摘要
We read with great interest the Article by Edward Gregg and colleagues (June 16, 2018, p 2430).1Gregg EW Cheng YJ Srinivasan M et al.Trends in cause-specific mortality among adults with and without diagnosed diabetes in the USA: an epidemiological analysis of linked national survey and vital statistics data.Lancet. 2018; 391: 2430-2440Summary Full Text Full Text PDF PubMed Scopus (236) Google Scholar Here, we note some of our concerns on the methodology. First, age standardisation is widely used in epidemiological trend analyses to reduce the bias introduced by the variations in age distribution across different calendar years. We commend the investigators for their use of age adjustment, which, however, modelled an even distribution of the age groups in the population (unless data were already age standardised) and probably changed the reported population mortality rates. Age adjustment is valid for inferential analysis but is inferior to age standardisation in trend analysis. Considering the study period of 27 years (1988–2015) and the different age trends in the diabetes (no changes) and non-diabetes (gradually increasing) groups, we recommend analysis of age-standardised mortality rates based on the US standard population from the 2000 national census. Second, insurance coverage,2Semilla AP Chen F Dall TM Reductions in mortality among Medicare beneficiaries following the implementation of Medicare Part D.Am J Manag Care. 2015; 21: s165-s171PubMed Google Scholar poverty,3Saydah SH Imperatore G Beckles GL Socioeconomic status and mortality: contribution of health care access and psychological distress among U.S. adults with diagnosed diabetes.Diabetes Care. 2013; 36: 49-55Crossref PubMed Scopus (100) Google Scholar rural residence,4Magliano DJ Cohen K Harding JL Shaw JE Residential distance from major urban areas, diabetes and cardiovascular mortality in Australia.Diabetes Res Clin Pract. 2015; 109: 271-278Summary Full Text Full Text PDF PubMed Scopus (7) Google Scholar and geographic location were linked to diabetic mortality. These socioeconomic factors might contribute to the large variations in diabetes-related mortality, as the authors rightfully noticed. Most of these socioeconomic factors were indeed surveyed in the National Health Interview Survey and perhaps should be included in the analyses. Furthermore, 3-year percentage changes are more sensitive in detecting turning points for different line slopes than the reported 10-year percentage changes and perhaps should have been used instead. Indeed, the coverage for Medicare Part D was implemented in 2006,2Semilla AP Chen F Dall TM Reductions in mortality among Medicare beneficiaries following the implementation of Medicare Part D.Am J Manag Care. 2015; 21: s165-s171PubMed Google Scholar and the Affordable Care Act became effective in 2014.5Kaufman HW Chen Z Fonseca VA McPhaul MJ Surge in newly identified diabetes among medicaid patients in 2014 within Medicaid expansion states under the Affordable Care Act.Diabetes Care. 2015; 38: 833-837Crossref PubMed Scopus (67) Google Scholar These events have affected the incidence of diabetes and its associated mortality.2Semilla AP Chen F Dall TM Reductions in mortality among Medicare beneficiaries following the implementation of Medicare Part D.Am J Manag Care. 2015; 21: s165-s171PubMed Google Scholar, 5Kaufman HW Chen Z Fonseca VA McPhaul MJ Surge in newly identified diabetes among medicaid patients in 2014 within Medicaid expansion states under the Affordable Care Act.Diabetes Care. 2015; 38: 833-837Crossref PubMed Scopus (67) Google Scholar Therefore, sensitive analyses using shorter intervals such as 3-year percentage changes might be warranted to detect the trend changes linked to these events. We declare no competing interests. Trends in cause-specific mortality among adults with and without diagnosed diabetes in the USA: an epidemiological analysis of linked national survey and vital statistics dataDeclining rates of vascular disease mortality are leading to a diversification of forms of diabetes-related mortality with implications for clinical management, prevention, and disease monitoring. Full-Text PDF Trend analysis of diabetic mortality – Authors' replyWe appreciate Xiaoling Yuan and colleagues' Correspondence about our Article.1 We used age adjustment instead of age standardisation because we considered it to be more efficient and appropriate to account for the reduced power in younger age strata in the context of discretised Poisson regression models, which we used to transform serial cohorts into nationally representative, year-specific estimates. Our analyses also modelled mortality trends as a continuous variable and evaluated non-linear terms, in addition to presenting point estimates at 10-year intervals. Full-Text PDF