Outcomes related to inflammation, glycemic control, and diet
Brinkworth et al. (2004), Australia . Long-term effects of advice to consume a high-protein, low-fat diet rather than a conventional weight-loss diet, in obese adults with type 2 diabetes: one-year followup of a randomized trial.
2-group pretest/posttest design with randomization, longitudinal follow-up. Sample: obese or overweight adults with type 2 diabetes.
Dual X-ray absorptiometry. 24-hour urine collection to measure protein intake from urinary urea : creatinine ratio.
(1) 8 weeks of 30% energy restriction, (2) 4 weeks of energy balance (during first 12 weeks met with dietician every 2 weeks), (3) 52-week maintenance period with minimal counseling (weigh-in every 3 months).
No control for confounding variables (i.e. differences in weight loss, changes in medications over the study period). 39% attrition rate, discussed adequately.
Repeated measures ANOVA. Adjustments for baseline values and body weight with ANCOVA. Regression analysis for relationships between changes in parameters over time.
There was a decrease in CRP for both groups of 14% (), but no difference between groups (). No difference overall in glycemic control at 64 weeks from baseline. Used urinary urea : creatinine ratio to determine adherence to the diet, which was adequate (i.e. higher in high-protein diet group).
Giannopoulou et al. (2005), U.S. . Effects of diet and/or exercise on the adipocytokine and inflammatory cytokine levels of postmenopausal women with type 2 diabetes.
3-group pretest-posttest design with randomization. 1 of 3 interventions: (1) only diet, (2) only exercise, or (3) diet and exercise. The intervention lasted 14 weeks. Sample: postmenopausal, obese women with type 2 diabetes.
Inflammatory markers: leptin, CRP, adiponectin, TNF-, and IL-6. All measured by blood samples. Glycemic control: HbA1c, fasting blood glucose, insulin. All measured by blood samples. Diet: 24 hr recall every two weeks, but did a postmeal evaluation of markers. Others: body composition (abdominal fat distribution), metabolic data, EKG.
24 hour recalls. Exercise stress test. BodPod used to determine percentage of body fat. MRI for abdominal fat distribution testing.
(1) Diet only: high MUFA Fat 40%, CHO 40%, Protein 20%, Individual modifications made to energy intake based on energy expenditure from exercise and metabolic rate. Initial nutritional counseling conducted at the beginning and at weekly intervals. (2) Exercise only: 60 min walking 3-4 times per week. (3) Combo.
Exercise intervention was “supervised” but not specified how. Subjects completed 24-hr dietary recalls but no explanation given as to what was done with these. No mention of where subjects came from. Only generalizable to postmenopausal obese women with type 2 DM. CRP correlated with weight, BMI, total body fat mass, SAT, and IL-6.
Descriptive data: 2-way ANOVA with post hoc analysis where appropriate. Covariates: total body fat mass, abdominal fat distribution.
For all groups: CRP decreased pre/post with . IL-6 change pre/post was NS for all groups. HbA1c change was NS for all groups. TNF-α change was NS for all groups.
Marfella et al. (2006), Italy . Effect of moderate red wine intake on cardiac prognosis after recent acute myocardial infarction of subjects with type 2 diabetes mellitus.
RCT. Sample: 131 participants with type 2 diabetes with recent (<2 months) myocardial infarction (MI), <70 yr, and can drink alcohol. All patients were treated with insulin. Baseline demographics between groups were similar. Subjects randomized to intervention or control group for 52 weeks.
Inflammatory markers: TNF-α, IL-6, IL-18, CRP. Glycemic control: insulin sensitivity (HOMA score), insulin, HbA1c. Diet: red wine consumption. Others: nitrotyrosine, BMI, lipids, myocardial synchronization, other measures of functional cardiac outcomes.
24-hr dietary recall: at first visit. 4-day food diary: every three months. Echocardiogram.
Both groups advised 2000 kcal/day diet similar to Mediterranean diet. Wine group: 4-oz. glass of red wine each day. Comparison group: no red wine consumption.
Only 115 patients completed the study (8 died). Both groups given dietary advice, which was not a variable in the analysis.
One-way ANOVA for baseline, then Scheffe’s test to compare pairs of data. For nonnormally distributed data (IL-6, TNFα, IL-18, HOMA, triglycerides), Wilcoxon’s test was used. Linear regression used to assess correlations.
Significant higher levels of TNFα, IL-6, IL-18, CRP, and nitrotyrosine ( for all) in the control group than wine group. Change in HbA1c was NS different between groups.
