Scottish Health Survey - topic report: The Glasgow Effect

3. Physical Health

SUMMARY

Heart attack

• The factors which were found to be significantly associated with doctor-diagnosed heart attack were: age, sex, residence, household tenure, marital status, smoking status, abstaining from alcohol, drinking over the recommended weekly alcohol limit, physical activity level, BMI and being a stepparent.
• Even after controlling for all of these factors, residents of Greater Glasgow and Clyde were nearly one-and-a-half times as likely to have had a doctor-diagnosed heart attack (42% increased risk compared to the rest of Scotland).

Limiting longstanding illness

• Although residence in Greater Glasgow and Clyde was associated with increased risk of limiting longstanding illness when age, sex and SIMD were adjusted for, the difference was explained with further adjustment for socio-economic variables, specifically receiving income-related benefits, economic activity, household tenure, equivalised income, educational qualifications and marital status.

Stroke

• There was no significant difference in risk of stroke between residents of Greater Glasgow and Clyde and the rest of Scotland, either before or after adjusting for age and sex.

CVD

• There was a significantly higher risk of CVD for residents of Greater Glasgow and Clyde compared to the rest of Scotland after adjusting for age and sex, however this was explained by further adjusting for SIMD.

COPD

• Despite adjusting for age and sex, the risk of COPD was 44% higher among residents of Greater Glasgow and Clyde compared to those living in other parts of Scotland; however, this was fully explained by the addition of SIMD to the model.

3.1 Introduction

Participants were classified as having had a heart attack, stroke or chronic obstructive pulmonary disease ( COPD) if they reported that it was doctor-diagnosed. Participants were asked about longstanding illness and whether it limited their activities. CVD was classified based on a series of questions on whether participants had suffered from any of the following conditions: angina, heart attack, stroke, heart murmur, irregular heart rhythm and 'other heart trouble'.

3.2 Heart attack

3.3% of participants had suffered a doctor-diagnosed heart attack, with a significantly higher prevalence for men (4.3% vs. 2.3%). Prevalence increased sharply with age, with less than 1% of those aged 16-54 having suffered a heart attack compared with 13% of those aged 75 and over. There was also a significant difference by SIMD quintile, with prevalence of 2.2% in the most deprived quintile compared to 4.1% in the least deprived quintile. The difference in prevalence between residents of Greater Glasgow and Clyde and the rest of Scotland was small but significant (3.9% vs. 3.1%).

The model development process can be found in Appendix 4, along with McFadden's pseudo R ²s for the different models. In view of these, the best fitting model was chosen, and the results for that model using the full data available are described here. Nurse and blood variables could not be included in the analysis due to the low prevalence of heart attacks, and the reduced sample sizes resulting from the addition of variables collected in the nurse subsample.

The factors which were found to be significantly associated with doctor diagnosed heart attack were: age, sex, residence, household tenure, marital status, smoking status, abstaining from alcohol, drinking over the recommended weekly alcohol limit, physical activity level, BMI and being a stepparent. Using all available data the weighted sample size was 11,685. The odds ratio for residents of Greater Glasgow and Clyde was 1.42, showing that even after adjusting for socio-economic, behavioural, biological, relationship and social mobility variables, residents of Greater Glasgow and Clyde were nearly one and a half times as likely to have had a doctor diagnosed heart attack. Being overweight and obese were also significantly associated with doctor-diagnosed heart attack (odds ratios of 1.71 and 2.44 respectively).

Females were less likely than men to have suffered a doctor-diagnosed heart attack (odds ratio of 0.37), and risk increased with increasing age, as would be expected. Private renters had significantly lower odds than owner occupiers (odds ratio of 0.53), whereas social renters had significantly higher odds than owner occupiers (odds ratio of 1.37). People who had never been married or in a civil partnership had lower odds of having had a heart attack than those who were married/in a civil partnership and living together (odds ratio of 0.47), and those who were widowed or surviving civil partners were 1.55 times as likely to have had a heart attack. Compared to people who had never smoked or were ex-occasional smokers, ex-regular smokers, current moderate smokers and current heavy smokers all had significantly higher odds, with odds ratios of 2.63, 2.36 and 2.44 respectively.

