Racial/ethnic disparities in type 2 diabetes remission after bariatric surgery
MetadataShow full item record
BACKGROUND: Previous studies have shown that there are racial disparities in type 2 diabetes (T2DM) remission following bariatric surgery, with African-Americans (AA) in particular experiencing a subsequent relapse. In recent years, some have attributed these findings to racial differences in fasting insulin levels, with AA having higher levels, as increasing evidence for an alternate model of T2DM pathophysiology gains support. In this model, basal hyperinsulinemia is considered a primary event in T2DM disease development, rather than a compensatory response to increased insulin resistance. This study aimed to compare glycemic outcomes after bariatric surgery in different races, namely African-Americans (AA), Hispanic-Americans (HA), and Caucasian-Americans (CA), and to determine whether there were any associated changes in insulin levels and insulin resistance that may lend support to this revised model of T2DM pathophysiology. METHODS: A retrospective medical record review of 1,326 patients (389 AA, 179 HA, and 758 CA) who underwent bariatric surgery at Boston Medical Center (BMC) from 2004 to 2015 was conducted. Baseline characteristics and maximum percent weight loss were compared using one-way ANOVA and Chi-square tests of independence. Changes in mean glycated hemoglobin (HbA1c), insulin levels, insulin resistance (HOMA-IR), and blood glucose levels were analyzed using linear mixed models, overall and by racial group. The same procedures were conducted in both the overall patient population and a T2DM subpopulation. RESULTS: Over an 11-year postoperative observation period, all racial groups underwent a significant decrease in HbA1c (P<0.001) within the first two years following surgery. While HbA1c levels remained stable in CA and HA, they began to rise at 2 years in AA only (P=0.043). Additionally, analyses of covariates, including age at surgery (P=0.005), initial BMI (P<0.001), and maximum weight loss (P=0.049), revealed that all three were significant factors affecting mean HbA1c levels. However, when included in the mixed model, the race x time interaction effect on mean HbA1c remained significant. There was also a significant overall decrease in both insulin and HOMA-IR. When stratified by race, analysis of the T2DM population showed that insulin levels began to increase again by the 2nd year after surgery in AA, while in CA and HA they continued to decrease and subsequently stabilize. Analysis of the total patient population showed that HOMA-IR levels in AA, as well as in CA and HA, continued to decrease at this 2-year time point. Decreases in blood glucose levels after surgery were significant overall (P<0.001), but not significant when stratified by race. CONCLUSIONS: After the initial “metabolic reset” that occurs within the first 2 years after bariatric surgery, during which HbA1c levels normalize in the vast majority of patients, it was observed only in the AA population that there was a steady increase in HbA1c to levels near those recorded at baseline. This coincided with an observation of increasing insulin levels despite decreasing insulin resistance seen in AA only. Our results suggest that current discussions regarding a revised model of T2DM pathophysiology, in which hyperinsulinemia precedes insulin resistance, may help explain the racial disparities in glycemic control observed in both post-surgical and non-surgical contexts of T2DM outcome. However, future prospective studies are needed to further the preliminary results of this study.