Wolever et al. (2008), Canada . The Canadian trial of carbohydrates in diabetes (CCD), a 1-y controlled trial of low-glycemic-index dietary carbohydrate in type 2 diabetes: no effect on glycated hemoglobin but reduction in C-reactive protein.
3-group pre-test/post-test design with randomization. Sample: . Subjects were adults age 35–75 y/o with type 2 diabetes and a BMI of 24–40.
Inflammatory markers: CRP. Glycemic control: insulin sensitivity (HOMA score), fasting glucose, fasting insulin, HbA1c. Diet: macronutrient intake. Others: lipid profile, weight, blood pressure, free fatty acids. CRP: log-transformed (those whose statin doses were changed during the study were not included in the analysis of CRP, lipids, and lipoproteins).
3-day food diary records collected 7 times during the 12 month period. Key food diaries from 91% of subjects.
12 months in length. All subjects advised “heart healthy” diet. High-GI diet: avoid low-GI foods, eat low-fat foods. Low-GI diet: exchange high-GI foods for low-GI foods, eat low-fat foods. Low-CHO: decrease SFA intake, increase MUFA intake. Individual dietary counseling provided at 2 and 4 weeks, then every 4 weeks.
No control group.
General linear mixed model. Stats CRP used log-transformed data presented as means and CI’s. Time treated as a regression variable. Model covariates: age, BMI, sex, and center that correlated significantly with the response variable. Time × diet: interactions. *data for subjects that were excluded from analysis were included up until the point they were dropped. Considered MAR.
CRP in low-GI diet 29% lower than high-GI diet. CRP was 1.95 (1.68, 2.27) for the low-GI diet, 2.75 (2.33, 3.24) for the high-GI diet, and 2.35 (2.01, 2.75) for the low-CHO diet ( for difference in groups). CRP increased with high-GI diet. HbA1c rose from 6.1 to 6.3%, but between group effect on HbA1c was NS from diet: high GI HbA1c = 6.34 ± 0.05, low GI HbA1c = 6.34 ± 0.05, low GI HbA1c = 6.35 ± 0.05. Significant effect of time on weight, but not significantly different between diet groups.
Barnard et al. (2009), United States . A low-fat vegan diet and a conventional diabetes diet in the treatment of type 2 diabetes: a randomized, controlled, 74-week clinical trial.
2 group pretest-posttest design. Sample: Subjects were adults with type 2 diabetes. (1) Vegan diet. (2) Conventional ADA diet (2003 guidelines).
Inflammatory markers: CRP. Glycemic control: HbA1c, fasting plasma glucose. Others: lipid profile, weight, BMI, waist circumference, albumin, blood pressure. Dietary adherence: vegan-no animal products in 24-h recalls, low fat, low cholesterol conventional-following prescribed energy restrictions, low SFA. Both groups: attended at least 10 of the 22 weekly dietary info sessions.
Initial 1-hr meeting with dietician for each group. Then, 22 weekly group sessions specific to group. Optional sessions every two weeks for the remaining 52 weeks. Medications were adjusted for safety by an endocrinologist.
Not all analyses were performed only using those participants with good adherence. Good adherence not defined. Only CRP measured, was not a primary focus of the study.
Repeated measures ANOVA for HbA1c (as ITT) using time, diet group, and interaction between time × diet with HbA1c as DV. Between subject -tests for all DVs to detect significant changes from baseline to 74 weeks (or until last measurement) and paired-comparison -tests.
After adjustment for medication changes, LDL and non-HDL cholesterol values decreased, and more so in the vegan group. CRP decreased significantly in both groups, but no difference between groups.
Dostlova et al. (2009), Czech Republic . Increased serum concentrations of macrophage inhibitory cytokine-1 in patients with obesity and type 2 diabetes mellitus: the influence of very-low-calorie diet.
Quasi experimental one group with two comparison groups (monitored in the hospital). Sample: total Group no. 1 (comparison): () obese, nondiabetic women. Group no. 2 (treatment group): () obese women w/type 2 DM. Group no. 3 (comparison): () healthy lean women.
Inflammatory markers: CRP. Glycemic control: glucose, homeostasis model assessment of insulin resistance (HOMA-IR). Others: MIC-1 levels lipids, BMI, body fat composition. Another primary focus of the study was to test levels of macrophage inhibitory cytokine-1 (MIC-1) and mRNA expression of MIC-1 in subcutaneous and visceral fat.
No measure of adherence to diet.
2-week very-low-calorie diet. Only obese women with and without DM were given the IV diet; a 2-week 550 kcal/day diet. All patients were hospitalized.
