A controlled-feeding feasibility study of the anti-inflammatory diet in insulin-resistant subjects

Abstract

Listen

Translate article

Purpose To evaluate the feasibility and initial effects of the anti-inflammatory (AI) Diet in type 2 diabetes and pre-diabetes on inflammatory, cardiovascular risk factors and weight outcomes. Methods A randomized, 6-week, controlled-feeding study of the AI Diet compared with an ADA Control Diet: with 2 weeks of isocaloric followed by 4 weeks of ad lib feeding. The AI Diet was based on the recommendations from the naturopathic community emphasizing foods thought to reduce inflammation (fish, nuts, flaxseed, vegetables, and fruits) and excluding foods considered antigenic (wheat, dairy, corn, soy, eggs, red meat, and citrus) as well as alcohol, caffeine, and refined sugar. The Control Diet was based on guidelines for medical nutrition therapy from the American Diabetes Association (ADA). The two diets were matched on total fat, CHO, protein, and cholesterol, but fatty acids, fiber, and micronutrients differed between the diets. Outcomes included fasting glucose (FG), lipids, hs-crp, cytokines, and weight (Wt), measured repeatedly. Participants had FG≥100 and Results Study population ( n =30) included 65% women (mean age 56 years and mean FG 111.7); 26 individuals completed the protocol. FG, TC LDL, HDL, and Wt were reduced in both diet groups; changes ( r ) for AI Diet vs. ADA Control for isocaloric (2 weeks) and ad lib (4 weeks) phases are as follows (see Table 1 ): Table 1 . Change in measures during isocaloric and ad lib feeding study of AI and ADA Control Diet. Measure: ΔFG (mg/dl) ΔTC (mg/dl) ΔLDL (mg/dl) ΔHDL (mg/dl) ΔWt (kg) Δhs-crp (mg) Diet: AI ADA AI ADA AI ADA AI ADA AI ADA AI ADA Isocaloric a −4.4 −4.2 −22.9 13.5 −30.2 −10 −4.8 −7.4 −1 −0.6 −1.6 0 Ad lib b −9.1 −7.5 −24.9 10.7 −31 −20.7 −7.2 −8.8 −3.7 −2.1 −2.2 0.6 a r =End of 2-week isocaloric minus baseline. b r =End of 4-week ad lib minus isocaloric. Parametric analyses revealed that participants on the AI Diet generally achieved better results than those on the Control Diet for lipids (i.e., greater and faster reductions in total and LDL cholesterol, increased HDL/LDL ratios), fasting glucose, and both BMI and weight (all greater and faster reductions). Non-parametric analyses found AI Diet participants more likely to achieve clinical targets (e.g., total cholesterol 0.4, fasting glucose Conclusions Data from this exploratory study show that both the AI Diet and the diabetic Control Diet have beneficial effects on glucose, lipids and weight. Neither diet affected IL6 or TNFa markers of inflammation, however significant decrease in hs-crp was seen in the AI Diet compared with control diet. The AI Diet tended to have a larger effect on weight-loss than the Control Diet. Limitations include: the control diet, which for ethical considerations reflected current guidelines for diabetes treatment rather than a comparison with the “usual American diet”, the short duration of feeding and small sample size.