HOME > Methodology > Sample Description

The sample now includes 3 generations grandparents (G1), parents (G2), and grandchildren (G3) and consists of families at high or low risk for depression based on the depression status of the original sample (G1). [Summary Table of Participants by Generation and Wave]

The G1 sample derives from the Yale Family Study of Depression (Weissman et al, 1984). All G1 were adults, white nonhispanic, and group matched by age and sex. One or more of the parents from the depressed families had received treatment for depression at a Yale outpatient clinic between 1968 and 1977. The diagnostic criteria were derived from the Research Diagnostic Criteria (RDC) with the following modifications: a minimum of 4 weeks duration and impairment in a major social role. The depression could not be preceded by another major disorder (e.g., alcoholism, drug abuse). The non-depressed families derived from a community study in New Haven, Connecticut. They were interviewed on at least 3 occasions and reported no history of treatment or psychiatric illness (Weissman et al, 1984).

There were 104 probands that had children between the ages of 6-17. At Wave 1 91 (87.5%) agreed to participate. There were a total of 220 children, 125 children from the depressed group and 95 from the normal group. There were a total of 105 males and 115 females. The mean age was 17 years. The diagnostic assessment of the children was based on the Schedule for Affective Disorders and Schizophrenia for School-Aged Children, Epidemiologic version (KSADS-E). DSM-III definite criteria was used. The parent were assessed using the Schedule for Affective Disorders and Schizophrenia, Lifetime Version (SADS-L). For a detailed description of methods and results from Wave 1 see Weissman et al, 1987.

Two years later all 91 families were recontacted and 85 (93%) agreed to participate in Wave 2. At Wave 2 all offspring were interviewed regardless of age. A total of 217 offspring were interviewed. The mean age was 20.6 years, 122 were females and 95 were males. The 91 families who had been originally classified by the status of the proband e.g., whether the proband was depressed or normal, were restratified as having at least one parent depressed or having neither parent depressed at Wave 2, to reflect the fact that some of the normal (never psychiatrically ill) probands as well as never psychiatrically ill spouses developed major depression over the follow up period (n=9). There are 65 families with 151 offspring who have 1 or more parent with depression and 26 families with 66 offspring where neither parent is depressed. The assessment measures used were the same as at Wave 1. For detailed information on stability of diagnoses over the 2 years between Wave 1 and 2 see Fendrich et al, 1990.

In Wave 3, 10 years after the initial assessment, all families were recontacted for a reassessment. G1 and G2 were clinically reassessed and G3 were assessed for the first time. In addition, in families with G3, the spouses of G2 (married-in) were assessed. Of the 220 G2 interviewed at Wave 1, 3 died and 1 was found to have Down’s Syndrome. Eighty-four percent (182/217) of G2 interviewed at Wave 1 were reinterviewed, 42 additional G2 were interviewed. Of the 182, 129 came from families with G1 depression and 53 from families without G1 depression, 100 were female and 82 male, and the mean age was 27.7 years. Eighty-four percent (73/87) of probands and 85% (52/61) of the spouses of the probands were reinterviewed (Weissman et al., 1997). At Wave 3 there were a total of 175 G3 (grandchildren). To be eligible G3 had to be older than 5 years of age and living in the geographic area or 18 years of age or older for a phone interview. Of the 175, 119 were eligible and 90 (76%) were interviewed. The mean age was 10 years, 44 were female and 46 were male. Seventeen came from families where G1 and G2 did not have depression, 12 from families where only G2 had depression, 18 from families where only G1 had depression and 39 from families where both had depression (Warner et al, 1999). The KSADS-E and SADS-L were modified to collect DSM-III-R criteria.

In Wave 4, approximately 20 years after the initial assessment, all families were recontacted for a reassessment. Probands and spouses of probands were not interviewed. Startle response and EEG testing was conducted as well as clinical reassessments. There were 2 more G2 deaths leaving 215 G2 from the original sample. Seventy percent (151/215) of the original sample was reinterviewed. Forty-seven additional G2 were interviewed. One hundred and ten participated in the startle response testing and 108 had usable startle response data (Grillon et al, 2005). One hundred and eleven participated in the EEG testing 87 were right-handed and had usable data (Bruder et al., 2005). For the sample of 198, the mean age was 36 years, 81 were male, 117 were female, 61 were from families without depression, and 137 were from families with G1 depression (Weissman et al., 2006). There were a total of 188 G3 available for inclusion in Wave 4, 156 (83%) were interviewed and the interview data for 15 G3 interviewed at Wave 3 was added to Wave 4. Seventy-five participated in startle response testing and 70 had usable startle response data (Grillon et al, 2005). Seventy-four participated in EEG testing and 49 were usable and right handed (Bruder et al, 2007). For the sample of 161, the mean age was 12.4 years, 86 were female, and 75 were male. Thirty-five came from families without G1 or G2 depression, 25 came from families with G2 depression only, 30 came from families with G1 depression only, and 71 came from families with both G1 and G2 depression (Weissman et al., 2005). The KSADS-E and SADS-L were modified to collect DSM-IV criteria.

In Wave 5, which ended in 2007, a subset of the original sample was targeted for clinical reassessment and MRI studies. Thirty-one G1 were interviewed at Wave 5, 20 probands and 11 spouses of probands. Twenty-nine usable structural and 27 usable functional scans were collected. Seventy-two of original 220 G2 who had participated at Wave 1 were clinically assessed at Wave 5. Nineteen additional G2 were clinically assessed. Ninety-three usable structural and 84 usable functional scans were collected. Sixty-one of the G3 who participated at Wave 3 or Wave 4 participated, 17 additional G3 participated. Seventy-eight usable structural and 69 usable functional scans were collected. Twenty-three married-ins (spouses of G2) were interviewed, 23 had usable structural and usable functional scans. There were a total of 216 usable functional or structural scans, one married in passed away, leaving 215 usable scans. One hundred and twenty-five were female and 90 were male. The mean age in 2007 was 38.0. Ninety-four were from families without G1 depression and 121 from families with G1 MDD. For additional details on the MRI studies see Peterson et al, in press.