The impact of demographic and risk factor changes on coronary heart disease deaths in Beijing, 1999–2010

Coronary heart disease (CHD) mortality rates have halved in most industrialized countries since the 1980s [1]. However, most developing countries, including China, have experienced substantial increases in CHD mortality in recent decades [2]. CHD is thus projected to be the leading global cause of death and disability by 2020 [3].

A variety of CHD policy models have therefore been developed to help explain the observed changes in CHD mortality in terms of changes in cardiovascular risk factors and treatments [4‐6]. Analyses conducted in different countries have consistently suggested that reductions in cardiovascular risk factors account for larger CHD mortality effects than treatments [7‐14]. Using these models to explain future trends in CHD mortality in terms of risk factor trends, or demographic change, can be potentially helpful for policy makers. Such analyses can also help identify the most effective and cost-effective prevention strategies, especially for economically developing countries.

The IMPACT Model was developed by Capewell and colleagues and successfully used to help explain the observed declines in CHD mortality resulting from improving risk factor levels and cardiological treatments in many developed countries, including the UK, the USA, New Zealand and Finland [9‐13, 15], and to estimate the number of additional deaths that could be prevented or postponed due to further declines in risk factors [16]. Subsequently, the model was used to evaluate potential future options [9‐13, 17]. More recently, the IMPACT model has also been applied to populations where CHD mortality rates are increasing [18]. In Beijing between 1984 and 1999, the CHD mortality rates rose by 50% in men and 27% in women aged 35–74 years, and most of these increases could be attributed to substantial increases in total cholesterol levels (mean population level rose by 1.1 mmol/L) [18].

In this paper, we now extend this Beijing population model from 1999 to estimate the number of CHD deaths anticipated in Beijing in 2010 under three contrasting scenarios: a) assuming simple population ageing, b) assuming that recent adverse trends in risk factors continue to worsen, and c) more optimistically assuming 0.5% annual reductions in major cardiovascular risk factors, as already achieved in many other populations [17]. These estimates might provide a potentially useful reference to inform the planning of future CHD prevention and treatment strategies in Beijing.

This study analyzed current trends in risk factor levels and population ageing to estimate the number of CHD deaths expected in the Beijing population by 2010. Compared with 1999, approximately 1,990 additional deaths might be expected simply due to population ageing, representing a 27% increase. Most worryingly, if recent adverse trends in major risk factors also continue, Beijing will experience over 12,000 CHD deaths in 2010, representing a 67% increase. Alternatively, if modest reductions in major risk factors (0.5% per year) were achieved, as in many other countries, these reductions would decrease CHD deaths by approximately 23%, in spite of population ageing (Figure 2).

The impact of population ageing on CHD mortality was very obvious in this study. In 2010, a substantial increase in older people (about 1.4 million more citizens aged 35–74 years) is expected in Beijing, representing an increase of approximately 22%. Even if the risk factor profiles and treatments remain as good as they were in 1999, these demographic trends mean that the projected CHD deaths in 2010 will rise above 9,300, representing a 27% increase compared with 1999. The number of elderly women is projected to increase far more substantially than the number of elderly men, resulting in a particularly large mortality burden in women. This gender difference is partly explained by a differential mortality in men from competing causes at earlier ages (mainly due to their very high smoking prevalence) [32, 33]. Conversely, CHD deaths in the youngest age groups will decrease due to their decline in population size (attributable to a reduction in the birth rate since 1975) [34].

Although several studies have examined the decline in CHD mortality in Western countries [7, 9‐13], very few studies have been conducted in developing countries or in Asia [18, 35, 36]. Beijing has witnessed a dramatic increase in CHD mortality during recent decades, much of which can be attributed to the rising population cholesterol levels [18]. Furthermore, the incidence rate of CHD showed a corresponding upward trend in Beijing during 1987 to 1993 [20]. Similarly, the CHD epidemic in Singapore can be largely explained by higher cholesterol levels [35]. These trends are consistent with the observed changes from traditional to Western diet [37], with a higher consumption of dietary saturated fat and lower consumption of polyunsaturated fats, vegetables and carbohydrates [35]. If the recently observed adverse trends in risk factors levels continue in Beijing, there would be 3,000 more deaths annually by 2010 than otherwise expected based on demographic trends (population ageing) alone. Impressively, approximately three-quarters of this increase would reflect increases in total cholesterol levels. In the mid-1980s, cholesterol and obesity levels in Beijing were the lowest among all the MONICA populations [38]. However, the subsequent massive increases in cardiovascular risk factor levels, especially cholesterol, have resulted in many additional CHD deaths because cholesterol is a particularly powerful CHD risk factor [18, 38]. Intriguingly, in Japan, decreasing CHD mortality rates have been reported despite modest increases in the total cholesterol levels and similarly high smoking rates to Beijing. However, the high consumption of cardioprotective oily fish in Japan may have helped to offset the unfavourable effects of other cardiovascular risk factors [36]. In Beijing, the overall population diastolic blood pressure level increased slightly and made only a small contribution to the predicted change in CHD mortality. Improvements in hypertension treatment may have contributed [18, 39].

