in free radical generation exposed during pregnancy that is critical but rather her by lipid peroxidation (Koner et al, 1998). It is well lifetime exposure and bioaccumulation that determines known that ROS generation impairs sperm motility the level of exposure of the fetus and breast-fed infant (Aitken et al, 1998). This means that there is a strong (Longnecker et al, 2000; Korrick et al, 2001). Many of possibility of increased ROS generation in this exposed the infants in these rural areas in South Africa are breast group of men, which will influence morphology and fed. Other effects could occur; in a study by Longnecker sperm motility parameters negatively. et al (2002), there was a modest to moderate association The etiology of TDS is suspected to be related to in boys with maternal levels of DDE greater than or genetic, environmental, or both factors, including equal to 85.6 mg/L with the development of cryptorchi- endocrine disruptors. Both p,p9-DDT and p,p9-DDE dism, hypospadias, and polythelia. Sunyer et al (2005) are considered to be hormonally active, with p,p9-DDT found that prenatal exposure to DDE residues (geo- having estrogenic activity via binding and activation of metric mean in cord serum: 1.06 mg/L) might contribute the estrogen receptor and p,p9-DDE being antiandro- to the development of asthma. Prenatal exposure to genic (Kelce et al, 1995). Irrespective of the exact DDT and, to a lesser extent, DDE was associated with mechanism, which remains to be elucidated in these neurodevelopmental delays during early childhood cases, either reduced testosterone production by Leydig (Eskenazi et al, 2006). cells (via DDT’s estrogenic suppression of the hypotha- Of concern in this nonoccupationally exposed lamic-pituitary-testicular axis) or by impeded androgen population is that these high levels of p,p9-DDT action (via DDE’s effect on the androgen receptor), and p,p9-DDE appear to have adverse effects on the the physiological consequence would be impaired seminal parameters, supporting the findings by de Jager Sertoli cell function (Parvinen, 1998). The primary role et al (2006). The elevated levels of exposure indicate of these cells is to support spermatogenesis (de Jager et a high degree of chronic exposure and imply an urgent al, 2006). The increased serum p,p9-DDT and p,p9-DDE need for continued epidemiologic studies in the Lim- levels could be exerting an effect on the Sertoli popo Province to determine the potential adverse effects cells, preventing normal spermatogenesis and result- of these pesticides. These findings are not only ing in abnormal sperm function as observed in this applicable to Limpopo Province, but also to other study. malaria areas in South Africa, Africa, and other parts of 432 Journal of Andrology N May June 2007 the world where DDT is used for malaria control. The Chantler E, Abraham-Peskir JV. Significance of midpiece vesicles and feasibility of cost-effective and environmentally safe functional integrity of the membranes of human spermatozoa after osmotic stress. Andrologia. 2004;36:87–93. alternative methods for pest control needs to be Chitra KC, Mathur PP. Vitamin E prevents nonylphenol-induced considered. oxidative stress in testis of rats. Indian J Exp Biol. 2004;42(2): 220–223. Colborn T, Dumanoski D, Myers JP. Our Stolen Future. New York: Acknowledgments Penguin; 1996. Colborn T, Vom Saal FS, Soto AM. Developmental effects of We thank Ms K Wannenburg (Andrology, Department of Urology, endocrine-disrupting chemicals in wildlife and humans. Environ University of Pretoria) and for the morphology assessments, and to Health Perspect. 