Repositioning of the global epicentre of non-optimal cholesterol

Abstract

High blood cholesterol is typically considered a feature of wealthy western countries^1,2. However, dietary and behavioural determinants of blood cholesterol are changing rapidly throughout the world³ and countries are using lipid-lowering medications at varying rates. These changes can have distinct effects on the levels of high-density lipoprotein (HDL) cholesterol and non-HDL cholesterol, which have different effects on human health^4,5. However, the trends of HDL and non-HDL cholesterol levels over time have not been previously reported in a global analysis. Here we pooled 1,127 population-based studies that measured blood lipids in 102.6 million individuals aged 18 years and older to estimate trends from 1980 to 2018 in mean total, non-HDL and HDL cholesterol levels for 200 countries. Globally, there was little change in total or non-HDL cholesterol from 1980 to 2018. This was a net effect of increases in low- and middle-income countries, especially in east and southeast Asia, and decreases in high-income western countries, especially those in northwestern Europe, and in central and eastern Europe. As a result, countries with the highest level of non-HDL cholesterol-which is a marker of cardiovascular risk-changed from those in western Europe such as Belgium, Finland, Greenland, Iceland, Norway, Sweden, Switzerland and Malta in 1980 to those in Asia and the Pacific, such as Tokelau, Malaysia, The Philippines and Thailand. In 2017, high non-HDL cholesterol was responsible for an estimated 3.9 million (95% credible interval 3.7 million-4.2 million) worldwide deaths, half of which occurred in east, southeast and south Asia. The global repositioning of lipid-related risk, with non-optimal cholesterol shifting from a distinct feature of high-income countries in northwestern Europe, north America and Australasia to one that affects countries in east and southeast Asia and Oceania should motivate the use of population-based policies and personal interventions to improve nutrition and enhance access to treatment throughout the world.

Fig. 1. Change in age-standardized mean total cholesterol between 1980 and 2018 by region for women and men.

a, Age-standardized mean total cholesterol in women. b, Age-standardized mean total cholesterol in men. The start of the arrow shows the level in 1980 and the head indicates the level in 2018. See Extended Data Fig. 3 for age-standardized mean HDL cholesterol. One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Fig. 2. Change in age-standardized mean non-HDL cholesterol between 1980 and 2018 by region for women and men.

a, Age-standardized mean non-HDL cholesterol in women. b, Age-standardized mean non-HDL cholesterol in men. The start of the arrow shows the level in 1980 and the head indicates the level in 2018. See Extended Data Fig. 3 for age-standardized mean HDL cholesterol. One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Fig. 3. Age-standardized mean non-HDL cholesterol by country in 1980 and 2018 for women and men.

a, Age-standardized mean non-HDL cholesterol in women in 1980. b, Age-standardized mean non-HDL cholesterol in women in 2018. c, Age-standardized mean non-HDL cholesterol in men in 1980. d, Age-standardized mean non-HDL cholesterol in men in 2018. See Extended Data Fig. 5 for age-standardized mean total cholesterol and Extended Data Fig. 6 for age-standardized mean HDL cholesterol. One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Fig. 4. Change in age-standardized mean non-HDL cholesterol per decade by country for women and men.

a, Change per decade in age-standardized mean non-HDL cholesterol in women. b, Change per decade in age-standardized mean non-HDL cholesterol in men. See Extended Data Fig. 7 for change per decade in age-standardized mean total cholesterol and Extended Data Fig. 8 for change per decade in age-standardized mean HDL cholesterol. One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Fig. 5. Deaths from IHD and ischaemic stroke attributable to high non-HDL cholesterol by region in 1990 and 2017 for women and men.

a, Deaths in women attributable to high non-HDL cholesterol. b, Deaths in men attributable to high non-HDL cholesterol.

Extended Data Fig. 1. Number of data sources by country.

The colour indicates the number of data sources for each country used in the analysis. Countries and territories that were not included in the analysis are coloured in grey.

Extended Data Fig. 2. Number of data sources by region and year.

The size of each circle shows the number of data sources for each region and year, and the colours indicate the relative size of national, subnational and community data sources.

Extended Data Fig. 3. Change in age-standardized mean HDL cholesterol between 1980 and 2018 by region for women and men.

The start of the arrow shows the level in 1980 and the head shows the level in 2018. One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Extended Data Fig. 4. Change in age-standardized mean HDL and non-HDL cholesterol between 1980 and 2018 by region for women and men.

One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Extended Data Fig. 5. Age-standardized mean total cholesterol by country in 1980 and 2018 for women and men.

One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Extended Data Fig. 6. Age-standardized mean HDL cholesterol by country in 1980 and 2018 for women and men.

One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Extended Data Fig. 7. Change per decade in age-standardized mean total cholesterol by country for women and men.

One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Extended Data Fig. 8. Change per decade in age-standardized mean HDL cholesterol by country for women and men.

One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.

Extended Data Fig. 9. The association between mean LDL and non-HDL cholesterol in studies that measured lipids in a laboratory that had data for both variables.

Each data point is one study–age–sex group (n = 6,864). One mmol l⁻¹ is equivalent to 38.61 mg dl⁻¹.