Diabetes
Abstract & Commentary
Prevention of Diabetes with Mediterranean Diet
By Traci Pantuso, ND
Adjunct Faculty, Bastyr University, Seattle, WA
Dr. Pantuso reports no financial relationships relevant to this field of study.
Synopsis:Previously, the authors reported the preliminary data from one of the 11 PREDIMED sites demonstrated that the Mediterranean diets enriched with high-fat vegetable foods decreased the incidence of diabetes. In this subgroup analysis, the authors report the data from all 11 PREDIMED sites demonstrating a decreased risk of diabetes with the use of a Mediterranean diet pattern in persons with high risk of cardiovascular disease.
Source:Salas-Salvado J, et al. Prevention of diabetes with Mediterranean diets: A subgroup analysis of a randomized trial. Ann Intern Med 2014;160:1-10.
Summary Points
- After controlling for confounding variables, a 30% relative risk reduction for diabetes was found in the merged Mediterranean diet groups vs the control diet.
- The mean scores of both the Mediterranean diet groups' adherence increased compared to the control group for all yearly comparisons (P < 0.010).
Lifestyle modifications have demonstrated efficacy in the prevention of diabetes. Limited
studies have investigated whether dietary pattern changes without energy restriction and increased physical activity also may be an effective intervention in the prevention of diabetes. The Mediterranean dietary pattern has been associated with a decreased risk of diabetes.1 The Mediterranean dietary pattern consists of large quantity and variety of plant-derived foods including whole grains and cereals, raw and cooked vegetables, fresh and dried fruits, fish, with nuts and olive oil as added fats, and a moderate intake of meat and dairy products with moderate amount of wine during meals.2
Methods
The PREDIMED study is a parallel-group, randomized, primary cardiovascular prevention trial conducted in Spain at 11 recruiting centers. The study included persons at high risk but without cardiovascular disease (CVD) at baseline. Participants were randomized to one of three of the following nutrition interventions: Mediterranean diet supplemented with extra-virgin olive oil (EVOO), Mediterranean diet supplemented with mixed nuts, or a control diet with the guidance to reduce intake of all types of fat. There were no energy restrictions or physical activity recommendations for any of the intervention groups.
The participants consisted of community-dwelling men (aged 55-80) and women (aged 60-80) without CVD who had either type 2 diabetes or at least three or more of the following cardiovascular risk factors: hypertension, hypercholesterolemia, low high-density lipoprotein (HDL) cholesterol levels, current smoking, overweight or obese, and family history of premature CVD. A total of 7447 eligible participants were enrolled in the trial from October 2003 to June 2009. Each participant was randomized to one of the three nutrient interventions by computer-generated random numbers at the recruitment site.
This subgroup study only included participants who did not have diabetes at baseline and had adequate follow-up. The authors calculated that 1130 participants were needed per group to obtain a statistical power greater than 90% with a two-tailed α = 0.05. The expected proportions of new diabetes cases were 11% and 7% in the control and intervention groups, respectively (relative risk, 0.64).
Mediterranean Diet Intervention
Dieticians gave personalized advice in both individual and group formats to participants regarding their dietary intervention at baseline and quarterly thereafter. A 14-item questionnaire was used to evaluate dietary adherence at each session and personalized advice was given to increase adherence to the Mediterranean diet interventions. Participants enrolled in the Mediterranean plus nut group received 30 g/day of mixed nuts (15 g of walnuts, 7.5 g of hazelnuts, and 7.5 g of almonds) and the EVOO group received 50 mL/day of EVOO at no cost to the participants.
Control Group Intervention
The same 14-item questionnaire as in the Mediterranean diet intervention was used yearly in the control group. The control group received guidance to reduce intake of dietary fat from animal and plant sources. Through 2006, participants in the control group only received a leaflet describing the low-fat diet. In 2007, participants in the control group also received personalized dietary advice and were invited to group sessions with the same intensity and frequency as those in the Mediterranean groups. A separate nine-item questionnaire was used to assess dietary adherence to the recommended low-fat diet.
