The authors have declared that no competing interests exist.
Conceived and designed the experiments: ARH JD LR. Performed the experiments: BAGJ CL. Analyzed the data: ARH JD LR. Wrote the paper: ARH CL JD BAGJ LR. A. RH A Contributed to conception and design, analysis and interpretation of data, drafted and critically revised as well as approved the version to be published ARH. Contributed to acquisition of data, critically revised the manuscript and approved the version to be published. CL Contributed to acquisition of data, critically revised the manuscript and approved the version to be published. BAGJ Contributed with analysis and interpretation of data, drafted and critically revised as well as approved the version to be published. JD Contributed to conception and design, analysis and interpretation of data, drafted and critically revised as well as approved the version to be published. LR.
Environmental exposures to endocrine disrupting chemicals have been suggested as a risk factor for male genital abnormalities such as hypospadias. The aim of this case-control study was to investigate the association between fetal exposure to persistent organochlorine pollutants (POP) and the risk for hypospadias.
The Southern Sweden Maternity Cohort (SSMC) contains serum samples collected in early pregnancy among women in Southern Sweden. Linkages with the Medical Birth Register, the Malformation Register and the In-patient Register resulted in 390 SSMC mothers who had given birth to a boy with hypospadias in year 1986–2002 (mean 1995). For 237 of these (cases) sufficient amounts of serum for the chemical analyses were available. For each case, a control boy from the SSMC was randomly selected, matched for maternal age, birth year, parity and maternal smoking. PCB-153,
The present study suggests that fetal exposure to HCB and
In western countries the prevalence of hypospadias varies from two to eight cases per 1,000 live births
The major DDT metabolite 1,1-dichloro-2,2-bis (
A recently published epidemiological study from Italy which included 37 serum samples from mothers with boys born with hypospadias indicated a strong association with HCB
All women were informed that a part of the sample should be stored in a biobank and could be used for research in the future. All participants gave their verbal informed consent at the time blood samples were collected. The study was performed in accordance with the Declaration of Helsinki and approved by the Research Ethics Committee at Lund University.
The Department of Clinical Microbiology, Laboratory Medicine Scania in Malmö is the only virus laboratory serving Southern Sweden. All samples submitted for virus diagnostics have been stored during 1986 and from 1988 onwards, resulting in a biobank that contains about 1,3 million serum samples from about 524.000 subjects. During a reorganization of the biobank unfortunately all samples from 1988 were thrown away. A particularly valuable part of the biobank is samples stored as a result of population-based serological screening for virus infections and rubella immunity during pregnancy (The Southern Sweden Maternity Cohort, SSMC). These samples are scheduled to be taken during week 14 of pregnancy and virtually all pregnant women in the population participate in this screening program. Up to year 2002, the SSMC contained serum samples taken during pregnancies resulting in 101262 single-born boys. A case-control study of the risk of hypospadias in relation to serum levels of POP of the mother during pregnancy was performed of the single-born boys within SSMC.
