Risk Factor Trends in Cardiometabolic Mortality Decline
Risk Factor Trends in Cardiometabolic Mortality Decline
In 2009, age-standardized cardiometabolic death rate was highest in Argentina and Mexico for both men and women. It was lowest in Australia, Switzerland, France and Japan. Age-standardized death rate in Argentinean men was about three times that of men in Australia, where adult male cardiometabolic mortality was lowest; the death rate of women in Mexico was five times that of Japanese women. Between 1980 and 2009, age-standardized cardiometabolic mortality declined in all 26 countries with the relative decline ranging from <1% per year in Mexico to ~5% in Australia for both men and women.
Age-standardized mean BMI increased in most countries, with trends for women in a few Western European countries statistically indistinguishable from no change. BMI increased the most in Mexico and Chile for women (by 0.16 kg/m per year) and in Mexico and the USA for men (0.11 kg/m per year). Except for Chilean men, SBP declined in all countries, by as much as 0.28 mmHg per year in Finland, Luxembourg and the USA for men and 0.43 mmHg per year in Finland and France for women. TC declined by 0.03–0.04 mmol/l per year or more in men and women in Finland, Sweden, New Zealand and the UK. The decline was smaller in countries in Southern Europe, Germany and Argentina; it was nearly zero in Mexico and Chile and increased by over 0.01 mmol/l per year in Japan. SBP and TC declined by a smaller amount where BMI rose more (Figure 1). In contrast, there was an inverse association between change in BMI and change in lung cancer, the proxy for smoking, in both sexes. Country mortality and risk factor trends are presented in Supplementary Figures 1–5, available as Supplementary Data at IJE online.
(Enlarge Image)
Figure 1.
The cross-country associations between change in BMI and changes in SBP, TC and lung cancer mortality. Each point shows one country. The lines show the fitted linear association
Before adjustment, changes in mean population BMI, SBP, TC and (for men only) lung cancer were positively associated with change in cardiometabolic mortality (Figure 2). The associations were steeper for men than for women except that of BMI. After adjustment for other risk factors and GDP, each 10-mmHg reduction in mean population SBP was associated with 46% (95% CI 4 to 70) decline in cardiometabolic mortality for men and 41% (95% CI -7 to 67) for women. One mmol/l reduction in mean TC was associated with 21% (95% CI -11 to 43) decline in cardiometabolic mortality in men and 18% (95% CI -13 to 40) in women. When adjusted for SBP and TC in addition to smoking and GDP, 1 kg/m rise in mean BMI was associated with 5% (95% CI -10 to 18) increase in cardiometabolic mortality in men and 6% (95% CI -3 to 14) in women; the association was larger without adjustment for SBP and TC (9% in men and 10% in women). Each doubling of LC mortality was associated with a 5% increase in cardiometabolic mortality among men and 16% among women. Our conclusions about the role of risk factors in cardiometabolic mortality at the population level were similar in Sensitivity analyses 1 and 2 (see Supplementary Table 1, available as Supplementary data at IJE online). The magnitudes of the associations were smaller for SBP and for BMI among men without adjustment for GDP; the confidence intervals of adjusted and unadjusted effect sizes overlapped in all instances.
(Enlarge Image)
Figure 2.
The cross-country associations between change in risk factors and change in cardiometabolic (CVD and diabetes) death rates. All variables were age-standardized as described in Methods. Each point shows one country. The lines show the fitted linear associations. See Supplementary Figure 6 for results for CVD alone (available as Supplementary Data at IJE online)
Risk factor trends may have adversely affected cardiometabolic mortality trajectory in Chilean men and women (i.e. slowed down the mortality decline), and had virtually no effect among Argentinean women. The populations of the other countries are likely to have benefited from lower risk factor levels in 2009 compared with 1980 (Figure 3). If the above associations reflect the causal effects of risk factor change on mortality decline, trends in these four risk factors alone might have led to over 60% of cardiometabolic mortality decline among men and women in Finland and Switzerland, men in New Zealand and France and women in Italy. Most of these countries had above average improvements in SBP and TC, and for men in smoking; Switzerland, Finland, France and Italy also had below-average rise in BMI. At the low end, the risk factors accounted for ~11–26% of mortality decline in Mexican, Portuguese, and Japanese men and for ~18% in Mexican women. These countries performed poorly in terms of risk factor trends compared with most other countries, with either smaller declines (e.g. in SBP) or larger increase (e.g. BMI in Mexican women and TC in Japanese men). Portugal was also one of the three countries where cumulative smoking increased among men and Japan one of the few without a decline. In the USA, risk factor trends accounted for an estimated 56% of the reduction in mortality among men and 28% among women. The male-female difference was due to smaller SBP decline and later decline of smoking among American women compared with American men. The male-female differential in the role of risk factors also seemed relatively large in Canada and The Netherlands. In Southern European countries, especially Portugal, and in Japan, risk factor trends may have accounted for noticeably more of the mortality decline among women compared to men.