A number of questionnaires were used to evaluate diet, physical activity, and other lifestyle factors at baseline and yearly during follow up. A 137-item validated semi quantitative food frequency questionnaire was used to evaluate diet. The Minnesota Leisure time physical activity questionnaire, which was validated in Spanish, was used to evaluate physical activity. A 47-item questionnaire regarding education, lifestyle, medical history, and medication use was used.
Statistical Analysis
To assess hazard ratios (HR) for Mediterranean diet groups in comparison to the control group, the authors used Cox regression models. They also conducted a crude age- and sex-adjusted model, and two other multivariate models were used for analysis to control for multiple variables. In the multivariate model A, age, sex, and body mass index (BMI) were adjusted. In the multivariate model B, age, sex, baseline smoking (never, current, former), fasting glucose level, presence of dyslipidemia or hypertension, total energy intake level (kcal/d), adherence to Mediterranean diet (14-item questionnaire), physical activity level, education level, and alcohol intake were adjusted.
All P values are 2-tailed at < 0.050. All analyses were performed using SPSS version 19 and Stata version 12.1.
At baseline, participants underwent a number of clinical and laboratory assessments including: electrocardiography, anthropometric measurements, and blood pressure measurements. Fasting blood and spot urine were assessed at baseline and follow-up years 1, 3, 5, and 7. Urine hydroxytyrosol and plasma alpha-linolenic acid were used as biomarkers to assess supplemental diet adherence during the first 5 years and follow up.
New onset diabetes was diagnosed during the follow-up period (October 1, 2003 to December 1, 2010) using the American Diabetes Association criteria: fasting plasma glucose levels ≥ 126.1 mg/dL or a 2-hour fasting blood glucose levels of ≥ 200.0 mg/dL after a 75 g oral glucose load.
Results
A total of 7447 participants were enrolled in the PREDIMED study and 3833 did not have diabetes at baseline. A total of 3541 of the 3833 participants had available information during follow-up to determine if they had diabetes and were eligible for this subgroup analysis.
A total of 252 participants were lost to follow-up for more than 2 years: 4.1% in the Mediterranean supplemented with EVOO, 6.9% in the Mediterranean supplemented with nuts, and 10.5% in the control diet groups.
At baseline, clinical characteristics of age, sex, BMI, weight, mean waist circumference, mean waist-height ratio, tobacco use, marital status, education level, hypertension status, dyslipidemia status, medication use (antihypertensives, statins, or other hypolipidemic drugs), fasting glucose level, total cholesterol, HDL and low-density lipoprotein cholesterol, triglycerides, non-HDL cholesterol, mean physical activity level, mean total caloric intake, and mean Mediterranean diet adherence scores were similar in all groups.
During the follow-up period
After a median follow-up of 4.1 years, 273 participants were diagnosed with new-onset diabetes — 80 participants in the Mediterranean diet plus EVOO (6.9%), 92 participants in the Mediterranean plus mixed nuts (7.4%), and 101 participants in the control group (8.8%) (see Table 1).
Table 1. Incidence of Diabetes
| Groups | Med Diet + EVOO | Med Diet + nuts | Control Diet |
| n = | 1154 | 1240 | 1147 |
| New Cases of Diabetes | 80 (6.9%) | 92 (7.4%) | 101 (8.8%) |
| Cumulative incidence (95% CI) | 6.93 (5.53-8.55) | 7.42 (6.02-9.02) | 8.81 (7.23-10.60) |
The HRs for diabetes in the Mediterranean diet group supplemented with EVOO compared to the control diet group were 0.60 (95% confidence interval [CI], 0.43-0.85) after adjusting for potential confounding variables (see Table 2).The Mediterranean diet group supplemented with nuts compared to the control diet group had a HR of 0.82 (CI, 0.61-1.10) after adjusting for potential confounding variables (see Table 2).
After controlling for confounding variables, a 30% relative risk reduction for diabetes in the merged Mediterranean diet groups vs control (HR, 0.70; CI, 0.54-0.92) was found (see Table 2).