Highly controlled environment for study. No adverse events reported. Not a sustainable intervention for long-term use. The substudy examining MIC-1 mRNA expression was totally unrelated to the 3 groups receiving the intervention, and these patients were receiving surgery and cannot be compared.
One-way ANOVA followed by post hoc tests. Spearman correlations performed to evaluate the relationship of MIC-1 expression to other variables.
For women with T2DM (), significant changes in the following parameters were observed after the VLCD: BMI, cholesterol, TGs, glucose, insulin, CRP, and HOMA-IR.
Kozłowska et al. (2010), Poland . Adiponectin, resistin, and leptin response to dietary intervention in diabetic nephropathy.
Quasiexperimental one group pretest/posttest design. Sample: Subjects were obese with T2DM, all took antihypertensives, antidiabetic drugs, phosphate binders, and diuretics. Study duration was 8 weeks.
Inflammatory markers: TNF-α, CRP Glycemic control: insulin sensitivity (HOMA score), insulin, HbA1c. Diet: not measured. Others: adiponectin, resistin, leptin, total protein, metabolic rate using oxygen and CO2 measurements.
DEXA scan for body composition.
Intervention diet: 20% energy deficit 0.8–1 g/kg ideal body weight, 30% cal from fat, 60% cal from CHO. Consult with dietician Q2wk.
No mention of how dietary intake was measured, just reported in a chart (Table 2). No control group. Small sample size, weak design.
Pearson or Spearman’s tests for correlations. Repeated measures. Student’s -test or Wilcoxon matched pairs.
No correlations between food intake and inflammatory markers made. After intervention, resistin () and TNF-α () concentrations were significantly decreased. Resistin concentrations were correlated with TNF (, ).
Vetter et al. (2010), U.S. . Effect of a low-carbohydrate diet versus a low-fat calorie-restricted diet on adipokine levels in obese diabetic participants.
Ancillary group in an RCT. Sample: . Participants were obese, with type 2 diabetes, outpatient. (1) Low-carb diet. (2) Low-fat diet.
Inflammatory markers: leptin, adiponectin, TNF-α. All measured by blood samples. Glycemic control: HbA1c, insulin, fasting blood glucose. All measured by blood samples. Diet: 24 hr recall. Others: weight, BMI, race, age, gender, diabetes related medications, presence of CAD, smoking status.
24-hour recalls. Nutribase Management software to establish micro/macro nutrient content of intake.
Randomly assigned to groups. First month: dietician-led weekly group sessions for each group. Following months: monthly group sessions. Given handouts, sample menus, recipes, etc.
24-hour recall assessed, but not used to assess adherence to the prescribed diet. High attrition rate (54.9%). Demographic variables between those who did and did not complete study were not significantly different. No discussion of how missing data were handled. No difference in macronutrient intake between groups, although it should have been.
Log-transformed data: leptin, adiponectin, and TNFα. Repeated measures ANOVA (with time × diet group) used to compare change in weight, adipokines, and dietary intake. Multiple linear regression to describe associations between weight loss and changes in adipokine values.
Inflammatory markers: CRP not measured. TNF-α: decrease did not differ with time () or between groups (). Glycemic control: HbA1c: decline in HbA1c was not significantly different between groups (no value provided). Others: weight: declined significantly over time (), however, it was not different between groups (). Leptin: decline did not differ significantly between groups. Adiponectin: increase did not differ significantly between groups Weight loss significantly associated with changes in leptin () and TNF-α ).
Azadbakht et al. (2011), study conducted in Iran . The Dietary Approaches to Stop Hypertension eating plan affects C-reactive protein, coagulation abnormalities, and hepatic function tests among type 2 diabetic patients.
Crossover intervention design with randomization. Sample: , Convenience sample of adults with type 2 diabetes at the Shaheed Motahari Hospital of Fooladshahr.
3d dietary recall to assess adherence to the prescribed diet. Monthly 3-day physical activity diary.
8-week crossover study with randomization to control or DASH diet. Control diet had fewer PUFA. Energy intake individually tailored. Monthly visits. 4-week washout period.
Unclear how study was conducted, not easy to follow methods. R/o carryover effects. No mention of how long each diet was maintained, if there was a washout period, or why those who had left the protocol left (r/o fatigue effect).
Assessing effect of intervention: paired -test with and without adjustment for weight. Groups compared using a % change calculation. Patients who “deviated from the study protocol” were not included in the analysis (p. 1084).
Patients who received the DASH diet showed a reduction in CRP, plasma fibrinogen, and liver transferase enzymes. These reductions were significant and remained significant after adjusting for the effect of weight.
