#PHTwitJC 6 discusses Cause specific mortality, social position, & obesity among women who had never smoked: 28 year cohort study, by Carole Hart, Laurence Gruer & Graham Watt, published in BMJ June 2011.
Join us on Twitter at 8.00 GMT, Sunday 13th November, using the hashtag #PHTwitJC. All are welcome – no expertise in research methodology is required, and ‘stupid’ questions are positively encouraged.
The paper analyses data relating to 3613 non-smoking women who were recruited to a larger cohort study in the 1970s. The Paisley and Renfrew Midspan study recruited more than 15000 men and women aged 45-64 (80% of the eligible population) in two Scottish towns in the years 1972-6.
Follow-up data analysis found that participants identified as ‘never-smokers’ had much better survival rates than smokers, and that was particularly true for women. However, there was also a difference in survival rates between never-smokers according to occupational class. At the 28-year follow-up point,
age adjusted survival rates for women who had never smoked were 65% in the highest occupational class and 56% for those in the lowest.
Higher rates of smoking among those in the lower occupational classes cannot, therefore, explain all of the class-based disparities in mortality. The further analysis reported in this paper aimed to explore in more depth the relationships between causes of death and social class position in women who never smoked.
Background: health inequalities, class, smoking, and obesity
The UK demonstrates significant health inequalities that are patterned by social class and geography. Official reports over the past thirty years: the Black Report (1980), Acheson Report ( 1998), Marmot Review (2010 -relating to England only) and Scotland’s Ministerial Taskforce report (2008) have demonstrated consistent patterns, with the ‘gap’ widening over time. These reports are supported by a wealth of academic research into health inequalitites, some of which is cited in this paper. Broadly, the pattern is for worse health outcomes (higher mortality, lower life expectancy) for those in the lower occupational class groupings; those who live in the most deprived neighbourhoods, and in households with the lowest incomes. The Marmot Review further underlined that there is a clear ‘gradient’ in health.
In the UK, smoking status is a health indicator that is (and has increasingly become) patterned by social class, with higher smoking rates in lower class groups. Research cited in the paper has indicated that smoking explains much excess mortality between social groups in the UK. This paper examines a large cohort of women who never smoked in order to identify further factors that might explain disparities in mortality between higher and lower social class groups.
Obesity: overweight and obesity are explored in the paper as further factors that might explain social class-related disparities in mortality between never-smokers. Obesity measured by BMI has been increasing for all population groups over recent decades, but this is something that again displays a social gradient (see e.g. p. 21 of this report for evidence from Scotland). Women from lower soci-economic groups/who live in more deprived neighbourhoods are more likely than those in the higher groups, to be obese.
The paper does not discuss the location of the study at any length; those unfamiliar with UK geography may appreciate some further context. Paisley and Renfrew are neighbouring towns in Renfrewshire, in proximity to Glasgow, which is the largest city in Scotland. As this teaching resource shows, this region displays some of the worst health indicators in the UK. If we consider women only (bottom right quadrant), six of the ten local authorities with the lowest life expectancy in the UK (out of423) are located in the West-Central region of Scotland. Renfrewshire itself is seventh lowest (2006 data). Over the past few decades, the area has experienced significant industrial decline, as employment in shipbuilding, textiles and other industries dwindled.
Methods & findings
On recruitment to the study, all participants completed questionnaires and underwent screening covering a range of health indicators, including lung function, body mass index, smoking status, blood pressure, cholesterol and heart function. They were categorised by occupational class (using the Registrar General’s social class categories – brief overview here) . These data were statistically analysed retrospectively along with known data at follow-up point on mortality and cause of death.
Main findings relating to the ‘never smoker’ women :
a) those in lower social class groups demonstrated worse health indicators overall than those in the higher class groups;
b) obesity and overweight among never-smokers were patterned by social class/position;
c) women who had never smoked were more likely than smokers of the same social class to be overweight or obese;
d) mortality was higher overall in the lower social class groups, although not for cancers;
e) however, for ‘normal weight’ women there were not significant differences in mortality between social class groups.
The authors assert that this study adds new knowledge about the ‘true impact’ of obesity on mortality, by isolating women who had never smoked within a large cohort.
Except for the women in the highest social positions, the differences in mortality rates between non-obese and severely obese women of the same social position were much greater than those between normal or overweight women in different social positions.
As obesity rates have been increasing significantly, and smoking rates declining, since the measuring point in the 1970s, the authors argue that this finding is potentially very important for understanding the patterning of health status now:
If, as in this cohort, obesity is socially patterned, it may add to health inequalities and disproportionately increase the pressure on health and social services serving more disadvantaged populations. On a more positive note, the women in the whole cohort who did not smoke and were not obese had the lowest mortality rates, regardless of their social position. The preventive message is clear.
Broadly based around this appraisal toolkit for cohort studies.
- Were the aims of this study clear?
- Were methods of recruitment to the cohort study, and selection of the sub sample for follow up analysis, clearly explained?
- Is the statistical analysis valid? Were outcomes measured appropriately; have the authors taken into consideration all possible confounding factors?
- How do the results of the analysis fit with other available evidence (within and outside the UK)?
- What are the implications of the study findings for a) research into obesity and health inequalities; b) public health policy & practice?