1993;101:378–384. Ms C Van Zijl for technical assistance on the project. We thank Mr Dalvie MA, Myers JE, Thompson ML, Dyer S, Robins TG, Omar S, R Mudzielwana from the Malaria Control Programme of the Riebow J, Molekwe J, Kruger P, Millar R. The hormonal effects of Department of Health and Social Development, Limpopo Provincial long-term DDT exposure on malaria vector-control workers in Government, for his assistance in facilitating communication between Limpopo Province, South Africa. Environ Res. 2004a;96:9–19. the community leaders, field workers, and project team. Dalvie MA, Myers JE, Thompson ML, Robins TG, Dyer S, Riebow J, Molekwe J, Jeebhay M, Millar R, Kruger P. The long-term effects of DDT exposure on semen, fertility, and sexual function of References malaria vector-control workers in Limpopo Province, South Africa. Environ Res. 2004b;96:1–8. Aitken RJ, Gordon E, Harkiss D, Twig JP, Milne P, Jennings Z, Irvine Dalvie MA, Myers JE, Thompson ML, Robins TG, Omar S, Riebow DS. Relative impact of oxidative stress on the functional J. Exploration of different methods for measuring DDT exposure competence and genomic integrity of human spermatozoa. Biol among malaria vector-control workers in Limpopo Province, Reprod. 1998;59:1037–1046. South Africa. Environ Res. 2004c;96:20–27. Aitken RJ, Koopman P, Lewis SEM. Seeds of concern. Nature. Danzo BJ. Environmental xenobiotics may disrupt normal endocrine 2004;432:48–52. function by interfering with the binding of physiological ligands to Arata de Bellabarba G, Tortolero I, Villarroel V, Molina CZ, steroid receptors and binding proteins. Environ Health Perspect. Bellabarba C, Velazquez E. Nonsperm cells in human semen and 1997;105:294–301. their relationship with semen parameters. Arch Androl. 2000;45(3): de Jager C, Farias P, Barraza A, Avila MH, Ayotte P, Dewailly E, 131–136. Rousseau F, Sanchez VD, Bailey JL. Reduced seminal parameters Auger J, Kunstmann JM, Czysglik F, Jouannet P. Decline in semen associated with environmental DDT exposure and p,p9-DDE quality among fertile men in Paris during the past 20 years. concentrations in men in Chiapas, Mexico: A cross sectional study. N Engl J Med. 1995;332:281–285. J Androl. 2006;27(1):16–27. Ayotte P, Giroux S, Dewailly E´ , Hernandez Avilla M, Farias P, Danis Edmunds JSG, McCarthy RA, Ramsdeli JS. Permanent and func- R, Villanueva Diaz C. DDT spraying for malaria control and tional male-to-female sex reversal in d-rR strain Medaka (Oryzias reproductive function in Mexican men. Epidemiology. 2001;12: latipes) following egg microinjection of o,p9-DDT. Environ Health 366–367. Perspect. 2000;108:219–224. Aziz N, Saleh RA, Sharma RK, Lewis-Jones I, Esfandiari N, Thomas Eimers JM, te Velde ER, Gerritse R, Vogelzang ET, Looman CW, AJ Jr, Agarwal A. Novel association between sperm reactive Habbema JD. The prediction of the chance to conceive in subfertile oxygen species production, sperm morphological defects, and the couples. Fertil Steril. 1994;61:44–52. sperm deformity index. Fertil Steril. 2004;2(81):349–354. European Society of Human Reproduction and Embryology Special Bay K, Asklund C, Skakkebaek NE, Andersson AM. Testicular Interest Group on Andrology. Guidelines on the application of dysgenesis syndrome: possible role of endocrine disrupters. Best CASA technology in the analysis of spermatozoa. Hum Reprod. Pract Res Clin Endocrinol Metabol. 2006;20(1):77–90. 1998;13:142–145. Beard J. DDT and human health. Sci Tot Environ. 2005;355:78–89. Eskenazi B, Marks AR, Bradman A, Fenster L, Johnson C, Barr DB, Belsey MA, Eliasson R, Gallegos AJ, Moghissi KS, Paulsen CA, Jewell NP. In utero exposure to dichlorodiphenyltrichloroethane Prasad MRN. Laboratory Manual for the Examination of Human (DDT) and dichlorodiphenyldichloroethylene (DDE) and neuro- Semen and Semen–Cervical Mucus Interaction. Singapore: Press development among young Mexican American children. Pediatrics. Concern; 1980. 2006;118(1):233–241. Bhatia R, Shiau R, Petreas M, Weinraub JM, Farhang L, Eskenazi B. Faqui AS, Dalsenter PR, Merker HJ, Chahoud I. Reproductive Organochlorine pesticides and male genital anomalies in the child toxicity and tissue concentrations of low doses of 2,3,7,8- health and development studies. Environ Health Perspect. 2005; tetrachlorodibenzo-p-dioxin in male offspring rats exposed 113(2):220–224. throughout pregnancy and lactation. Toxicol Appl Pharmacol. Bidleman TF, Falconer RL, Walla MD. Toxaphene and other 1998;150:383–392. organochlorine compounds in the air and water at Resolute Bay, Fisher JS, Turner KJ, Fraser HM, Saunders PT, Brown D, Sharpe NWT, Canada. Sci Tot Environ. 