Almost 350 million adults now have diabetes, according to results from a sweeping new study evaluating 3 decades of fasting–plasma glucose trends worldwide.
The findings, published online June 25 in the Lancet (doi:10.1016/S0140-6736(11)60679-X), challenge recent estimates of the global diabetic population at about 285 million (Diabetes Res Clin Pract 2009; 87: 4-14) and also suggest, in contrast to other views, that diabetes in east and southeast Asia is rising mainly because of aging and growing populations, rather than lifestyle changes.
The increased global estimate reflects a larger number of national studies analyzed for this study, along with different analytical methods, reported members of the Global Burden of Metabolic Risk Factors of Chronic Diseases Collaborating Group, which conducted the study.
Worldwide, the mean age-standardized fasting–plasma glucose (FPG) concentrations have risen on average by 0.08 mmol/L per decade (0.07 for men and 0.09 for women) since 1980, which corresponds to an increase in overall diabetes prevalence of about 7% per decade (from 8.3% in 1980 to 9.8% in 2008 among men, and from 7.5% to 9.2% among women).
The number of people with diabetes is estimated to have more than doubled, from about 153 million in 1980 to about 347 million in 2008. Of these, 40% live in China or India.
About 70% of the worldwide increase in prevalence can be attributed to population growth and aging, the investigators wrote, but the rise in age-standardized FPG indicates "an important epidemiological component" accounting for the other 30%, although the reasons vary by region. The authors found, by comparing their results to a recent body mass index (BMI) study of similar scope (Lancet 2011;377: 557–67), that the correlation between changes in fasting glucose and changes in BMI was 0.71 for women and 0.57 for men.
In no part of the world was a significant decline in FPG noted over the study period, except among Singaporean women, although levels in many regions remained flat.
For their research, which was funded by the Bill and Melinda Gates Foundation and the World Health Organization, the investigators looked at data from all available health examination surveys and epidemiological studies, including national and regional studies, for the 29-year study period. The study covered 199 countries and territories, although 92 countries provided no population-based data.
The researchers sought information on FPG among adults aged 25 years and older, by sex. They chose mean FPG as the main glycemic measure because it is widely used in population-based studies (including more than 71% of the studies used in their analysis), and "there is a continuous association between FPG and cardiovascular disease, including at levels below clinical thresholds for diabetes diagnosis," the investigators wrote.
Data from studies using different glycemic indicators were converted to FPG for this analysis. For each sex, investigators used Bayesian modeling to estimate mean FPG and its uncertainty by age, country, and year, to adjust for whether a study was nationally, regionally, or locally representative. Diabetes prevalence was estimated using mean FPG.
The harmonization of data into a readily comparable analysis, spanning large populations over a long time period, was the main strength of this study, compared with previous studies, the investigators wrote. Its size was another: The study looked at data for 2.7 million participants over 370 country-years. An added strength was the incorporation of data from several large national studies from east and southeast Asia.
The study’s main weaknesses, they wrote, were that some countries did not provide population-based data, and that many country-years did not have data, especially in the 1980s; the authors adjusted for these gaps with modeling.
Oceania, the Caribbean regions, and North Africa and the Middle East were found to have the highest FPG and diabetes prevalence in 2008, with an age- standardized mean FPG of 6.5 mmol/L or higher in several South Pacific island nations; diabetes prevalence was estimated at 21%-25% in men and up to 32% of women.
Countries in southeast Asia, east Africa, and Andean Latin America had the lowest mean FPG in 2008 (as low as 5 mmol/L or less). Among high-income countries, mean FPG and diabetes prevalence were highest in the United States, Greenland, Malta, New Zealand, and Spain, and were lowest in the Netherlands and Austria for both sexes (and in France for women). Mean FPG in the Netherlands, Austria, and France was lower than in Japan and South Korea, despite higher BMIs.
FPG increased the most among South Pacific Islanders, by 0.22 mmol/L per decade in men and 0.32 in women. Tropical and southern Latin America and South Asia saw large increases for men, and south Asia, central Asia, north Africa, and the Middle East recorded large increases for women. FPG remained essentially flat in western Europe, increasing by 0.07 mmol/L per decade in men and by 0.03 in women. In North America, by contrast, FPG rose by 0.18 mmol/L per decade in men and by 0.14 in women.