Scottish Health Survey - topic report: mental health and wellbeing

Summary of results

This report explores factors associated with mental wellbeing and mental health among adults in Scotland using data from the Scottish Health Survey (SHeS). Analyses are based on survey years 2012 to 2013 and include participants aged 16 years and older.

The factors included in the analyses include socio-demographic, health behaviours and conditions. Results are presented for:

• The Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS) - used to measure mental wellbeing
• The General Health Questionnaire (GHQ) - used to identify individuals showing signs of the presence of a possible psychiatric disorder.

The mean score on the WEMWBS scale in 2012/2013 was 50.3 for men and 49.6 for women. Mental wellbeing is significantly associated with age, with mean scores high in the youngest adult age groups (50.1 and 50.2 for ages 16-24 and 25-34 respectively), dropping among adults aged 45-54 (48.9), rising to a peak between ages 65 and 74 (51.1), and then dropping off again among the oldest group (ages 75+, mean score 49.5).

One in eight (13%) men have GHQ12 scores of four or higher, indicating the presence of a possible psychiatric disorder, compared to 18% of women. A lower proportion of men across all age groups displayed signs of a possible psychiatric disorder.

GHQ12 and WEMWBS show a moderate negative correlation. The median WEMWBS score declines as GHQ12 score increases, most rapidly nearest the two extremes on the GHQ12 scale.

Chapters 4 and 5 present an analysis of factors significantly associated with low mental wellbeing among adults, indicated by WEMWBS scores, and an analysis of factors associated with adults who display signs of the presence of a possible psychiatric disorder, indicated by scores of four or higher on the GHQ12 scale. Logistic regression of 2012/2013 SHeS data provides a robust analysis examining the factors associated with these mental health and wellbeing outcomes across the adult population. By controlling for various independent factors simultaneously, the association of each factor with mental health and wellbeing can be established. A benefit of these analyses is being able to disentangle confounding factors. For example, this allows us to test whether lower levels of wellbeing observed among people who provide 35 hours or more unpaid care per week is due to the age profile of this subgroup. Other results, standardised by age, show the proportion of adults with a GHQ12 score of four or higher, and the mean WEMWBS score of each population subgroup.

It is important to note that the factors examined in the multivariate models in this report are likely to have bi-directional relationships with low mental health and wellbeing. Therefore, while many of these findings support other research which shows a relationship between demographic or health-related factors and mental wellbeing, the associations identified in this analysis show correlations between variables, as opposed to indicating causality. Furthermore, the results are limited to those factors which are reported in SHeS.

The factor most strongly associated with low mental health and wellbeing, under both measures, is economic activity, in particular the group of adults who are permanently unable to work. Although these results may partially reflect the effect of unemployment on mental wellbeing, as cited by other research, the results may be confounded by those who are unable to take up work as a result of poor mental health. However, these associations are still useful to identify population subgroups at greatest risk of poor mental wellbeing.

Many of the health-related factors for which the results indicate an association with poor mental wellbeing - for example, smoking and alcohol dependence - are socio-demographically patterned whereby prevalence is generally higher in lower socio-economic groups (for example, the most deprived areas or lower income households). Therefore, although area deprivation (SIMD) is shown not to be a significant predictor of GHQ12 scores of four or higher after controlling for other factors, it is still true that prevalence of some of those most significant risk factors for low mental wellbeing is highest in those areas.

Physical inactivity, smoking and possible dependence on alcohol are all significantly associated with low WEMWBS scores and GHQ12 scores of four or higher, after controlling for other factors. For both measures, physical inactivity and possible dependence on alcohol were the strongest behavioural predictors.

Adults who provide unpaid care for 35 or more hours per week are more likely to have a low WEMWBS score, and a GHQ12 score of four or higher, after controlling for other factors. These results were stronger among women than men. In particular, female carers who provide 35 hours or more support per week were significantly more likely to respond negatively when asked if they have 'been feeling relaxed'.

A number of socio-demographic factors are significant predictors of poor mental wellbeing - for example, age, household income and marital status. While area deprivation was a significant predictor for WEMWBS, it was one of the weakest of those selected factors and not significant for GHQ12.