Pubertal hormones and brain structure: Exploring the value of hair assays

Neuroimaging research has highlighted the role of puberty in structural brain development in humans, but studies investigating the mechanistic role of hormones in this association have produced inconsistent findings. Limitations of current approaches to hormonal assessments have long been recognized, as hormone levels are susceptible to momentary influences (in particular, circadian rhythmicity and menstrual cyclicity). In this study, we present a comprehensive characterization of hormonal associations with brain structure using two methods of assessing hormone concentrations that are less susceptible to extraneous factors. We focus on effect sizes estimations as varied analytic strategies employed in prior research may be contributing to inconsistencies, and also demonstrate the effects of controlling for age or pubertal stage. Associations were examined in a cross-sectional sample of early adolescent females (N = 112, 10-13 years of age). We collected an average of four saliva samples across a month to calculate “basal” pubertal hormone concentrations (estradiol, testosterone, and DHEA). We also used the relatively novel method of assaying these hormones in hair; 5cm hair samples were collected proximal to the scalp, reflecting approximately 5 months of hormonal exposure. Participants underwent structural MRI scans, and estimates of cortical thickness and subcortical volume were obtained. Findings revealed that DHEA and testosterone concentrations in saliva samples exhibited strongest associations with parieto-occipital cortices, and controlling for age and pubertal stage considerably decreased the strength of these associations. Comparatively, DHEA and estradiol concentrations in hair exhibited strong negative associations with cingulate and lateral prefrontal cortical thickness, and controlling for age and pubertal stage resulted in minimal change in these associations. They also provide preliminary evidence that hair and saliva samples may be tracking different aspects of neuroendocrine processes, and highlight the need for further investigations into the potential utility of hair samples in pubertal neuroimaging research.