DEPRESSION HAS HIGH prevalence and recurrence rates relative to other mental diseases, and its incidence in women is twice that in men, with that of menopausal women being even higher (Kessler 2003 and Shors and Leuner 2003). Hay et al. (1994) reported that the probability of experiencing depression within 4 years of menopause is 35%, and Sagsoz et al. (2001) reported that menopausal women who experienced changes in menstruation commonly experience mental disorders such as depression and anxiety. Maartens et al. (2002) also reported that psychotic symptoms such as depression, anger, somatization, and anxiety are frequent in menopausal women. Moreover, many clinical and epidemiological studies have found that the incidence rate of depression is higher in menopausal women, although the underlying reason is still unclear (Sagsoz et al., 2001). Researchers who ascribe menopausal depression to physiological mechanisms have presented a hypothesis that estrogen deficiency during menopause is the primary reason for menopausal depression, and they have attempted to show a causal relationship by showing the efficiency of estrogen supplement therapy in treating depression. However, the studies performed within the past 20 years have found no significant relationship between estrogen and menopausal depression, and that estrogen has no direct antidepressive effect (Birkhauser, 2002). Therefore, recent studies have focused on psychosocial factors rather than biological factors (Takamatsu et al., 2004), although none of these studies have systematically investigated these factors. Moreover, factors evaluated as significant in some studies have been found to be insignificant or conflicting in other studies. This is attributable to the factors of menopausal depression not being covered comprehensively, with only the relationship between a few factors being investigated or their independent influences being verified. These conflicts have resulted in controversies about the development of menopausal depression intervention programs. These problems could be resolved by constructing and verifying a theoretical model that includes the cause-and-effect relationship or interactions between several factors related to menopausal depression. This could provide a useful theoretical guideline for developing a program for the prevention, early detection, and curing of menopausal depression.

Development process for a theoretical model

In this study, significant factors for predicting menopausal depression and their relationships were constructed based on previous studies, and these are illustrated in Figure 1. It was hypothesized that educational and economic statuses not only influence menopausal depression directly, but also indirectly via other parameters such as attitudes towards menopause and aging, social support, and social conflicts. It was also hypothesized that attitudes towards menopause and aging directly influence menopausal depression. The empirical support for each of these paths is presented.

The hypothetical model.

In many previous studies, educational and economic statuses were presented as specific indices for individuals predisposed to depression, and they were presented as demographic variables with a considerable influence on depression (Bloch 2002; Cho et al 1999; Kessler 2003; Kim 2000; Nam and Cho 1997 and Shors and Leuner 2003). Especially, in the variable of economic status, almost all previous studies reported that middle-aged women with low income showed a higher degree of depression compared with those with higher income.

Previous studies have shown that the perceived health status of middle-aged women is directly related to menopausal depression (Choi 1998 and Kim 2000). There was also a close relationship shown between attitudes towards menopause and aging and menopausal depression (Chi 1983; Choi 1998; Kaufert et al 1998; Lopez and Silva 1999 and Rita et al 2002). Most studies have shown that middle-aged menopausal women have a negative attitude towards menopause, thinking that menopause would accelerate aging and cause problems in sexual life, and aggravate their health status.

Positive support from members of social networks, such as husband, children, relatives, and friends, was significantly related with protection from depression (Ham et al 2002; Han 1997; Major et al 1997; Vinokur and van Ryn 1993 and Yoo 2000). Especially, emotional and informational support from friends who are also suffering from middle-age crises had a positive effect on alleviating the negative effects of menopausal stress on depression (Park, 1999).

Previous studies on menopausal depression focused only on the positive function of social relationships, which is a social support, and paid little attention to the negative effects of social network. However, recently it was reported that the social network could be a positive resource like social support, and at the same time could cause tension and stress (Ham et al 2002; Han 1997; Major et al 1997; Park 1999; Rauktis et al 1995 and Yoo 2000). Moreover, when we consider the results of studies (Ham et al 2002; Park 1999 and Yoo 2000) showing that even when frequencies of social conflicts are lower than those of social support, the experiences of the former had a greater effect on menopausal depression, indicating that studying both the positive and negative aspects of the social network is necessary. From the above considerations, it was hypothesized in this study that a poor perception of health status and lower educational and financial statuses were related to little positive support from members of the social network, and that more conflict with members of the social network and negative attitudes toward menopause and aging would be related to more menopausal depression.

Methods

Design

This study is a cross-sectional structural model-verification study in which a hypothetical model predicting menopausal depression based on previous studies was constructed, and the appropriateness of the model was verified by collected data.

Sampling

In this study, the data were collected by distributing questionnaires to middle aged women who conveniently selected in two Korean cities. From the distributed 500 questionnaires, 401 were collected (response rate 80.2%). Among these, the data from 305 subjects who belong to the aged 45 to 55 years (the typical range of menopausal ages) were used for ultimate analysis.