Authors;
Vasiliu O., Muttineni J., Karmaus W.

Title:
In utero exposure to organochlorines and age at menarche

Source:
Human Reproduction 19(7): 1506-1512, 2004

Summary:
Polychlorinated biphenyls (PCBs) and dichlorodiphenyltrichlorethane (DDT) are important organochlorine pollutants found in the Great Lakes. These chemicals bioaccumulate in the food chain and can be detected in human tissues and serum. Previous studies have suggested that Michigan anglers and fish eaters possess higher serum levels of PCB's and dichlorodiphenlydichloroethylene (DDE), the metabolite of DDT, than the general population. Studies have also shown PCB's and DDE to be endocrine active compounds which may have an impact on sexual maturation. The current study by Vasiliu et al. was conducted to determine if the female offspring of Michigan anglers exposed prenatally to PCB's and DDE exhibit an earlier age of menarche (onset of menstruation at puberty).

A retrospective cohort study was conducted over two generations: mothers and their female offspring. The initial cohort was derived from a series of 3 surveys conducted between 1973 and 1991 on Michigan anglers in which organochlorine serum samples were collected. Inclusion criteria for this cohort included women who were capable of having children between 1950 and 1980, and had recorded organochlorine measurements. Of 319 eligible participants, 259 women were recruited; a response rate of 66.2%. The female offspring of these women (aged 20-50 years at the time of the study) were also recruited and formed the second generation of the study. Of 213 eligible offspring, 151 (71%) agreed to participate. Using data collected from the initial surveys, backwards linear regression was used to estimate the organochlorine serum levels of the mothers at the approximate time of the birth of their offspring. Multiple organochlorine serum measurements were available for a large proportion (71.3%) of women who participated in more than one of the initial surveys. This data was used to validate the predicted results. Additional data (e.g. age of menarche, birth weight) was obtained by surveying each generation of the cohort. Each offspring was linked to their mother's extrapolated DDE and PCB serum concentrations at birth. Linear regression was then used to estimate the association between maternal PCB and DDE levels at birth and age of menarche for the female offspring. Adjustments were made for birth year cohorts, maternal height and age, birth weight, breastfeeding and educational status of the offspring.

Age of menarche of the offspring ranged between 9 and 17 years and was normally distributed. A significant association (p=0.038) was found between DDE and age of menarche, wherein a 15µg/l increase in DDE concentration reduced the age of menarche by one year. However, when body size at menarche was incorporated in the model the association ceased to be significant. No association between maternal PCB levels and age of menarche was observed.

A strength of the study was the fact that the techniques and standards of organochlorine serum determination remained constant for all three iterations of the initial survey and enabled the authors to make meaningful comparisons across cohorts. The multiple iterations of the survey also allowed for validation of the predicted numbers obtained from the regression line, giving greater confidence to the results.

The study also had several limitations. No serum DDE measurement was taken during the first cycle of the survey, thus the predicted serum concentrations at time of birth may be less robust for earlier birth years. Outliers do not appear to have been excluded from the regression analysis examining the association between DDE levels at birth and age at menarche, which may skew the results to show an association. Also no data was collected on dietary consumption during childhood. It is possible that offspring whose mothers had higher DDE levels were also exposed to higher levels of DDT through their diet (e.g. Great Lakes fish) as a child. Consequently, the earlier age of menarche may be due to these dietary or other chemical exposures and not in utero exposure. In addition the retrospective nature of the study creates the potential for recall bias. Women were asked to recall information such as their body size at menarche, and age at menarche. While in the latter case misclassification can be expected to be non-differential, bias due to recalling body size may substantial. Finally, it is unlikely that the study findings can be generalized to the population at large. The study population was comprised of anglers and their spouses who presumably consume large amounts Great Lakes fish, and are thus exposed to higher levels of contaminants than the general population. Different outcomes may be associated with lower exposure levels.

The effect of body size on age of menarche is particularly of note. Previous studies have found a higher body mass index (BMI) in adolescence to be associated with a lower age of menarche. This relationship is complex as adipose (fat) tissue may contribute to the pool of available estrogens, thereby contributing to the onset of puberty and thus, menarche. This relationship may therefore explain the confounding effect of body size on the DDE and age of menarche association observed in this study.