Longitudinal association between saliva and hair cortisol concentration: A systematic comparison

Cortisol assays from hair have become increasingly common in psychoneuroendocrinological research as indicators of long-term output relevant to stress and health outcomes. Comparisons of hair cortisol concentration (HCC) with salivary samples have produced mixed findings, and it remains unclear which aspects of the diurnal salivary profile correspond most closely to HCC, and what time intervals between saliva and hair sampling are most relevant, taking the rate of hair growth into account. This longitudinal study aimed to evaluate the correspondence between HCC and parameters of total salivary cortisol output in the morning (CAR_auc and CARi) and during the rest of the day excluding the early morning period (DAY_auc), by systematically studying three time periods - two weeks, four weeks, and six weeks - before hair sampling. At each time period, 54 female university students (mean age: 20.85 ± 1.16 years) provided three saliva cortisol samples on day 1 at 11 am, 3 pm, at bedtime, then two samples the following day on waking and 30 min after awakening. Hair strand collection (1 cm nearest the scalp) took place two weeks after the last saliva sample. Results of multivariable regressions indicate that HCC was consistently associated with DAY_auc for all three time periods and with the aggregate DAY_auc across days after adjusting for age, body mass index, smoking, oral contraceptive use, hair washing frequency and hair treatments. The strongest associations were found for DAY_auc two weeks before hair sampling (β = 0.578, p < 0.001) and the aggregated DAY_auc across all three time periods (β = 0.596, p < 0.001), although the confidence intervals overlapped those for four and six week analyses. There was no significant association between HCC and either CAR_auc or CARi. Our study confirms that hair cortisol could be a reliable retrospective biomarker of basal and long-term cortisol output secretion at least up to six weeks earlier. The results contribute to a better understanding of the different associations between HCC and salivary cortisol in the morning and the rest of the day, while also having implications for the use of HCC as an outcome measure in intervention and treatment research.