Anthropometric changes may predict diabetic response after bariatric surgery

Pornthep Prathanvanich, MD, Sharfi Sarker, MD, P Marco Fisichella, MD, Bipan Chand, MD. Loyola University Chicago Stritch School of Medicine.

Background
Type 2 diabetes mellitus (T2DM) is a life-threating obesity linked condition often treated with bariatric surgery. Predictors of positive surgical response include duration and severity of diabetes, type of operation employed and various patient factors. Central obesity, as measured by waist circumference and waist/hip ratio, is associated with increased cardiometabolic risk. In this study, patient and surgical factors, including anthropometric changes, were analyzed in predicting T2DM response.

Method
145 consecutive morbidly obese surgical patients at Loyola Health University Chicago were enrolled in this IRB approved cohort study, from Jan 2012 to Sep 2013. Analyzed were 35 patients (22 women; mean age = 44.86±9.81) having T2DM. Demographic data, diabetic type (non-insulin or insulin dependent) and type of surgery were analyzed. Anthropometric factors were measured before and at intervals of 1 week, 1 month and every 3 months post-operatively. Follow up time was 8.09 ± 3.72 (range, 3-15) months. Predictors of diabetic remission (complete remission, improvement and non-response) were analyzed. Definition of complete remission (CR) was considered when there was a return to “normal” glucose metabolism (HbA1C < 6%, fasting glucose < 100 mg/Dl). Diabetic improvement (DI) was defined as lower drug dosage with sub-diabetic hyperglycemia (fasting glucose 100–125 mg/dl).

Results
Surgical intervention included 19/35 (54.28%) gastric bypass (LRYGB), 10/35 (28.57%) sleeve gastrectomy (LSG) and 6/35 (17.15%) gastric banding (LAGB). There were no demographic or anthropometric measurement differences between type of surgery and gender (p= 0.370), race (p=0.503) or diabetic type (p= 0.267). Mean preoperative BMI (kg/m2) was 45.03±6.31 (LRYGB), 53.25±10.49 (LSG), and 50.41±13.70 (LAGB). When evaluating postoperative factors, only a decrease in waist circumference (WC) at 3 months was associated significantly with gastric bypass (p=0.030; 95% Confidence Interval, CI: 0.56-10.47) but not with either LSG or LAGB. Eighteen (51.42%), eleven (31.42%) and six (17.16%) of the 35 patients had complete remission, improvement and non-response of their diabetes respectively. In univariate analysis, preoperative independent predictors of success (CR and DI) included; white race (p=0.018), LRYGB (p=0.015), non-insulin dependent (p=0.025), pre-BMI ≤ 60 kg/m2 (p=0.019) and pre-WC ≤ 160 cm (p=0.010). Postoperative parameters of success included; decrease in waist/hip ratio at 1 and 3 months (0.012 and 0.00 respectively) and greater change in BMI. Waist and neck circumference (NC) changes at 1 and 3 months were associated significantly with complete remission and improvement in diabetes. In multivariate logistical regression analysis, type of surgery (p=0.019; CI: 0.106-1.069), greater ΔWC at 1 month and greater ΔNC at 3 months remained independent predictors of success.

Conclusion
Bariatric surgery remains the most effective treatment of diabetes (overall 82% of success) in the morbidly obese patient. Predictors of positive outcome may not only include the type of surgery and duration/ severity of T2DM, but changes in fat distribution. Gastric bypass, in this short-term study, led to a greater degree of change in central obesity. Long-term follow-up may show that gastric bypass is superior in resolving central adiposity and therefore decreasing cardiometabolic risk and improving diabetes.

View Poster