(Enlarge Image)
Figure 3.
Proportion of decline in cardiometabolic mortality due to changes in BMI, SBP, TC and smoking. A negative fraction should be interpreted as a situation in which risk factor trends slowed down mortality decline
Not only did countries differ in the proportional contributions of risk factors to cardiometabolic mortality decline, but they also differed in their absolute contributions (Figure 4). The estimated absolute benefits were largest among men in Finland and New Zealand, with about 381 and 346 fewer deaths per 100 000, respectively, in 2009 due to risk factor trends. In contrast, the estimated absolute mortality benefits were close to zero in Mexico, and between 70 and 90 deaths per 100 000 in Japanese, Portuguese and Greek men. Chilean men had about 109 more deaths per 100 000 in 2009 than they would have experienced if their risk factor profiles had not changed. The absolute benefits were smaller in women, who experienced smaller overall reduction in cardiometabolic mortality than men in most countries. The largest estimated benefits among women were in Finland, Israel, Italy and New Zealand about 150–180 fewer deaths per 100 000 in 2009 due to risk factor trends. Women in Chile had about 29 more deaths per 100 000 in 2009 than they would have had if their risk factor profiles had not changed.
(Enlarge Image)
Figure 4.
Change in age-standardized cardiometabolic death rates between 1980 and 2009 (total length of each bar) and change that would have been expected solely due to risk factor trends (the coloured section). The changes are shown as both (A) absolute and (B) relative
In 2009, there were about 810 000 deaths from CVD and diabetes among men and about 490 000 among women aged 25–79 years in these 26 countries. This is approximately 1 210 000 fewer (for men) and 710 000 fewer (for women) than would have occurred had death rates remained at their 1980 levels. The observed risk factor trends alone would have achieved an estimated 48% (men) and 40% (women) of these avoided deaths. If all countries had experienced the risk factor changes of the best performing countries, the number of cardiometabolic deaths in these countries in 2009 would be lower by an impressive 27% for men and 42% for women, larger than the proposed global goal for reducing noncommunicable disease (NCD) mortality.
Results
Mortality and Risk Factor Trends
In 2009, age-standardized cardiometabolic death rate was highest in Argentina and Mexico for both men and women. It was lowest in Australia, Switzerland, France and Japan. Age-standardized death rate in Argentinean men was about three times that of men in Australia, where adult male cardiometabolic mortality was lowest; the death rate of women in Mexico was five times that of Japanese women. Between 1980 and 2009, age-standardized cardiometabolic mortality declined in all 26 countries with the relative decline ranging from <1% per year in Mexico to ~5% in Australia for both men and women.
Age-standardized mean BMI increased in most countries, with trends for women in a few Western European countries statistically indistinguishable from no change. BMI increased the most in Mexico and Chile for women (by 0.16 kg/m per year) and in Mexico and the USA for men (0.11 kg/m per year). Except for Chilean men, SBP declined in all countries, by as much as 0.28 mmHg per year in Finland, Luxembourg and the USA for men and 0.43 mmHg per year in Finland and France for women. TC declined by 0.03–0.04 mmol/l per year or more in men and women in Finland, Sweden, New Zealand and the UK. The decline was smaller in countries in Southern Europe, Germany and Argentina; it was nearly zero in Mexico and Chile and increased by over 0.01 mmol/l per year in Japan. SBP and TC declined by a smaller amount where BMI rose more (Figure 1). In contrast, there was an inverse association between change in BMI and change in lung cancer, the proxy for smoking, in both sexes. Country mortality and risk factor trends are presented in Supplementary Figures 1–5, available as Supplementary Data at IJE online.
(Enlarge Image)
Figure 1.