Those who drank over the recommended weekly alcohol limit had lower odds of having had a heart attack (odds ratio of 0.68), whilst those who abstained from alcohol had higher odds compared to those who drank (odds ratio of 1.77). It should be noted that drinking behaviour measured in the survey relates to current and not previous behaviour. Those who have had a heart attack may have been advised to stop drinking by a doctor. Having a medium or high level of physical activity was associated with lower odds of having had a heart attack compared to having a low level of physical activity (odds ratios of 0.57 and 0.50 respectively). The odds ratio for being a stepparent was 0.05.

Table 3

3.3 Longstanding limiting illness

25% of participants reported a limiting longstanding illness, with significantly higher prevalence among women than men (27% vs. 23%). Prevalence increased significantly with age, from 9% of those age 16-24 to 54% of those age 75 and over. The most deprived SIMD quintile had almost double the prevalence of the least deprived quintile (35% vs. 18%). Residents of Greater Glasgow and Clyde had a significantly higher prevalence of limiting longstanding illness than the rest of Scotland (28% vs. 25%).

In the initial logistic regression model containing age, sex and residence, residents of Greater Glasgow and Clyde had an odds ratio of 1.33, meaning their odds of having a limiting longstanding illness were 33% higher than the rest of Scotland. When SIMD was added to the model all four variables were significant predictors with the odds ratio for Greater Glasgow and Clyde reduced to 1.17, showing that half of the additional odds of having a limiting longstanding illness was explained by SIMD. However there was still a 17% increase in odds of having a limiting longstanding illness for residents of Greater Glasgow and Clyde compared to the rest of Scotland.

When the group of socio-economic variables were added to the model and backward selection performed, residence was no longer significant at a 5% level of significance. The variables which remained in the model (with p<0.05) were age, sex, receiving income-related benefits, economic activity, household tenure, equivalised income quintiles, educational qualifications and marital status. When residence dropped from the model only these variables remained in the model, showing that these variables fully explain the effect previously seen of residence in Greater Glasgow and Clyde.

3.4 Stroke

2.5% of participants reported a doctor-diagnosed stroke, with no significant difference by sex. Prevalence increased with increasing age, from less than 1% of 16-54 year olds to 11% of those aged 75 and over. There was also a significant difference by SIMD quintile, with the prevalence in the most deprived quintile more than twice that of the least deprived quintile (3.7% vs. 1.7%). There was no significant difference in prevalence of stroke between residents of Greater Glasgow and Clyde and the rest of Scotland, nor any difference in odds of having had a doctor-diagnosed stroke after controlling for age and sex.

3.5 Cardiovascular disease ( CVD)

15% of adults had a doctor-diagnosed CVD condition, with no significant difference by sex. There was a significant relationship with increasing age, with a prevalence of 5% among 16-24 year olds, which rose to 31% of those age 65-74 and 41% of those age 75 and over. A significant difference was also found by SIMD quintile, with those living in the most deprived quintile more likely to have a doctor-diagnosed CVD condition than those in the least deprived quintile (17% vs. 12%). There was no significant difference for residents of Greater Glasgow and Clyde compared to the rest of Scotland.

Despite this, in the initial logistic regression model containing only age, sex and residence, residents in Greater Glasgow and Clyde had an odds ratio of 1.20, meaning that they had a 20% higher risk of having a CVD diagnosis compared to the rest of Scotland. Once the model also adjusted for SIMD the odds ratio dropped to 1.12, which was not significant at the 5% level of significance. SIMD was highly significant, with an increased risk of having a CVD diagnosis in more deprived areas. This shows that the difference in prevalence of doctor-diagnosed CVD between Greater Glasgow and Clyde and the rest of Scotland was explained by their different SIMD distributions.

3.6 Chronic obstructive pulmonary disease ( COPD)

3.5% of participants reported doctor-diagnosed COPD, with a significantly higher prevalence in women than men (3.8% vs. 3.1%). The prevalence increased with increasing age, from 1.5% among 16-54 year olds to 8.7% for those aged 75 and over. The prevalence also increased with increasing deprivation, with over three times the prevalence in the most deprived SIMD quintile than the least deprived SIMD quintile (6.0% vs. 1.9%). The prevalence among residents of Greater Glasgow and Clyde was significantly higher than in the rest of Scotland (4.2% vs. 3.2%).

