. Author manuscript; available in PMC: 2019 Aug 1.

Published in final edited form as: Health Psychol. 2018 Aug;37(8):701–715. doi: 10.1037/hea0000637

Table 2.

Cumulative hazard ratios of cumulative childhood adversity on cardiometabolic disease.

	Cumulative Hazard Ratio	95%CI	Q_T (heterogeneity)
All effects (k=15)	1.42^*	1.20 – 1.67	61.61^*
Using study as the unit of analysis (K=9)	1.38^*	1.15 – 1.66	52.11^*
Participants
Sex^a
Women (k=3)	1.62^*	1.08 – 2.43	6.30^*
Men (k=2)	1.01	.59 – 1.73	.002
Q_M (1, 3)			1.92
Outcome^c
Type
Metabolic outcomes (k=3)	1.63^*	1.16 – 2.29	10.89^*
CVD clinical outcomes (k=12)	1.36^*	1.13 – 1.63	41.11^*
Q_M (1, 13)			.84
Measurement
Self-report (k=7)	1.59^*	1.23 – 2.05	42.91^*
Independently-assessed disease (k=8)	1.30^*	1.02 – 1.66	15.51^*
Q_M (1, 13)			1.26
Cumulative Adversity Index
Type of Adversity Measure
ACE Study Questionnaire (k=3)	3.41^*	2.31 – 5.03	1.81
Other Scale/Ad-hoc Composite (k=12)	1.23^*	1.09 – 1.38	24.44^*
Q_M (1, 13)			24.29^*
Index Included Childhood SES
No (k=4)	2.26^*	1.66 – 3.07	19.11^*
Yes (k=11)	1.22^*	1.04 – 1.43	19.46^*
Q_M (1, 13)			12.13^*
Index Included Household Dysfunction
No (k=5)	1.16	0.86 – 1.56	5.63
Yes (k=10)	1.61^*	1.29 – 2.01	55.74^*
Q_M (1, 13)			3.04^†
Index Included Abuse/Neglect
No (k=7)	1.27^*	.98 – 1.66	14.21^*
Yes (k=8)	1.56^*	1.24 – 1.96	46.99^*
Q_M (1, 13)			1.28
Analytic Strategy
Number of Covariates in Analytic Models
5 or fewer (k=5)	1.99^*	1.48 – 2.62	27.48^*
6 or more (k=10)	1.22^*	1.01 – 1.47	19.49^*
Q_M (1, 13)			7.48^*
Psychosocial Covariate
No (k=4)	2.26^*	1.66 – 3.07	19.11^*
Yes (k=11)	1.22^*	1.04 – 1.43	19.46^*
Q_M (1, 13)			12.13^*
Health Behavior Covariate
No (k=5)	1.97^*	1.48 – 2.62	27.48^*
Yes (k=10)	1.22^*	1.01 – 1.47	19.49^*
Q_M (1, 13)			7.48^*

Note. Q_M = test of between group difference. K denotes studies and k denotes effects (i.e., cumulative adversity on one outcome). Cumulative effect sizes are represented using hazard ratios. Results reflect estimated effects of cumulative adversity on combined cardiometabolic outcomes, with the exception of results for moderation by “Outcome Type.” Bolded values highlight significant moderator variables.

Greater than 80% of the sample had to be of the same sex or results presented separately by sex to code the sample as predominantly composed of men or women.

p<.05.

^†

p<.10.