The cross-country associations between change in BMI and changes in SBP, TC and lung cancer mortality. Each point shows one country. The lines show the fitted linear association
Associations of Risk Factor Changes With Change in Cardiometabolic Mortality
Before adjustment, changes in mean population BMI, SBP, TC and (for men only) lung cancer were positively associated with change in cardiometabolic mortality (Figure 2). The associations were steeper for men than for women except that of BMI. After adjustment for other risk factors and GDP, each 10-mmHg reduction in mean population SBP was associated with 46% (95% CI 4 to 70) decline in cardiometabolic mortality for men and 41% (95% CI -7 to 67) for women. One mmol/l reduction in mean TC was associated with 21% (95% CI -11 to 43) decline in cardiometabolic mortality in men and 18% (95% CI -13 to 40) in women. When adjusted for SBP and TC in addition to smoking and GDP, 1 kg/m rise in mean BMI was associated with 5% (95% CI -10 to 18) increase in cardiometabolic mortality in men and 6% (95% CI -3 to 14) in women; the association was larger without adjustment for SBP and TC (9% in men and 10% in women). Each doubling of LC mortality was associated with a 5% increase in cardiometabolic mortality among men and 16% among women. Our conclusions about the role of risk factors in cardiometabolic mortality at the population level were similar in Sensitivity analyses 1 and 2 (see Supplementary Table 1, available as Supplementary data at IJE online). The magnitudes of the associations were smaller for SBP and for BMI among men without adjustment for GDP; the confidence intervals of adjusted and unadjusted effect sizes overlapped in all instances.
(Enlarge Image)
Figure 2.
The cross-country associations between change in risk factors and change in cardiometabolic (CVD and diabetes) death rates. All variables were age-standardized as described in Methods. Each point shows one country. The lines show the fitted linear associations. See Supplementary Figure 6 for results for CVD alone (available as Supplementary Data at IJE online)
Contributions of Risk Factor Trends to Mortality Decline
Risk factor trends may have adversely affected cardiometabolic mortality trajectory in Chilean men and women (i.e. slowed down the mortality decline), and had virtually no effect among Argentinean women. The populations of the other countries are likely to have benefited from lower risk factor levels in 2009 compared with 1980 (Figure 3). If the above associations reflect the causal effects of risk factor change on mortality decline, trends in these four risk factors alone might have led to over 60% of cardiometabolic mortality decline among men and women in Finland and Switzerland, men in New Zealand and France and women in Italy. Most of these countries had above average improvements in SBP and TC, and for men in smoking; Switzerland, Finland, France and Italy also had below-average rise in BMI. At the low end, the risk factors accounted for ~11–26% of mortality decline in Mexican, Portuguese, and Japanese men and for ~18% in Mexican women. These countries performed poorly in terms of risk factor trends compared with most other countries, with either smaller declines (e.g. in SBP) or larger increase (e.g. BMI in Mexican women and TC in Japanese men). Portugal was also one of the three countries where cumulative smoking increased among men and Japan one of the few without a decline. In the USA, risk factor trends accounted for an estimated 56% of the reduction in mortality among men and 28% among women. The male-female difference was due to smaller SBP decline and later decline of smoking among American women compared with American men. The male-female differential in the role of risk factors also seemed relatively large in Canada and The Netherlands. In Southern European countries, especially Portugal, and in Japan, risk factor trends may have accounted for noticeably more of the mortality decline among women compared to men.
(Enlarge Image)
Figure 3.
Proportion of decline in cardiometabolic mortality due to changes in BMI, SBP, TC and smoking. A negative fraction should be interpreted as a situation in which risk factor trends slowed down mortality decline
Not only did countries differ in the proportional contributions of risk factors to cardiometabolic mortality decline, but they also differed in their absolute contributions (Figure 4). The estimated absolute benefits were largest among men in Finland and New Zealand, with about 381 and 346 fewer deaths per 100 000, respectively, in 2009 due to risk factor trends. In contrast, the estimated absolute mortality benefits were close to zero in Mexico, and between 70 and 90 deaths per 100 000 in Japanese, Portuguese and Greek men. Chilean men had about 109 more deaths per 100 000 in 2009 than they would have experienced if their risk factor profiles had not changed. The absolute benefits were smaller in women, who experienced smaller overall reduction in cardiometabolic mortality than men in most countries. The largest estimated benefits among women were in Finland, Israel, Italy and New Zealand about 150–180 fewer deaths per 100 000 in 2009 due to risk factor trends. Women in Chile had about 29 more deaths per 100 000 in 2009 than they would have had if their risk factor profiles had not changed.
(Enlarge Image)
Figure 4.
Change in age-standardized cardiometabolic death rates between 1980 and 2009 (total length of each bar) and change that would have been expected solely due to risk factor trends (the coloured section). The changes are shown as both (A) absolute and (B) relative
In 2009, there were about 810 000 deaths from CVD and diabetes among men and about 490 000 among women aged 25–79 years in these 26 countries. This is approximately 1 210 000 fewer (for men) and 710 000 fewer (for women) than would have occurred had death rates remained at their 1980 levels. The observed risk factor trends alone would have achieved an estimated 48% (men) and 40% (women) of these avoided deaths. If all countries had experienced the risk factor changes of the best performing countries, the number of cardiometabolic deaths in these countries in 2009 would be lower by an impressive 27% for men and 42% for women, larger than the proposed global goal for reducing noncommunicable disease (NCD) mortality.
Source...