Dr Nadine Smith: Mental health: A longitudinal analysis of the self-reported mental health of Australian women.

Aim: The primary aim of this thesis is to determine the nature of the relationships between mental health, physical health and social disadvantage including cross-sectional associates and longitudinal predictors of mental health, physical health and social disadvantage. A secondary aim is to explore psychometric issues related to the measurement and analysis of mental health, including psychometric evaluation of self-report mental health measures and the impact of mental health and recall on self-reported events.
Study sample: This research analyses data collected for the Australian Longitudinal Study on Women’s Health (ALSWH) from three cohorts of women who were aged 18-23 years (the Younger cohort, N=14779), 45-50 years (the Mid-age cohort, N=14099) and 70-75 years (the Older cohort, N=12940) when first surveyed in 1996. The ALSWH sent postal questionnaires containing about 300-500 items to each woman surveyed. For this research, the ALSWH variables analysed include multiple measures of mental health, physical health, health behaviours, positive psychological characteristics, life events, social support and socio-demographics.

Findings: Considerable change in depression across time was found. The cross-sectional and longitudinal analyses were consistent in showing that depressed women were more likely than others not to be partnered, to have difficulty managing on their income, to be inactive, to smoke, to report more symptoms of physical illness, and to recall more GP visits. Furthermore, declining mental health was associated with going from being partnered to not being partnered and from being an acceptable weight to being overweight. Increasing mental health was associated with going from being inactive to being active.
The distinctive features of anxiety and depression were found to be less pronounced than expected amongst the older cohort. The different mental health indicators examined were found to all have approximately equivalent discriminating abilities for detecting depression and anxiety outcomes.

The methodological challenges in using life events lists (and more generally any self-reported data) were emphasised by the empirical evidence that mental health and errors of recall impacted on the reporting of life events. Women with low mental health scores reported life events more frequently than women with high mental health scores, especially for “perceived” rather than “factual” life events. Furthermore, remote life events were reported to have occurred more recently suggesting that recall of life events was subject to telescoping.



Conclusions: The findings of this thesis draw attention to the complex pattern of interaction between mental health status, social disadvantage, physical ill-health and unhealthy lifestyles. This study also highlights the methodological issues related to the measurement and analysis of self-reported data, particularly for research on mental health