Bozzetto et al. (2011), Italy . The association of hs-CRP with fasting and postprandial plasma lipids in patients with type 2 diabetes is disrupted by dietary monounsaturated fatty acids.
Preexperimental posttest only Cross-over with randomization (no washout). Sample: 12 participants with type 2 diabetes HbA1c in good control. 4 weeks on each diet. No washout period mentioned.
Inflammatory markers: CRP. Glycemic control: not measured. Diet: not measured. Others: triglyceride-rich lipoproteins (chylomicrons, large VLDL, small VLDL), cholesterol and triglyceride.
Nothing provided in paper.
Diets were isoenergetic: both had kcal of 948 per day. Randomized, crossover design. CHO/fiber/low-GI: C 52%, MUFA 17%,
High-MUFA diet: C 45%, MUFA 23%,
No details given concerning dietary advisement. Dietary adherence not measured. Confounding variables not explored. Body weight did not change significantly over the study period, but it should have changed, considering the limited caloric prescription.
Paired -tests to identify differences in outcome measures between the two diets (groups were merged for analysis). Repeated measures ANOVA to evaluate differences in postprandial testing (multiple tests). They claim no carryover effect was detected.
Diet and inflammation: after intervention, fasting CRP values were not significantly different between study diets, although CRP did decrease significantly after the MUFA meal (), not the CHO/fiber meal.
Davis et al. (2011), . Differential effects of low-carbohydrate and low-fat diets on inflammation and endothelial function in diabetes.
2-group pre-test/post-test clinical trial subgroup of larger trial. Sample: subjects were obese with type 2 diabetes with an HbA1c between 6% and 11%. 63% of subjects were Black in both diet groups, in the low-carb group, 11% of the subjects were Hispanic.
EndoPat to measure peripheral microvascular endothelial function.
Total 24 weeks: subjects randomized to receive either a low-carbohydrate, Atkins-style diet or a low-fat diet (similar to DPP diet). Participants received structured menus for the 1st two weeks. Medications were adjusted based on a predefined algorithm. Diet was reinforced every 6 weeks at scheduled visits.
No control group. No measure of adherence to prescribed diet.
CRP and IL-6 were log-transformed. Unpaired -tests (or Wilcoxon rank) used to detect differences in parameter outcomes between dietary arms. Pearson or Spearman correlations.
Low-fat diet: CRP decreased from 4.0 to 3.0 (). Low-carb diet did not cause a significant change in CRP (). Low-carb diet: sICAM decreased from 234 to 199 (), E-selectin decreased from 92 to 82 ().
Itsiopoulos et al. (2010), Australia . Can the Mediterranean diet lower HbA1c in type 2 diabetes? Results from a randomized cross-over study.
Randomized cross-over (2 groups pretest, posttest). Sample: Subjects recruited from newspaper had “well-controlled” type 2 diabetes.
Health and lifestyle questionnaire (not defined). Seven-day diet record (completed by participant). Used changes in plasma levels of several different carotenoids to check compliance with the IV diet.
Intervention was a Mediterranean diet for 12 weeks, then crossover (24 weeks total). 70% of intervention diet meals prepared and provided in bulk for free. Advised to eat 3 servings fruits per day.
No washout period between diet periods. Measurement tools not standardized. Olive oil company provided the olive oil, may lead to biased results. Underpowered.
Data from the two diets were pooled. Repeated-measures ANOVA. Used an interaction term to control for absence of washout period.
Significant reduction in HbA1c in the intervention diet. Trend toward decrease in HOMA and BMI. Increase in plant food intake.
Khoo et al. (2011), Australia . Comparing effects of a low-energy diet and a high-protein low-fat diet on sexual and endothelial function, urinary tract symptoms, and inflammation in obese diabetic men.
2 group pretest-posttest design with randomization. Sample: convenience sample of 31 White men with type 2 diabetes and erectile dysfunction. Randomly assigned to either modified low-calorie diet (LCD) or high-protein low-fat diet (HP diet) for 8 weeks. After first 8 weeks, all subjects were placed on the HP diet for the remaining 44 weeks.
Inflammatory markers: CRP, IL-6, and soluble E-selectin. Glycemic control: quantitative insulin sensitivity check index (QUICKI), glucose, insulin. Diet: compliance, not used in analysis. Others: androgen levels, erectile function, sexual desire, and lower urinary tract symptoms.
Food diaries to monitor compliance. IIEF-5, Sexual Desire Inventory, International Prostate Symptom Scale.