To identify boys with hypospadias within the SSMC, linkages between the Medical Birth Register (MBR), the Malformation Register and the In-patient Register were performed. In the present study boys were defined as hypospadias cases if they were present in two of the three registers, i.e. minor hypospadias were excluded. This resulted in 390 boys with hypospadia. For 237 (61%) of them, sufficient amount of serum samples for the POP analyses were found in the SSMC biobank. Additional information about the pregnancy was retrieved from the MBR (
Cases | Controls | |||
n | (%) | N | (%) | |
|
||||
Birth weight | ||||
1986–1987 | 18 | (8) | 18 | (8) |
1988–1993 | 73 | (31) | 74 | (31) |
1994–1998 | 65 | (27) | 62 | (26) |
1999–2002 | 81 | (34) | 83 | (35) |
Maternal age (years) | ||||
<25 | 50 | (21) | 50 | (21) |
25–29 | 91 | (38) | 92 | (39) |
30–34 | 61 | (26) | 60 | (25) |
>35 | 35 | (15) | 35 | (15) |
Parity | ||||
1 | 120 | (51) | 119 | (50) |
2 | 77 | (32) | 76 | (32) |
>2 | 40 | (17) | 42 | (18) |
Maternal smoking in early pregnancy |
||||
No | 180 | (81) | 183 | (82) |
Yes | 42 | (19) | 41 | (18) |
Unknown | 15 | 13 | ||
|
||||
Gestational age (weeks) | ||||
<37 | 24 | (10) | 14 | (6) |
37–38 | 45 | (19) | 39 | (16) |
39–40 | 114 | (48) | 122 | (52) |
>40 | 54 | (23) | 62 | (26) |
Birth weight |
||||
<2500 | 21 | (9) | 9 | (4) |
>4000 | 38 | (16) | 43 | (18) |
Family situation |
||||
Cohabiting with the father | 213 | (94) | 207 | (92) |
Else | 14 | (6) | 17 | (7) |
Working status in early pregnancy |
||||
Full time | 104 | (51) | 106 | (52) |
Part time | 52 | (26) | 66 | (32) |
Not working | 47 | (23) | 34 | (16) |
Caesarean section | 32 | (14) | 30 | (13) |
Data missing for 15 cases and 13 controls;
Data missing for 2 cases and 2 controls;
Data missing for 10 cases and 13 controls;
Data missing for 34 cases and 31 controls.
For each case a control boy was randomly selected from the SSMC. The aim was to match the controls for maternal age, birth year, parity and maternal smoking habits in early pregnancy, which was successful for almost all controls (
In the present study 2,2′,4,4′,5,5′-hexachlorobiphenyl (PCB-153) was selected as a biomarker for PCB exposure as well as dioxin exposure since PCB-153 correlates very well (r = 0.98) with total PCB concentrations in plasma and serum from Swedish individuals and with the TCDD equivalent (TEQ) in plasma from PCB (r = 0.89)
PCB-153,
The association between the levels of maternal POP concentrations and the risk for hypospadias in the offspring was evaluated by conditional logistic regression (using the statistical software EGRET), given odds ratios (OR) as the risk measure with 95% confidence intervals (CI). The exposure variables (PCB-153,
Due to the relatively high correlation between PCB-153,
For three cases and for five controls it was not possible to determine PCB-153 and
The median maternal serum concentrations of PCB-153,
No statistically significant dose-response associations were observed between maternal serum concentrations of PCB-153,
Cases | Controls | OR | 95% CI | p-value for trend | |
PCB-153 (ng/mL) | 0.13 | ||||
−0.25 (ref) | 68 | 62 | 1.00 | – | |
>0.25–0.48 | 59 | 53 | 0.96 | 0.57, 1.62 | |
>0.48–0.76 | 54 | 57 | 0.75 | 0.42, 1.35 | |
>0.76 | 48 | 57 | 0.60 | 0.30, 1.19 | |
p,p’-DDE (ng/mL) | 0.09 | ||||
−0.10 (ref) | 46 | 59 | 1.00 | – | |
>0.10–1.00 | 52 | 54 | 1.33 | 0.73, 2.44 | |
>1.00–2.20 | 67 | 59 | 1.69 | 0.93, 3.08 | |
>2.20 | 64 | 57 | 1.68 | 0.92, 3.08 | |
HCB (ng/mL) | 0.12 | ||||
−0.15 (ref) | 63 | 64 | 1.00 | – | |
>0.15–0.19 | 47 | 60 | 0.80 | 0.46, 1.38 | |
>0.19–0.26 | 52 | 55 | 1.04 | 0.59, 1.84 | |
>0.26 | 75 | 58 | 1.59 | 0.86, 2.93 |
Odds ratios (OR) and 95% confidence intervals (CI) were obtained with conditional logistic regression analysis.