In the initial logistic regression model containing only age, sex and residence, residents in Greater Glasgow and Clyde had an odds ratio of 1.44, meaning that they had 44% increased odds of having a COPD diagnosis compared to the rest of Scotland. However once the model also adjusted for SIMD the odds ratio dropped to 1.17, which was not significant at the 5% level of significance. SIMD was highly significant, with increasing odds of having a COPD diagnosis for increasing levels of deprivation. This shows that the difference in prevalence of doctor-diagnosed COPD between Greater Glasgow and Clyde and the rest of Scotland was explained by their different SIMD distributions.

3.7 Assessing the impact of the socio-economic variables individually

For the physical health outcomes the model containing economic activity as the only socio-economic variable provided the model which best fit the data (as it had the highest McFadden's pseudo R ²), indicating that this measure was the best predictor of physical health after adjusting for the rest of the variables in the final models. For heart attack it was also possible to investigate which socio-economic variable explained the most difference between Greater Glasgow and Clyde and the rest of Scotland, with household tenure explaining the most difference.

3.8 Conclusions and Discussion

Despite adjusting for socio-economic, behavioural, biological, relationship and social mobility variables, residents of Greater Glasgow and Clyde had 42% increased odds of having a doctor-diagnosed heart attack. It is important to note that this outcome only relates to survivors of heart attacks, as data are obviously not collected from people who have had fatal heart attacks. The odds ratio for residence in Greater Glasgow and Clyde decreased slightly between adjusting for age and sex, and adding SIMD to the model, but further adjusting for a range of socio-economic and behavioural variables did not further reduce the odds.

In the final model the odds ratio for residence in Greater Glasgow and Clyde was 1.42, showing a 42% higher risk of having had and survived a heart attack. As expected the lower odds were for people who had never smoked or were ex-occasional smokers, with increasing odds for current light smokers, current moderate smokers, current heavy smokers and ex-regular smokers. The increased risk for ex-smokers could be due to the effect of smoking on cardiovascular risk factors; although the levels decline after smoking cessation, it has been found to take five years for the levels to return to those of people who have never smoked ²². Another possibility is that people stop smoking after they have had a heart attack, as they are advised to do so.

Those people with low levels of physical activity, have around double the odds of those who have a medium or high level of physical activity. The level of physical activity may also be linked to having had a heart attack in both directions; people may be more limited in the physical activity they can carry out or may be scared to do a lot of physical activity, alternatively their levels of physical activity may increase due to recommendations from a health professional.

There are similar issues with interpreting the alcohol variables - abstainers have an odds ratio of 1.77 compared to those who drink alcohol, but it may be that people who have had a heart attack are advised to reduce their alcohol consumption. Another factor to bear in mind is that people who become abstainers may have drunk heavily earlier in their lives. However both lifelong abstainers and former drinkers have been found to have an increased prevalence of coronary heart disease ²³. For drinking over the weekly alcohol limit the odds were 0.68, which again is not what would instinctively be expected. It is also possible that people's BMI, drinking, smoking and physical activity levels affect their chance of surviving a heart attack.

The analyses described in this report using the 2008 and 2009 data found no difference in prevalence of doctor diagnosed CVD between residents of Greater Glasgow and Clyde and the rest of Scotland; however after adjusting for age and sex, residence in Greater Glasgow and Clyde was associated with higher odds of CVD. The effect was removed by adjusting for SIMD. In Gray's report ¹⁰ using the 1995, 1998 and 2003 data the results differed for men and women, with no effect for women in Greater Glasgow and a lower risk for men after adjusting for socio-economic factors.

No difference in prevalence was found for stroke in the 2008 and 2009 data, whereas using the 1995, 1998 and 2003 data ¹⁰ a significantly higher prevalence was found among women in Greater Glasgow, which was explained by the socio-economic factors. There was no difference in prevalence found between men in Greater Glasgow and the rest of Scotland.

Using the 2008 and 2009 data residents of Greater Glasgow and Clyde had significantly increased odds despite adjusting for age, sex and SIMD; however the effect was removed by further adjustment for socio-economic variables. Gray ¹⁰ investigated longstanding illness, which is different to limiting longstanding illness which was investigated in this report; however, the results are still of interest. For both men and women in Greater Glasgow a significant difference was found in longstanding illness after adjusting for age and survey year, with the difference becoming non-significant after adjusting for socio-economic factors, as with limiting longstanding illness in this report.