Puberty and the human brain: Insights into adolescent development

Abstract

Alongside the exponential flourish of research on age-related trajectories of human brain development during childhood and adolescence in the past two decades, there has been an increase in the body of work examining the association between pubertal development and brain maturation. This review systematically examines empirical research on puberty-related structural and functional brain development in humans, with the aim of identifying convergent patterns of associations. We emphasize longitudinal studies, and discuss pervasive but oft-overlooked methodological issues that may be contributing to inconsistent findings and hindering progress (e.g., conflating distinct pubertal indices and different measurement instruments). We also briefly evaluate support for prominent models of adolescent neurodevelopment that hypothesize puberty-related changes in brain regions involved in affective and motivational processes. For the field to progress, replication studies are needed to help resolve current inconsistencies and gain a clearer understanding of pubertal associations with brain development in humans, knowledge that is crucial to make sense of the changes in psychosocial functioning, risk behavior, and mental health during adolescence.

Keywords: Adolescence; Brain development; Functional MRI; Hormones; Puberty; Structural MRI.

Figure 1.

Pubertal associations with regional cortical grey matter (from crosssectional studies), after accounting for age. Darker arrows represent lateral findings, while lighter arrows represent medial findings. Dashed lines connect findings from same study that fall within same anatomical subdivisions. Upward and downward arrows represent positive and negative correlations, respectively. Readers who would like further orientation to brain structure may consider exploring the interactive viewer available at http://www.brainfacts.org/3D-Brain. Pubertal stage: Hu et al., 2013; Koolschijn et al., 2014; Peper et al., 2009b; Pfefferbaum et al., 2015. Testosterone: Bramen et al., 2012; Koolschijn et al., 2014; Neufang et al., 2009 (null effects: Peper et al., 2009a). Estradiol: Brouwer et al., 2015; Koolschijn et al., 2014; Neufang et al., 2009; Peper et al., 2009a.

Figure 2.

Pubertal associations with subcortical grey matter (from cross-sectional studies) in males (M) and females (F). Circles represent findings from studies that ran analyses both controlling and not controlling for age. Pubertal stage: Blanton et al., 2012*; Bramen et al., 2011*; Hu et al., 2013*; Koolschijn et al., 2014; Neufang et al., 2009; Peper et al., 2009b; Satterthwaite et al., 2014*; Urosevic et al., 2014. Testosterone: Bramen et al., 2011*; Koolschijn et al., 2014; Neufang et al., 2009; Peper et al., 2009a; Brouwer et al., 2015. Estradiol: Koolschijn et al., 2014; Neufang et al., 2009; Peper et al., 2009a; Brouwer et al., 2015. *Studies only examined amygdala and hippocampus.

Figure 3.

Pubertal associations with amygdala development from longitudinal studies. a) Goddings et al., 2014; b,c) Herting et al., 2014.

Figure 4.

Pubertal associations with reward-related brain function. a) Darker arrows represent lateral findings, while lighter arrows represent medial findings. Dashed circles highlight findings that did not control for age. NB: Results using lowered (uncorrected) statistical thresholds from Op de Macks et al. (2011) are not presented. b) Striatal findings (all studies used ROIs). NB: Longitudinal findings from Braams et al., (2015) are not presented. T statistics and correlation coefficients were converted to Z statistics for consistency across studies.

Figure 5.

Pubertal associations with affect-related brain function. a) Darker arrows represent lateral findings, while lighter arrows represent medial findings. Dashed circles highlight findings that did not control for age. Dashed lines connect findings from same study that fall within the same anatomical subdivision. b) Amygdala findings (studies using whole brain analyses highlighted in bold). NB: Longitudinal findings from Spielberg et al. (2014; 2015) on task-based connectivity are not presented. T statistics and correlation coefficients were converted to Z statistics for consistency across studies.