When the upper two quartiles (i.e. above median) of
Cases | Controls | OR | 95% CI | |
p,p’-DDE (ng/mL) | ||||
−0.10 (ref) | 46 | 59 | 1.00 | – |
>0.10–1.00 | 52 | 54 | 1.33 | 0.73, 2.44 |
>1.00 | 131 | 116 | 1.69 | 0.97, 2.93 |
HCB (ng/mL) | ||||
−0.26 (ref) | 162 | 179 | 1.00 | – |
>0.26 | 75 | 58 | 1.65 | 1.02, 2.69 |
Odds ratios (OR) and 95% confidence intervals (CI) were obtained with conditional logistic regression analysis.
No statistically significant trends were observed between maternal serum concentrations of POP mixtures and the risk for hypospadias (
Cases | Controls | OR | 95% CI | p-value for trend | |
PCB-153+ p,p’-DDE + HCB |
1.00 | ||||
Low (ref) | 16 | 15 | 1.00 | – | |
Medium | 188 | 190 | 0.92 | 0.42, 2.02 | |
High | 25 | 24 | 0.98 | 0.35, 2.77 | |
PCB-153+p,p’-DDE |
0.74 | ||||
Low (ref) | 26 | 30 | 1.00 | – | |
Medium | 173 | 168 | 1.24 | 0.65, 2.35 | |
High | 30 | 31 | 1.16 | 0.49, 2.79 | |
PCB-153+ HCB |
0.69 | ||||
Low (ref) | 37 | 29 | 1.00 | – | |
Medium | 153 | 165 | 0.69 | 0.39, 1.25 | |
High | 39 | 35 | 0.85 | 0.38, 1.88 | |
p,p’-DDE + HCB |
0.13 | ||||
Low (ref) | 21 | 28 | 1.00 | – | |
Medium | 166 | 167 | 1.41 | 0.74, 2.67 | |
High | 42 | 34 | 1.86 | 0.83, 4.20 |
Odds ratios (OR) and 95% confidence intervals (CI) were obtained with conditional logistic regression analysis.
Low = children whose mothers had exposure concentrations in the lowest quartile for all exposure variables, High = children whose mothers had exposure concentrations in the highest quartile for all exposure variables, and Medium = all other.
No statistically significant dose-response associations were observed between maternal serum concentrations of PCB-153,
A recently published study by Giordano and colleagues found a fivefold increased risk of having a boy with hypospadias among Italian women with serum HCB concentrations above 0.16 ng/g (i.e. the median concentration in the study) as compared to women with lower HCB concentrations
In recently published studies on PCB concentrations in breast milk and placenta and risk for cryptorchidism, another male genital disorder which share many risk factors with hypospadias, no increased risk of cryptorchidism where seen
However, Damgaard et al found a positive association between chlorinated pesticides and cryptorchidism
The registers in Sweden are almost complete (98%) and by the use of different registers we will most probably miss very few hypospadias cases. A possible weakness is that the material may be skewed to severe cases that required operative treatment and the mild cases have not been included. It has been postulated that environmental factors would be particularly important in the pathogenesis of these cases, whereas the severe cases with proximal hypospadias may be more often associated with genetic defects. However, these cases of hypospadias are very few.
In addition, information about potential confounders is available from the MBR. On the other hand did we lack or did not have sufficient amount of maternal serum for almost 40% of the cases. It is, however, not reasonable to believe that this selection is due to POP levels. The fact that the distributions for the few socio-economic variables we had information about, i.e. family situation and working status in early pregnancy, were very similar among the cases and the controls do not support that selection bias is a major problem in the present student. Another possible limitation in the present study was the small amount of serum available from the biobank. The chemical analyses were therefore restricted to PCB-153,
The foetal period is a vulnerable period for toxic substances. POPs can pass the placental barrier and most POPs have very long half-lives, five to ten years. It is then reasonable to believe that POP levels measured in serum women reflect the
The hormone hypothesis of male reproductive disorders suggests that both endogenous and exogenous hormones might be a risk factor for hypospadias. Endocrine disrupting chemicals, such as PCBs,
The present study indicates that