Modest reductions in major risk factors could hugely affect the number of future CHD deaths in Beijing, despite population ageing. Annual 0.5% reductions in major risk factors could result in some 3,730 fewer deaths in 2010, representing a 23% decrease compared with 1999. Effective strategies targeting modifiable risk factors could therefore partly offset the adverse impact of unmodifiable risk factors, the most powerful being age.

Small but powerful risk factor reductions have actually been observed elsewhere [17]. Similarly modest falls in the smoking rate, cholesterol level and blood pressure have all been achieved in other industrializing populations, such as Poland and the Czech Republic [31], and thus appear to be feasible in Beijing. Encouragingly, a citywide public smoking ban was enacted in Beijing in 1996, and a nationwide survey confirmed that the smoking rates in China have fallen in recent years [40]. The signing of the Framework Convention on Tobacco Control by the government of China will further promote reductions in smoking. However, reductions in obesity and diabetes appear more challenging to achieve, with adverse trends continuing in the USA and Western Europe [41]. Furthermore, even if the proposed modest reductions in smoking and blood pressure were actually achieved by 2010, the levels in Beijing would remain higher than in most other countries [12, 17, 38, 42]. There is clearly a need for further, more substantial reductions in smoking and blood pressure levels. Such declines would be expected to greatly reduce the burden of CHD [43].

This modeling work can provide useful information, promote evidence-based health policy making, and improve the effectiveness of prevention strategies. Our study results showed that there is substantial potential benefit from reductions in major cardiovascular risk factors by modifying lifestyle, even with the growing and ageing population. The achievable interventions targeting these major risk factors could produce substantial declines of coronary mortality, as already achieved elsewhere [44].

Like any forecasting tool, this study has a number of strengths including the use of a transparent modelling methodology previously replicated in a range of populations. It also contains a number of limitations. Firstly, it considers only mortality, but not years of life lost or morbidity. Life expectancy or life-years gained should be included in future versions of this model [43, 45, 46]. Secondly, potential interactions among risk factors and potential lag effects were not considered in this comparatively simple, cell-based model. Thus, four important assumptions were made in the model. Firstly, that by using independent coefficients from multivariate analyses, reducing one risk factor would not affect another one. In reality, multiplicative interactions between various risk factors may further decrease the CHD risk [47]. Secondly, assuming an immediate CHD mortality benefit from risk factor modifications. In reality, delay might be surprisingly short. Thus, the relative risk of CHD death declines by approximately 37% at 1 year after quitting smoking [48]. Furthermore, the surprisingly rapid CHD mortality reversals seen in Poland, Hungary and the Czech Republic [attributed to dietary changes] have recently been mirrored in communities adopting smoke-free legislation [49‐51]. Thirdly, we considered only CHD deaths, not "competing causes" [52]. However, reductions in major risk factors would actually decrease deaths from other non-CHD causes, such as diabetes and common cancers [53]. Fourthly, the treatment uptakes in 2010 were assumed to remain constant at levels measured in 1999. In fact, more substantial improvements may be anticipated [18, 54]. Our model estimates should therefore not be viewed as precise projections, but are usefully indicative of the changes in CHD mortality rates which might be expected under our three alternative scenarios.

In conclusion, without specific interventions, a substantial increase in CHD deaths in Beijing is likely to occur. This will reflect worsening risk factors compounded by population ageing. However, modest, but eminently achievable, improvements in major risk factors could actually decrease the CHD burden. Future prevention strategies clearly need to be substantially strengthened. Furthermore, these results may be cautiously generalizable to other middle income countries addressing their own cardiovascular disease epidemics.