Early-onset type 2 diabetes, lower baseline BMI predict progression to insulin use in Asians
AUSTIN, Texas — In Chinese adults, type 2 diabetes onset before age 40 years, higher baseline HbA1c, and BMI 18.5 kg/m² or lower were all important predictors of progression to insulin therapy over nearly 10 years of follow-up, according to a Hong Kong database analysis presented here.
“Over the last 20 years, diabetes prevalence has gone up from 1% to over 10% in China, which translates to over 100 million people with diabetes in China alone,” Ronald C.W. Ma, MB BChir, of the department of medicine and therapeutics at Chinese University of Hong Kong, said during a presentation at the American Association of Clinical Endocrinologists Scientific and Clinical Congress. “Some of the features that distinguish Asians with diabetes vs. the European population is that diabetes tends to occur at a younger age in the Asian population, it tends to occur at lower BMI, and there is a larger proportion with a positive family history, pointing toward genetic effects in beta-cell dysfunction.”
Ma and colleagues analyzed data from 7,570 consecutive Chinese patients with type 2 diabetes enrolled in the Hong Kong Diabetes Registry between 1995 and 2007; patients already on insulin therapy at baseline or patients who required insulin within 1 year of follow-up were excluded. Researchers followed the cohort for a median of 9.3 years, identifying all patients who progressed to continuous insulin therapy; time to insulin use was calculated for all patients, who were tracked through a central Hong Kong registry, Ma said. Researchers used Cox regression analysis to identify clinical predictors of progression to insulin use. Time-to-insulin was defined as the first prescription of insulin for 6 months or more.
Within the cohort, 2,882 progressed to insulin use over the course of follow-up (38%; 1,861 with available genome-wide association data). Those who did progress to insulin use on average were younger, had earlier-onset of diabetes (onset younger than age 40 years), longer duration of diabetes at time of recruitment and shorter duration of follow-up, Ma said. Incident insulin users also had higher mean HbA1c, triglycerides and LDL cholesterol, and lower HDL cholesterol at baseline, Ma said, and were more likely to have complications, including sensory neuropathy, retinopathy, albuminuria or chronic kidney disease at baseline.
“Those with a BMI of less than 18.5 [kg/m²] had a markedly higher rate of progression to insulin use and were approximately started on insulin 5 years earlier than the other groups,” Ma said.
In multivariate analysis with BMI treated as a categorical variable, independent predictors of progression to insulin use included age at diagnosis (HR = 0.97; 95% CI, 0.97-0.98), smoking status, higher baseline HbA1c (HR = 1.32; 95% CI, 1.3-1.35), lower estimated glomerular filtration rate (HR = 0.99; 95% CI, 0.99-0.99), presence of diabetic retinopathy (HR = 1.68; 95% CI, 1.54-1.83) and low baseline BMI (mean of 18.5 kg/m² or less; HR = 1.71; 95% CI, 1.36-2.17).
“BMI as a continuous trait was not an independent predictor for the subsequent need for insulin,” Ma said. “However, when we divided BMI according to categorical variables, again, individuals who have low BMI are at approximately 70% increased risk to be started on insulin during [the] follow-up period.”
Ma cautioned that the data may not apply to other ethnic populations; however, preliminary analysis of the data using the glycemic deterioration endpoint yielded similar analysis in terms of risk factors predicting progression to glycemic deterioration conducted in other countries. Additional work assessing genetic and other biomarkers is ongoing to better understand the factors associated with diabetes progression in the Asian population, Ma said. – by Regina Schaffer
Ma RCW, et al. ABSTRACT #275. AACE Annual Scientific and Clinical Congress; May 3-7, 2017; Austin, Texas.
Disclosures: Ma reports receiving speaking and consulting fees from Astra Zeneca, Boehringer Ingelheim, Eli Lilly, Pfizer, Sanofi and Takeda, and clinical trial support for his institution from Astra Zeneca, Bayer and MSD.
We calculated time-to-insulin use for each patient. We
to insulin use.