Female vulnerability to depression

Evidence from several sources indicates that there is an increased risk of neuroses among women. This evidence derives from psychiatric in-patient samples (Cochrane & Stopes-Roe, 1981), from general practice patients (Shepherd et al. 1966) and, most important, from community studies (Bebbington et al. 1981). Weissman & Klerman (1977) have reviewed the evidence and argue that social factors contribute most to the sex difference in morbidity. Over the years a variety of social risk factors have been examined. These include, among others, maternal loss in childhood (Birtchnell, 1970), marital problems (Birtchnell & Kennard, 1983), the stress of child rearing (Richman, 1976), and lack of employment opportunities (Warr & Parry, 1982a). The evidence for some of these factors, such as parental loss (Tennant et al. 1981) or lack of employment (Warr & Parry, 19826), is far from conclusive. Indeed, it was on the basis that the above factors in their own right did not seem to correlate significantly with depression in women that Brown & Harris (1978) invoked their 'vulnerability' model. They found that four 'vulnerability' factors (identified as 'maternal loss before the age of 15 years', 'three or more children under the age of 14 years', 'lack of marital intimacy' and 'lack of employment outside the home') contributed to depressive illness only when combined with a stressful life event ('provoking agent'). They argued that a synergistic (interaction) effect on the risk of depression was exerted by the combination of a 'vulnerability' factor and 'provoking' life event. The findings of their study have been discussed at some length in the literature (Tennant & Bebbington, 1978; Bebbington et al. 1981; Brown & Harris, 1980). The synergy identified by Brown & Harris was, however, 'model' dependent, since the use of an alternative statistical model (i.e. multiplicative) failed to confirm the interaction (Tennant & Bebbington, 1978). Brown & Harris' additive model is less frequently preferred by statisticians (Everitt & Smith, 1979; Costello, 1982). Replications of Brown & Harris' findings have now been attempted in a variety of studies. These include studies of social risk factors in hospital psychiatric patients, where the synergistic model is not assessed (Roy, 1978), to community studies of social risk factors, where synergy is assessed (Warr & Parry, 1982a; Surtees et al. 1983). Others are perhaps more valid because they are community studies and use the same vulnerability factors as Brown and his colleagues (Costello, 1982; Solomon & Bromet, 1982; Campbell et al. 1983; Cooke, 1981; Brown & Prudo, 1981; Bebbington et al. 1984). In the first of these 'replications' Roy (1978) found that 'loss of mother before the age of 11 years', 'three or more children at home under 14 years of age' and a 'lack of full time employment' were associated with depression in working-class women patients. For middle-class women patients only the lack of a confiding relationship was associated with depression. This study was a poor replication, since it was based on a depressed out-patient sample rather than community subjects. Furthermore, 'case' criteria and the definition of some variables were quite different from those of Brown and his colleagues. Perhaps most important, Roy did not test the vulnerability model (synergy) since life event stress had not been assessed. Two community studies failed to demonstrate findings relating to a synergistic or vulnerability model for depression. In the first of these Warr & Parry (19826) assessed risk factors for depression (using the Zung depression inventory) in working-class women with young children. The risk factors included employment status, attitude to home role, strain between home and work roles, social support, attitude to employment, and child care demands. The first 4 of these 6 factors were associated with depression. When, however, employment status was assessed in conjunction with each of the other risk factors, no interaction was found; each variable operated as an independent risk factor for depression. In the second study Surtees et al. (1983) studied Edinburgh women and