Discussion

When Charles Dickens published his novel ‘A Tale of Two Cities’,14 he described the social stratification in 19th century England. He noted that living conditions for people could be completely different although they were living in two relatively comparable cities. In this study, we have compared the frequency of cardiovascular outcomes and risk factors for Scottish and Swedish middle-aged women living in comparable but nonetheless different social environments. To summarise, the overall picture of cardiovascular risks for Scottish women is, by almost all measurements, substantially worse than for Swedish women of the same age and education.

There were remarkable lifestyle differences between the Swedish and the Scottish women. Many more middle-aged women in Scotland than in Sweden drank alcohol frequently, had a higher frequency of smoking, were obese, reported a low vegetable consumption and generally exposed a more sedentary lifestyle.

There are many possible explanations why Scottish women tend to have a higher alcohol consumption than Swedish women. With the exception of Denmark, there is a Nordic model of alcohol control—most notably in Sweden—characterised by high taxation and restrictions on alcohol sales.15 This has been linked to lower levels of consumption, lower levels of liver cirrhosis mortality and other alcohol-related mortality and social problems due to alcohol. In contrast, Scotland (and the UK) has historically had a much more permissive approach to alcohol and a ‘wet’ drinking culture, whereby consumption is integral to everyday life and complete abstinence is atypical.16 A range of policies has been introduced in recent years in an attempt to reduce alcohol-related harm in Scotland, but the impact of such interventions will take time to manifest in the population.17 Recent estimates show increasing sales of alcohol in Scotland in supermarkets and off-licences, where prices have been stable, and a decline in sales in licensed premises, where prices have been increasing.18 Hence, alcohol control policies are one example of a policy area in which the Nordic experience is of obvious interest.19

The difference in obesity rates is also striking. Self-reported BMI data, as collected in the Swedish survey, has a tendency to underestimate obesity prevalence. However, the reported obesity prevalence for Swedish women in this study (15%) matches that recorded for all Swedish adults in 2016.20 The proportion of women with a diagnosis of high blood pressure was substantially higher for Scottish women. We cannot tell from this data whether they actually have higher rates of high blood pressure or whether Scottish women are simply diagnosed more often by their general practitioner (GP). During the time period covered by this study, GPs in Scotland were financially incentivised to identify cases of hypertension. No such system exists in Sweden, which might partly explain the differences.21

Although severe CVDs are quite rare in middle-aged women, we found a fourfold higher prevalence of doctor-diagnosed angina pectoris, myocardial infarction and ‘CHD’ (both combined) among Scottish women compared with Swedish women. More women in Scotland also reported worse psychosocial health and a lower grade of general health. In sum, although we adjusted for education level attained, the sociodemographic variable usually most strongly associated with social class differences in chronic disease rates, virtually all the measures of CVD risk factors and outcomes in our dataset showed much worse profiles in middle-aged Scottish women than in their Swedish counterparts.

While some might argue that genetic differences may contribute to this striking picture, modern population-genetic studies have revealed that many Scots have significant Nordic origins. Even Scots from other ethnic backgrounds based historically in the UK (over 95% of the current population of Scotland), including its invading peoples over the last few millennia, are not regarded as very genetically different from Swedes—compared with, for example, European ethnic groups with markedly different languages, such as the Basques.22 In our view, these profound differences in CVD risk and outcome frequency in the two populations are much more likely to have arisen from differences in the social, cultural, political and economic environments in which the two groups of women live. The worse health status and worse health inequalities by social class of Scotland overall (compared with most of Western Europe) are well documented.23 24

Studies have shown that large income differences are damaging to health and have negative social consequences. Income inequality is increasing in many western countries, but it is certainly greater in Scotland than in Sweden. It is important to take into account that in order to improve health, long-term work to reduce inequalities is needed.25 The role of the social, economic and cultural factors is more evident when determining such a frequent disease as CVD, especially when it comes to women. The societal development has led to increased labour force participation of women and thereby changes in their economic role. This development cannot be overlooked when describing the epidemiology of CVD and its risk factor in European women.5

The proportion of women in the labour force is generally higher among Swedish women than the Scottish, which is also evident in this study. Family policy legislation, particularly dual-earner family support, seems to have become of importance for cross-national differences in infant mortality in present-day Organisation for Economic Co-operation and Development countries, while GDP seems to have become less important. These family policies are particularly developed in the Nordic countries.19

The Nordic welfare states are distinguished by their emphasis on universal social policies rather than a reliance on targeted, selective and means-tested policies. The Nordic countries have been successful in reducing poverty and fostering equality of opportunity as well as equality of outcomes, with regard to class, income and gender. By providing many citizens with welfare resources through welfare state institutions, universal social policies are also likely to affect public health. In general, command over the resources by which we can control and consciously direct our conditions of life is of vital importance to health. These resources include both the material and the intangible.19 26 People with lower material and intangible resources tend to lose the motivation to take care of their own health. These factors could contribute to the reasons why we see higher levels of obesity, alcohol consumption and sedentary behaviour in women in Scotland. There are important regulatory differences as well—Sweden has much stricter alcohol control policies (both price and supply), stricter regulation of the food environment and the transport infrastructure gives more opportunities for walking and cycling and less reliance on cars.

Scottish women reported higher levels of psychological distress than Swedish women. This difference is likely to be related to the impact of inequality on the quality of social relationships and the scale of status differentiation in the two societies. Studies have shown that greater income inequality in rich societies is associated with higher levels of mental illness and drug misuse.27

This study has some strengths and limitations. One strength is that validated instruments and questions have been used. Many of the questions in the Swedish questionnaire were replicated from the SHeS Questionnaire, which improves the comparison between data from the two countries. Besides self-reports, we also had access to diagnoses set by doctors for the Swedish data, while the Scottish data were based on self-reported diagnoses. This might somehow impair the comparability of the estimates. However, the self-reported Scottish diagnoses are likely to have been underestimates of the true prevalence of conditions, so the difference observed between the two countries is likely to have been under, rather than over, estimated. As the incidence of cardiovascular diagnosis in general is relatively low for middle-aged women, the absolute number of affected women in the Swedish data must be considered in the comparisons. As with most surveys, there might have been a problem reaching those in the population most seriously affected, such as the unhealthiest and also those with language difficulties. However, this is likely to have been a problem in both countries.

Yet, this is one of the first studies to clearly show that Scottish middle-aged women are—relative to Swedish women—particularly affected by a worse profile of CVD risks, even after adjustment for education. The good news is that virtually all of the cardiovascular risk factors documented here (height excepted, although this can be changed over generations) as being of higher levels in Scottish women are—unlike genes—completely preventable and/or treatable. There is thus much scope for improving the health of Scottish women in this age group. For example, this could be done via better identification and management of traditional ‘Framingham’ risk factors for CVD, via regulatory actions to target the price and supply of alcohol and unhealthy food and via policies designed to address the structural causes of health inequalities.

In conclusion, the most likely reason why we see these different pictures of health in each country seems to have cultural, historical, economic and political origin. In future research, using an anthropological design to investigate how these two cultures, with their specific economy and political environments affect people's ways of thinking and acting could be interesting.