Diet compliance monitored every 2–4 weeks. LCD: 2 liquid meals per day plus small meal for total of 900 kcal/day. HP diet: 300 g lean meat/poultry/fish, 3 servings/day of cereals/bread and low-fat dairy foods, and two fruit and vegetable servings/day. Should induce a 600 kcal deficit from diet. All subjects received menu plans, recipes, and cooking advice. Activity was not restricted.
Adherence not reported or used in analysis, although diet diaries were used. Major focus of the study was sexual and endothelial function. Handling of missing data not discussed.
Maximum likelihood repeated measures mixed models to compare pre/post measures. Used time × diet interaction. Post hoc pairwise tests made.
Inflammation: significant decrease in CRP (at 8 and 52 wk), E-Selectin (at 8 and 52 wk), and IL-6 (at 8 wk only) in HP diet group (), but not LCD group. Glycemic control: insulin sensitivity (QUICKI) significant improved in LCD group, but not HP diet (). Plasma glucose did not improve significantly in either group. Other: significantly weight loss in both groups.
Mraz et al. (2011), Czech Republic . The effect of a very-low-calorie diet on mRNA expression of inflammation-related genes in subcutaneous adipose tissue and peripheral monocytes of obese patients with type 2 diabetes.
Quasiexperimental design with two comparison groups. Sample: women only. The intervention was only implemented in the 12 obese women with T2DM. Two other control groups: 15 healthy women, and 8 obese women without T2DM. Intervention lasted 2 weeks and was conducted in a hospital setting for close monitoring.
Only obese women with DM were given the IV diet; a 2-week 600 kcal/day diet. All patients were hospitalized.
No adverse events or medication changes reported from the intervention period. Inadequate detail of the intervention. No discussion of possible causes for the abnormal finding of increased TNF following the intervention.
Paired -test or Wilcoxon signed-rank test used to detect differences between measurements pre/post intervention.
TNF-α levels were significantly increased in T2DM women after 2 weeks of following the VLCD ( for pre/post Δ). HbA1c not assessed after intervention (only 2 weeks). Δ in CRP pre/post = .94 mg/L, . Δ in IL-6 pre/post = .8 pg/mL, .
Åsgård et al. (2007), Sweden . High intake of fruit and vegetables is related to low oxidative stress and inflammation in a group of patients with type 2 diabetes.
Cross-sectional design. Sample: 53 participants recruited from newspaper. Type 2 diabetes with HbA1c <10%, 40–75 yr, BMI < 35, and stable body weight ×3 months.
3-day diet recall with precoded instrument that was altered for this study (originally it was for 7 days). Program used to determine micro and macronutrient content from food recalls, name of program not provided.
No limitations section in paper. No adjustments for age or weight. The study participants were similar, but weight higher in females. This participant group answered a newspaper, may be more interested in diet, their diet was in accordance with the Nordic recommendations. Lower mean HbA1c at baseline (<6.2%).
Spearman correlation coefficients with considered significant. For correlations with inflammation or oxidative stress, significance value was set to <.05. Wilcoxon two sample tests used with non-normally distributed data to determine differences between genders.
IL-6 lower with higher levels of carotenoids (), but not with the larger food groupings of fruits and vegetables. -tocopherol had positive correlation with CRP.
Qi et al. (2006), U.S. . Whole-grain, bran, and cereal fiber intakes and markers of systemic inflammation in diabetic women.
Nested cohort in longitudinal analytic study of the Nurses’ Health Study. Sample: Women between 30–55 with type 2 diabetes as defined by National Diabetes Data group (diagnosis made prior to 1997).
Inflammatory markers: CRP, TNF-R2, Glycemic control: not assessed. Diet: glycemic load, glycemic index, quintiles of increasing intake of whole grains, bran, cereal fiber, total fiber and germ Q1–Q5. Others: endothelial function (s-ICAM, soluble E-selectin) oxidative stress and chromosomal damage markers.
Semiquantitative food frequency questionnaire. Questionnaire from the Nurse’s Health Study to collect anthropometric and lifestyle data.
“Whole grains were previously described” but the paper referred to was written by different authors and was in a study about men. No limitations section. Data collected between 1989-1990. Only women included.
Associations evaluated with linear regression. Intake of dietary variables assigned to quintiles. Inflammatory values log-transformed and median value assigned for each of the 5 quintiles. Adjusted for age, BMI, smoking, alcohol, physical activity, aspirin use, HbA1c, dx of hypertension or hypercholesterolemia, postmenopausal hormone use, glycemic index and magnesium.
Women with higher intakes of whole grains, bran, and cereal fiber had lower levels of CRP and TNF after adjusting for age, BMI, lifestyle, and dietary covariates.