Table 2

Overall summary estimates of effect for the associations between cortisol measures and well-being from a series of meta-analyses

Included cohortsNumber of individualsMean difference in WEMWBS score (95% CI) per SD increase in cortisol
Model 1*Model 2†
Regression coefficient
(95% CI)
p ValueTests of heterogeneity
I2(%)
p Value‡Regression coefficient
(95% CI)
p ValueTests of heterogeneity
I2(%)
p Value‡
Early morning cortisolAll75150.02
(−0.17 to 0.21)
0.82.30.40.01
(−0.22 to 0.24)
0.918.00.3
Evening cortisolCaPS
NSHD
1756−0.47
(−1.00 to 0.05)
0.0820.30.3−0.31
(−0.83 to 0.21)
0.219.10.3
Flatter diurnal slopeCaPS
NSHD
NCDS
6490−0.07
(−0.27 to 0.14)
0.50.00.6−0.08
(−0.28 to 0.13)
0.50.00.6
CARCaPS
NSHD
1612−0.19
(−0.62 to 0.24)
0.40.01.0−0.16
(−0.59 to 0.26)
0.40.00.9
  • *Sex, age at cortisol assessment, follow-up time to well-being assessment.

  • Model 1 plus body mass index, smoking status, adult social class.

  • ‡p Values from Cochran’s Q statistic performed as a test of between-study heterogeneity.

  • CaPS, Caerphilly Prospective Study; CAR, cortisol awakening response; HCS, Hertfordshire Cohort Study; NCDS, National Child Development Study; NSHD, National Survey of Health and Development; WEMWBS, Warwick Edinburgh Mental Wellbeing Scale.