Abdelouahab et al. Respond to "Maternal PBDEs and Thyroid Hormones"

In his commentary on our article, Dr. Jonathan Chevrier (1) raises an important question concerning the lack of consistency among epidemiologic studies on the association between polybrominated diphenyl ethers (PBDEs) and thyroid hormone levels. This contrasts with the highly consistent experimental data showing decreases in thyroid hormone levels in animals exposed to PBDEs even at low doses relevant to humans. Strikingly, human studies addressing neurotoxicity and lead exposure show high interstudy consistency and are highly congruent with animal studies (2). The epidemiologic studies on lead used similar designs applied to diverse populations, which differed greatly in terms of sociodemographic characteristics. Clearly, appropriate study design remains the major challenge when thyroid function is addressed in relation to low-dose exposures in humans. We hypothesize that interstudy inconsistencies for PBDEs result from 1) the divergence in study designs, especially lack of control for major risk factors for thyroid deficit, and 2) the fundamental limitations of epidemiology in the detection of subtle adverse health effects, as was elegantly discussed by Taubes (3). The majority of large studies on PBDEs and thyroid function have used existing biobanks from cohort studies designed for other purposes, which probably explains why thyroid function has been addressed at different time points and why a large number of parameters specific to thyroid diseases (e.g., thyroid-specific antibodies, iodine, selenium) have not been controlled for. The limited sensitivity and specificity of epidemiologic studies related to low-dose exposures (4) probably results in some “statistically significant” associations that are almost certainly false-positive. We agree with the arguments on iodine intake, but selenium status is also an important determinant of thyroid status (5–7). In our study, whole-blood selenium was positively correlated with total thyroid hormone and free triiodothyronine levels early in pregnancy and at delivery: It explained 1%–5% of the total variance in these hormones in pregnant women, but it was not correlated with hormone levels in umbilical-cord blood (8). We considered the possibility of reverse causality; however, the published experimental data on PBDEs are consistent with direct causality (i.e., PBDEs directly reducing thyroid hormone levels). In our longitudinal study, reverse causality cannot apply because thyroid hormones in cord blood cannot influence maternal PBDE and lipid levels. There is scant information about how thyroid hormones influence both blood lipids and lipophilic contaminants in normal pregnancy, when the metabolism of both thyroid hormones and lipids changes dramatically. It is unlikely that the knowledge obtained in nonpregnant subjects with clinically expressed hypothyroidism can be directly extrapolated to pregnant women with normal thyroid hormone levels. Moreover, epidemiologic studies on thyroid status have not examined the hypothesis of a correlation between lipids and lipophilic contaminants. Paradoxically, in our study, PBDEs, which are lipophilic contaminants, were not correlated with total lipid concentrations in blood, while polychlorinated biphenyls, which are also lipophilic contaminants, were highly positively correlated (data not shown). We used another database including PBDE levels and blood lipid levels (determined by the Phillips formula) from 42 nonpregnant subjects from our region (9) and found the same results (data not shown). This indicates that arguments for reverse causality as well as for lipid adjustment, the default method in most studies, are more relevant for polychlorinated biphenyls than for PBDEs. We agree that the question of lipid adjustment needs to be carefully reevaluated. However, the majority of published studies on PBDEs and thyroid hormones, findings from which are inconsistent, used very similar methods of lipid measurement and used lipid-standardized concentrations of PBDE. Hence, this is clearly not a source of inconsistency, but the question of whether lipid adjustment is valid in the case of certain lipophilic contaminants, such as PBDEs, requires debate. There is a clear need for consensus on the methodology of human studies addressing thyroid disruption by environmental contaminants, from initial design to data analysis and clinical interpretation. Author affiliations: Department of Pediatrics, Faculty of Medicine, University of Sherbrooke, Sherbrooke, Quebec, Canada (Nadia Abdelouahab, Laetiscia Lavoie, Larissa Takser); Endocrinology Service, Department of Medicine, Faculty of Medicine, University of Sherbrooke, Sherbrooke, Quebec, Canada (Marie-France Langlois); Department of Biochemistry, Faculty of Medicine, University of Sherbrooke, Sherbrooke, Quebec, Canada (François Corbin); and Department of Gynecology and Obstetrics, Faculty of Medicine, University of Sherbrooke, Sherbrooke, Quebec, Canada (Jean-Charles Pasquier). Conflict of interest: none declared.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.