Curvilinear associations between family income in early childhood and the cortisol awakening response in adolescence

Abstract

Previous evidence on cortisol output and socioeconomic status (SES) has been mixed, with studies finding that lower SES can be associated with higher or lower cortisol output, and null associations have also been reported. We hypothesized that these inconsistencies may be due to an underlying curvilinear, inverted-U pattern of association, such that low income is related to increased likelihood of both low and high cortisol output. We tested these curvilinear links among family income and cortisol indices in the Avon Longitudinal Study of Parents and Children (N = 803). Maternal reports of family income when the study children were 33 and 47 months of age were averaged to estimate early-childhood family income. Three cortisol indices were derived from samples collected in adolescence (15.5 years of age): the cortisol awakening response (CAR), area under the curve (AUC) cortisol, and the diurnal cortisol slope. As hypothesized, the CAR exhibited a curvilinear, inverted-U relation with childhood income, with low childhood income being associated with both the lowest and the highest CARs. These findings suggest that discrepancies in prior findings on low SES and the CAR may be due to curvilinear patterns of association. However, childhood income was not significantly associated with adolescent cortisol diurnal slope or AUC. Future work should clarify the factors that might predispose to high versus low CAR given equivalent low SES in childhood.

Keywords: ALSPAC; Cortisol; Curvilinear associations; Family income; HPA axis.

Figure 1.

Quadratic association between childhood income and the cortisol awakening response. Lower childhood income predicted both hypo- and hyperactive cortisol awakening responses. The probability density function of the cortisol awakening response is represented on the top of the scatterplot and the probability density function of childhood income is represented on the right side of the scatterplot.

Figure 2.

Quadratic association between parental education and the cortisol awakening response. Lower parental education in early childhood predicted both hypo- and hyperactive cortisol awakening responses. The probability density function of the cortisol awakening response is represented on the top of the scatterplot and the probability density function of parental education is represented on the right side of the scatterplot.