METHODS

Using a simple questionnaire, number of sons and daughters of 49 men working in filling stations in Shiraz (Fars province, south of Iran) were determined. We identified 115 offspring (47 males, 68 females) within these families. The mean duration of employment in the filling station was 7.7 years (range 1.8–29 years). Because it is reported that paternal age and birth order have some effect on offspring sex ratio,^3,10 for each exposed subjects, three unexposed persons from the general population of Shiraz (without occupational exposure to gasoline) were matched by age of fathers (±2 years) and number of children as a control group. In the control group, 345 offspring (178 males, 167 females) were identified.

Blood samples were obtained from 37 of 49 filling station workers participating in the study. Genomic DNA was isolated from whole blood. The polymerase chain reaction (PCR) conditions for determining GSTM1 and GSTT1 genotypes were the same as that reported previously.^5,6 The absence of amplified product was consistent with the null genotype. Successful amplification by β-globin specific primers confirmed the proper function of the PCR reaction.¹¹

The sex ratio expressed as the proportion of total live births that were male (male proportion). The odds ratio (OR) (the odds of being male) and 95% confidence intervals (CI) were calculated. An OR>1.0 indicates an increase in the sex ratio and an OR<1.0 indicates a decrease in the sex ratio. Comparison of paternal age and number of children was done using two tailed unpaired Student’s t test. A probability of p<0.05 was considered a significant difference.

RESULTS

The sex ratio at birth of exposed and unexposed groups were 0.409 (total birth = 115) and 0.516 (total birth = 345), respectively. The offspring sex ratio at birth in the filling station workers was significantly lower than the ratio in the control group (OR = 0.65; 95% CI = 0.42 to 0.99).

Table 1 shows the association of paternal GSTM1 and GSTT1 genotype combinations and offspring sex ratio in exposed group. Because it is reported that in general population, the GSTM1 and GSTT1 genotypes and their combinations showed no association with offspring sex ratio,¹¹ the sex ratio in each combination, was compared with the control group. The offspring sex ratio in exposed group with active GSTM1 and GSTT1 genes was the same as the ratio in control group (OR = 0.66; 95% CI = 0.34 to 1.28). However, when the GSTT1 became inactive (active GSTM1 and null genotype of GSTT1) the sex ratio in the offspring significantly decreased (OR = 0.45; 95% CI = 0.21 to 0.96). Statistical analysis for the two other combinations was not performed, because the sample size was too small.

It should be noted that there was no statistical difference for paternal age (t = 0.86, df = 28, p = 0.40) and number of children (t = 0.72, df = 28, p = 0.48) between the group with two active genes and the group with active GSTM1 and GSTT1 null genotype.

DISCUSSION

It is reported by Yang et al that the association between exposure to petroleum air pollution near a petroleum refinery plant in Taiwan and sex ratio at birth was not significant.¹² Recently we reported the increased sex ratio of live births in Masjid-i-Sulaiman, south west of Iran, which is contaminated by sub-surface leakage of natural gas containing hydrogen sulphide.¹³ The present results showed that sex ratio decreased in the offspring of filling station workers. The increased sex ratio in Masjid-i-Sulaiman, however, can be attributed to the exposure of both parents to the toxin.

Recently, we reported that in the general population, the GSTM1 and GSTT1 genotypes and their combinations showed no association with offspring sex ratio.¹¹ Our results may indicate that the offspring sex ratio is associated with combination genotypes of GSTM1 and GSTT1 of exposed group. It is reported that GSTT1 is involved in detoxification of several toxins including some of the compounds present in gasoline.⁷ A possible explanation for the change in sex ratio when the fathers have null genotype of GSTT1/active GSTM1 is disruption of detoxification processes of some gasoline toxins or their metabolites. In this status conditions became unfavourable. The finding reported here is in agreement with the hypothesis of Trivers and Willard. They hypothesised that under favourable environmental conditions males would be favoured and that under adverse conditions females would be favoured in polygynous species.¹⁴ There is some evidence that supports the Trivers and Willard hypothesis in human populations.^15,16 A few strong men could sire many grandchildren, whereas women have a more sure but limited reproductive potential. Taken together, it is suggested that under unfavourable conditions, the null genotype of GSTT1 shows a selective advantage.