Authors
M Vrijheid, B Armstrong, H Dolk, M van Tongeren, B Botting.

Title
Risk of hypospadias in relation to maternal occupational exposure to potential endocrine disrupting chemicals.

Source
Occupational and Environmental Medicine: 60(8):543-550, Aug 2003.

Summary
Hypospadias is a congenital anomaly characterized by an abnormal external urethral opening in males situated anywhere between the glans and the perineum along the lower side of the shaft of the penis. The prevalence of hypospadias and other related abnormalities such as cryptorchidism (undescended testes), testicular cancer and male infertility has been increasing in Europe and the US. Endocrine disrupting chemicals have been hypothesized to be associated with these conditions. Occupational groups such as farmers and gardeners have been of particular concern because of their exposure to pesticides which have endocrine disrupting properties. The results of previous studies however, have been inconclusive.

The present study, for the first time, focused on analyzing the relationship between maternal occupation, exposure to potential endocrine disrupting chemicals and risk of hypospadias in England and Wales. Data from the National Congenital Anomaly System (NCAS) between the years 1980 and 1996 was used. Occupations were coded using C080 and OC90 job classification systems in the 1980's and 1990's respectively. OC90 codes were translated back to C080 codes resulting in 348 job titles for the analysis. A job exposure matrix was developed for this study to classify each of the occupational codes as unlikely, possibly and probably exposed to endocrine disrupting chemicals (EDCs) including pesticides, polychlorinated organic compounds, phthalates, alkyl phenolic compounds, biphenolic compounds, heavy metals (cadmium, lead and mercury) and other hormone disrupting chemicals. In 1990 a stricter criteria for excluding minor congenital anomalies (such as glanular hypospadias) was adopted by the Office of National Statistics (ONS) which resulted in different case groups for the periods before and after 1990. Therefore, a separate analysis was performed for time periods 1980-1989, 1992-1996, as well as both time periods combined. Cases with no information on maternal occupation and districts with poor recording of occupational information were excluded from the analysis.

The purpose of the main analysis was to identify the proportion of congenital anomaly cases with hypospadias by occupational code and by category of exposure to endocrine disrupting chemicals. A subsidiary analysis using live births as a denominator for estimating the proportions was performed for checking the above results for the period 1992-1996. A total of 2,794 cases of hypospadias and 29,250 congenital anomaly cases were available for the main analysis whereas 889 hypospadias cases and 1,81,964 live births were available for the subsidiary analysis. The observed to expected ratio (O/E ratio) of hypospadias cases for each job title was estimated and adjusted for potential confounders such as year of birth, region, maternal age, social class of mother, social class of father and deprivation index of ward of residence (only for period 1992-1996) using logistic regression. Confidence limits were calculated using an exact binomial distribution. Heterogeneity in the observed to expected ratios was assessed using negative binomial regression.

An increasing trend of hypospadias cases was observed for the period 1980-1989. A similar trend was observed for maternal social classes (increasing from professional to unskilled social classes). Proportional analysis of hypospadias cases by the 348 occupational codes was suggestive of increased risk of hypospadias for the following occupational groups: management consultants (1980-1989 and all years combined), physical and geological scientists (1980-1989 only), tailors and dress makers (1980-89 and all years combined), hair dressers (1992-96, before adjusting for maternal social class), and vocational and industrial trainers (all years combined). Hair dressers constituted the largest occupational group classified as probably exposed to EDCs. The analysis of all seven EDCs separately and in combination revealed no evidence of increased proportion of hypospadias cases in the possible and probable exposure categories when compared to the category of unlikely exposure. Significant heterogeneity of odds ratios was observed for phthalates between the periods 1980-89 and 1992-96 which could be in part due to the differences in the types of hypospadias reported. However, the significance no longer persisted after adjusting for social class of mother. The results of using live births as denominators revealed no significant findings with the exception of alkylphenolic compounds which demonstrated a significant trend before adjusting for social class of mother.

Potential misclassification of exposure to EDCs and bias due to incomplete occupational information were major drawbacks of this study. Misclassification of EDCs by job exposure matrix is usually non-differential resulting in biased estimates towards null. The incompleteness in the occupational information might result in differential misclassification if the notification of congenital anomaly was dependent on occupation resulting in the bias in any direction. Using congenital anomaly cases in the denominators for the proportional analysis reduced the risk of information bias in this study. However, the inclusion of congenital anomaly cases that were associated with the exposure (EDCs) in the control group would result in an underestimation of odds ratios for the relationship between hypospadias and EDCs.

Though some indication of increased risk of hypospadias in the children of hairdressers and occupations related to phthalate exposure was observed in this study, these results should be interpreted with caution due to multiple comparisons, lack of consistent results in different time periods, lack of significant global heterogeneity and crude classification of exposures (EDCs) by job exposure matrix which did not take into account the duration and timing of exposure, and exact task. The significance of these findings in the general population however, remains uncertain due to lower concentrations of EDCs in the general population.