Abstract P181: Temporal Relationship Between Changes in Body Weight and Insulin Resistance and Its Impact on Blood Pressure in Response to 2-year Weight-loss Diet Interventions: the POUNDS Lost Trial

Liyuan Han,Tao Zhang,Dianjianyi Sun,Wenjie Ma,Tiange Wang,Yoriko Heianza,Wei Chen,George A. Bray,Frank M Sacks,Lu Qi
DOI: https://doi.org/10.1161/circ.135.suppl_1.p181
IF: 37.8
2017-01-01
Circulation
Abstract:Introduction: Changes of body weight and insulin resistance (IR) are closely correlated, and their temporal sequences in affecting blood pressure (BP) remain poorly defined. We examined the temporal sequences of weight loss and change of IR, and their relations with blood pressure in the Pounds Lost trial, a randomized weight-loss diet intervention study. Hypothesis: We hypothesized a one-directional temporal relation between weight loss and change of IR in affecting BP during weight-loss diet interventions. Methods: The present study included overweight/obese adults, who were randomized in a 2х2 factorial design to diets containing 20% or 40% fat and 15% or 25% protein (diets with 65%, 55%, 45% and 35% carbohydrate). Weight, IR, systolic and diastolic BP levels were measured at baseline, 6 and 24 months. After excluding the subjects who took antihypertensive or weight-loss drugs, cross-lagged path and mediation analysis were performed among 540 participants. Results: After adjusting for age, race, sex, and diets groups, the cross-lagged path coefficient from baseline weight to 24-month IR (β 1 =0.135, P =0.04) was significantly greater than the path coefficient (β 2 =0.022, P >0.05) from baseline IR to 24-month weight ( P <0.05 for the difference in βs), indicating that weight-loss preceded change of IR. The mediation effects of 24-month IR on the 24-month systolic BP and diastolic BP were estimated at 17.03%( P =0.004) and 20.94% ( P =0.034), respectively. Conclusions: Our dataindicate that weight loss precedes change of IR in affecting changes of BP in response to weight loss diet interventions. Figure 1 . Covariate-adjusted path coefficients from baseline weight to 24-month IR; β 1 and β 2 are cross-lagged path coefficients; r 2 and r 3 are tracking correlations; The difference between β 1 and β 2 was tested using Fisher Z test; * P<0.01, † P<0.05 for difference between β 1 and β 1 .
What problem does this paper attempt to address?