Table 3.

Prospective influence of centrality on depression

	Dependent variable: current CES-D score
	Co-eff	s.e.	P-value
Previous network centrality	– 8.285	4.142	0.045
Previous number of friends	–0.278	0.094	0.003
Previous number of family	0.456	0.019	0.000
Previous CES-D score	0.036	0.028	0.195
Age	0.022	0.008	0.006
Years of education	–0.168	0.037	0.000
Female	1.110	0.159	0.000
Exam 7	1.161	0.169	0.000
Constant	–4.469	1.454	0.002
Deviance	244061
Null deviance	327588
N	6113

Abbreviation: CES-D, Center for Epidemiological Studies Depression Scale.

Model of ego's CES-D score at current exam regressed on measures from the previous exam including ego's network centrality, number of friends, number of family, plus other covariates. The model was estimated using a general estimating equation (GEE) with clustering on the ego and an independent working covariance structure. Models with an exchangeable correlation structure yielded poorer fit. Fit statistics show sum of squared deviance between predicted and observed values for the model and a null model with no covariates. The main results (coefficients in bold) show that network centrality is associated with a decrease in future depressive symptoms, even controlling for the number of friends and family. This suggests that connection to more socially distant alters (for example, friends of friends) also influences depression.