Table 2.

Summary of meta-analytic results

	k	Cohen’s d	95% CI	p
Overall analyses
Hippocampal volume					QW (df)	p	I2
Total	23	−0.53	−0.74 to −0.31	<0.001	124.83 (22)	<0.001	0.82
Left	17	−0.61	−0.96 to −0.26	<0.001	112.97 (16)	<0.001	0.86
Right	17	−0.73	−1.15 to −0.32	<0.001	152.65 (16)	<0.001	0.90
Moderator analyses: sample characteristics
Alcohol use					QB (df)	p
Clinical	19	−0.65	−0.90 to −0.39	<0.001
Subclinical	4	−0.07	−0.47 to 0.34	0.752
					5.64 (1)	0.018
Age
Adult	17	−0.68	−0.92 to −0.44	<0.001
Adolescent	6	−0.04	−0.31 to 0.23	0.763
					11.84 (1)	<0.001
Age (clinical alcohol use)
Adult	16	−0.70	−0.98 to −0.42	<0.001
Adolescent	3	−0.35	−1.00 to 0.30	0.295
					0.93 (1)	0.335
Sex
Male	11	−0.69	−1.05 to −0.32	<0.001
Female	4	−0.83	−2.07 to 0.42	0.193
					0.05 (1)	0.832
Moderator analyses: methodological variables
Segmentation method					QB (df)	p
Hand tracing	14	−0.62	−0.92 to −0.32	<0.001
Automated	9	−0.38	−0.70 to −0.07	0.016
					1.15 (1)	0.283
Scanner strength
1.0 T/1.5 T	16	−0.58	−0.82 to −0.33	<0.001
3 T	7	−0.42	−0.86 to 0.02	0.062
					0.37 (1)	0.543
Covariate adjustment
Unadjusted	14	−0.55	−0.94 to −0.16	0.006
Adjusted	9	−0.49	−0.70 to −0.28	<0.001
					0.07 (1)	0.791

Notes: k = number of studies. Negative Cohen’s ds indicate reduced hippocampal volumes associated with alcohol use. By convention, ds > |0.20| are considered to be modest, >|0.50| moderate, and >|0.80| large (Cohen, 1988). CI, confidence interval. Q_W = heterogeneity within a set of studies. Q_B = heterogeneity between two sets of studies. df = degrees of freedom. I² = percentage of total variance across studies due to heterogeneity.