1995;161:55–63. RM. Immunoexpression of aquaporin-1 in the efferent ducts of the Bouwman H. South Africa and the Stockholm Convention on rat and marmoset monkey during development, its modulation by persistent organic pollutants. S Afr J Sci. 2004;100:323–328. estrogens, and its possible role in fluid resorption. Endocrinology. Buehler SS, Basu I, Hites RA. Causes of variability in pesticide and 1998;139:3935–3045. PCB concentrations in air near the Great Lakes. Environ Sci Fry DM. Reproductive effects in birds exposed to pesticides and Technol. 2004;38:414–422. industrial chemicals. Environ Health Perspect. 1995;103(suppl. Bush B, Bennett A, Snow J. Polychlorinated biphenyl congeners p-p9- 7):165–171. DDE and sperm function in humans. Arch Environ Contam Gergely A, Kovanci E, Senturk L, Cosmi E, Huszar G. Morphometric Toxicol. 1986;15:333–341. assessment of mature and diminished maturity human spermato- Aneck-Hahn et al N DDT Exposure and Seminal Parameters in South Africa 433

zoa: sperm regions that reflect differences in maturity. Hum Mortimer D. Practical Laboratory Andrology. New York: Oxford Reprod. 1999;12:2007–2014. University Press; 1994; Godduhn A, Duffy LK. Multi-generation health risks of persistent Mortimer D, Fraser L. Consensus workshop on advanced diagnostic organic pollution in the far north: use of the precautionary andrology techniques. Hum Reprod. 1996;11:1463–1479. approach in the Stockholm Convention. Environ Sci Pol. Mortimer ST. CASA—practical aspects. J Androl. 2000;21(4):515– 2003;6:341–353. 523. Gomez E, Buckingham DW, Brindle J, Lanzafame F, Irvine DS, Nordic Association of Andrology, European Society of Human Aitken RJ. Development of an image analysis system to monitor Reproduction and Embryology–Special Interest Group on Androl- the retention of residual cytoplasm by human spermatozoa: ogy. Manual on Basic Semen Analysis. 2002. Available at: http:// correlation with biochemical markers of the cytoplasmic space, www.ki.se/org/nafa/manual/Manual2002.pdf. Accessed August oxidative stress, and sperm function. Fertil Steril. 1996;2(81): 2006. 349–354. Pant N, Mathur N, Banerjee AK, Srivastava SP, Saxena DK. Guillette LJ Jr, Gross TS, Gross DA, Rooney AA, Percival HF. Correlation of chlorinated pesticides concentration in semen with Gonadal steroidogenesis in vitro from juvenile alligators obtained seminal vesicle and prostatic markers. Repro Toxicol. 2004;19: from contaminated or control lakes. Environ Health Perspect. 209–214. 1995;103(suppl. 4):31–36. Parvinen M. Cyclic function of Sertoli cells. In: Russell LD, Griswold Guo YL, Hsu PC, Hsu CC, Labert GH. Semen quality after prenatal MD, eds. The Sertoli Cell. Clearwater, Fla: Cache River Press; exposure to polychlorinated biphenyls and dibenzofurans. Lancet. 1998:331–347. 2000;356:1240–1241. Pesticide Action Network North America [Internet]. Advancing Hauser R, Chen ZY, Pothier L, Ryan L. Altshul L. The relationship alternatives to pesticides worldwide. Newsroom. Pesticide Action between human semen parameters and environmental exposure to Network Update Service (PANUPS). Global Treaty Targets polychlorinated biphenyls and p,p9-DDE. Environ Health Perspect. Dangerous Pollutants [April 27, 2006; Accessed 2006 Sept 12]. 2003;111(12):1505–1511. Available from http://www.panna.org/resources/newsroom. Hess RA, Bunick D, Bahr J. Oestrogen, its receptors and function in Rawn DFK, Halldorson THJ, Woychuk RN, Muir DCG. Pesticides in the male reproductive tract—a review. Mol Cell Endocrinol. 2001; the Red River and its tributaries in southern Manitoba: 1993–1995. 178:29–38. Water Qual Res J Can. 1999;34:183–219. Hirano Y, Shibahara H, Obara H, Suzuki T, Takamizawa S, Rogan WJ, Chen A. Health risks and benefits of bis(4-chlorophenyl)- Yamaguchi C, Tsunoda H, Sato I. Relationships between sperm 1,1,1-trichloroethane (DDT). Lancet. 2005;366:763–773. motility characteristics assessed by the computer-aided sperm Rozati R, Reddy PP, Reddanna P, Mujtaba R. Role of environmental analysis (CASA) and fertilization rates in vitro. J Assist Reprod estrogens in the deterioration of male factor infertility. Fertil Steril. Genet. 2001;18(4):213–218. 2002;78(6):1187–1194. Hougard JM, Fontenille D, Chandre F, Darriet F, Carnevale P, Rylander L, Nilsson-Ehle P, Hagmar L. A simplified precise Guillet P. Combating malaria vectors in Africa: current directions method for adjusting serum levels of persistent organohalogen of research. Trends Parasitol. 