Table 2.

Prospective influence of friends and family on depression and vice versa

	Dependent variable
	Current CES-D score			Current number of friends			Current number of family
	Co-eff	s.e.	P-value	Co-eff	s.e.	P-value	Co-eff	s.e.	P-value
Previous CES-D score	0.456	0.019	0.000	–0.002	0.001	0.008	–0.001	0.001	0.142
Previous number of friends	–0.289	0.093	0.002	0.901	0.007	0.000	0.933	0.003	0.000
Previous number of family	0.024	0.025	0.323	–0.003	0.002	0.042	–0.029	0.007	0.000
Age	0.022	0.008	0.007	–0.002	0.000	0.000	0.001	0.001	0.004
Years of education	–0.169	0.037	0.000	0.002	0.002	0.206	–0.006	0.003	0.020
Female	1.106	0.159	0.000	–0.015	0.009	0.091	0.015	0.012	0.213
Exam 7	1.159	0.169	0.000	0.005	0.009	0.543	0.039	0.012	0.001
Constant	–4.373	1.450	0.003	0.113	0.076	0.135	–0.250	0.093	0.007
Deviance	244229			723			1291
Null deviance	327768			4893			57482
N	6113			6113			6113

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

Results for linear regression of ego's CES-D score, number of friends, and number of family members at current exam on previous CES-D score, number of friends, and number of family plus other covariates. Models were 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 the sum of squared deviance between predicted and observed values for the model and a null model with no covariates.