2002;18:283–286. pollutants to total serum lipids. Chemosphere. 2006;62(3):333– Irvine DS. Falling sperm quality. BMJ. 1994;309:476. 336. Kelce WR, Stone CR, Laws SC, Gray LE, Kemppainen JA, Wilson Schrader SM, Chapin RE, Clegg ED, Davis RO, Fourcroy JL, Katz EM. Persistent DDT metabolite p,p9-DDE is a potent androgen DF, Rothman SA, Toth G, Turner TW, Zinaman M. Laboratory receptor antagonist. Nature. 1995;375:581–585. methods for assessing human semen in epidemiologic studies: Kidd SA, Eskenazi B, Wyrobek AJ. Effects of male age on semen a consensus report. Repro Toxicol. 1992;6:275–279. quality and fertility: a review of the literature. Fertil Steril. 2001; Sharpe RM. The roles of oestrogen in the male. Trends Endocrinol 75(2):237–248. Metab. 1998;9(9):371–377. Koner BC, Banerjee, Ray A. Organochlorine pesticide-induced Skakkebaek NE, Rajpert-De Meyts E, Main KM. Testicular oxidative stress and immune suppression in rats. Indian J Exp dysgenesis syndrome an increasingly common development Biol. 1998;36:395–398. disorder with environmental aspects. Hum Reprod. 2001;16: Korrick SA, Chen C, Damokosh AI, NI J, Liu X, Cho SI, Altshul L, 972–978. Ryan L, Xu X. Association of DDT with spontaneous abortion: Snedeker SM. Pesticides and breast cancer risk: a review of DDT, a case control study. Ann Epidemiol. 2001;11:491–496. DDE and dieldrin. Environ Health Perspect. 2001;109(suppl 1): Longnecker MP, Klebanoff MA, Brock JW, Zhou H. DDE is 35–47. associated with increased risk of preterm delivery and small-for- Sunyer J, Torrent M, Mun˜oz-Ortiz L, Ribas-Fito´ N, Carrizo D, gestational-age birthweight in humans. Organohalogen Compd. Grimalt J. Prenatal dichlorodiphenyldichloroethylene (DDE) and 2000;48:161–162. asthma in children. Environ Health Perspect. 2005;113(12): Longnecker MP, Klebanoff MA, Brock JW, Zhou H, Gray KA, 1787–1790. Needhan LL, Wilcox AJ. Maternal serum level of 11-dichloro-2,2- Toft G, Hagmar L, Giwercman A, Bonde JP. Epidemiological bis(p-chorophenyl)ethylene and risk of cryptorchidism, hypospa- evidence on reproductive effects of persistent organochlorines in dias and polythelia among male offspring. Am J Epidemiol. humans. Rep Toxicol. 2004;19:5–26. 2002;155(4):313–322. Toppari J, Larsen JC, Christiansen P, Giwercman A, Grandjean P, Mann T, Lutwak-Mann C. Passage of chemicals into human and Guillette LJ, Je´gou B, Jensen TK, Jouanette P, Keding N, Leffers animal semen: mechanisms and significance. Crit Rev Toxicol. H, McLachlan JA, Meyer O, Mu¨ller., Rajpert-De Meyts E, Scheike 1982;11:1–14. T, Sharpe R, Sumpter J, Skakkebaek NE. Male reproductive Menkveld R, Stander FSH, Kotze TJW, Kruger TF, Van Zyl JA. The health and environmental xenoestrogens. Environ Health Perspect. evaluation of morphological characteristics of human spermatozoa 1996;104(suppl 4):741–776. according to stricter criteria. Hum Reprod. 1990;5:586–592. Trummer H, Habermann H, Haas J, Pummer K. The impact of Morozova DV, Riboli E, Turusov VS. Estrogenic effect of DDT in cigarette smoking on human semen parameters and hormones. CBA female mice. Exp Toxicol Pathol. 1997;49:483–485. Hum Reprod. 2002;17(6):1554–1559. 434 Journal of Andrology N May June 2007

Turusov V, Rakitsky V, Tomatis L. Dichlorodiphenyltrichloroethane cals. Stockholm Convention.] Available at: http://www.chem. (DDT): ubiquity, persistence and risks. Res Rev. 2002;110(2): unep.ch. Accessed September 12, 2006. 125–128. World Health Organization. DDT and its Derivatives. Vol 9. Geneva: United Nations Environment Programme. UNEP decision 18/32 of World Health Organization. 1979. the UNEP Governing Council: persistent organic pollutants. May World Health Organization. WHO Laboratory Manual for the 25, 1995. Available at: http://www.pops.int/documents/back- Examination of Human Semen and Sperm–Cervical Mucus In- ground/gcdecision/18-32/gc1832en.html. Accessed September 12, teraction. 3rd ed. New York: Cambridge University Press; 1992. 2006. World Health Organization. WHO Laboratory Manual for the United Nations Environment Programme. UNEP chemicals home- Examination of Human Semen and Sperm–Cervical Mucus In- page. [Division of Technology, Industry, and Economics. Chemi- teraction. 4th ed. New York: Cambridge University Press; 1999.