Table 2. Hazard Ratios of Diabetes
| Raw Data HR (95% CI) | Adjusted for Age and Sex HR (95% CI) | Multivariate A* Adjusted HR (95% CI) | Multivariate B** Adjusted HR (95% CI) | |
|
Med Diet + EVOO vs Control Diet |
0.69 (0.51-0.92) | 0.68 (0.51-0.92) | 0.68 (0.51-0.92) | 0.60 (0.43-0.85) |
|
Med Diet + Nuts vs Control Diet |
0.81 (0.61-1.08) | 0.80 (0.60-1.06) | 0.82 (0.61-1.09) | 0.82 (0.61-1.10) |
|
Both Med Diets vs Control Diet |
0.75 (0.58-0.96) | 0.74 (0.58-0.95) | 0.75 (0.58-0.96) | 0.70 (0.54-0.92) |
*Multivariate A Model adjusted for age, sex, and BMI.
**Multivariate Model B adjusted for age, sex, baseline smoking (never, current, former), fasting glucose level, presence of dyslipidemia or hypertension, total energy intake level (kcal/d), adherence to Mediterranean diet (14-item questionnaire), physical activity level, education level, and alcohol intake.
The mean scores of both the Mediterranean diet groups' adherence increased in both Mediterranean diet groups compared to the control group for all yearly comparisons (P < 0.010). As would be expected, both the Mediterranean diet groups had more participants with a diet score of ≥ 10 than in the control group (P < 0.010) indicating that the participants in the Mediterranean groups were adhering to their diets and that there were no significant changes in the control group diet.
Urinary hydroxytyrosol and Plasma alpha-linolenic acid biomarker levels
Participants in the Mediterranean diet with EVOO had increased urinary hydroxytyrosol levels compared to baseline at 3-year follow-up (P < 0.050), indicating that they were consuming the EVOO. The plasma alpha-linolenic acid levels increased in the Mediterranean diet with mixed nuts group at 3-year follow-up (P < 0.050) indicating that the participants were consuming the mixed nuts.
No changes were found in urinary hydroxytyrosol or plasma alpha-linolenic acid in the control group. No significant difference in body weight, waist circumference, physical activity levels, or medications that could potentially affect the development of diabetes (estrogens, corticoids, antiepileptic drugs, statins, and antihypertensives) were found between groups.
Commenatary
This subgroup analysis demonstrates more evidence that a Mediterranean diet decreases the incidence of type 2 diabetes in older white persons at high risk for CVD, without addressing physical activity or energy restrictions. However, the study has a number of limitations, including that it is a subgroup analysis and it was a secondary endpoint of the PREDIMED study, with the primary endpoint being CVD risk.
In addition, participants who withdrew from the study had a worse cardiovascular risk profile at baseline than the participants who completed the study, which may have affected the data. The study personnel and participants were not blinded to group allocation, as there were noticeable food differences. The PREDIMED investigators who diagnosed new-onset diabetes were blinded. The authors also state that they "cannot discard measurement errors affecting physical activity and alcohol intake during follow up." Overall, the strengths of this trial outweigh the limitations in that it was randomized, the treatment groups were well powered and balanced, and the authors did control for confounding variables. There was an adequate follow-up time of 4.1 years.
One interesting finding of this study is the adherence to the Mediterranean diet groups increased compared to the control group in yearly comparisons. This study had regular meetings between the participants and the dieticians, the participants also filled out questionnaires frequently to ascertain their adherence to the diet, and the Mediterranean groups had access to the EVOO and mixed nuts. This appears to be a great strategy in teaching patients and increasing their adherence to the Mediterranean diet pattern for the prevention of not only diabetes, but for a long, healthy life.
There do not appear to be any reasons not to recommend the Mediterranean diet pattern to both healthy people and those at risk for diabetes or cardiovascular disease. The addition of good quality olive oil and mixed nuts may increase food costs and should be taken into account on a patient-by-patient basis. In addition, should these research results be extrapolated to patients who already have diabetes, the Mediterranean diet pattern may need to be altered slightly by determining the ideal amount of carbohydrates in a specified portion size for individual patients. Also, patients who currently are taking medications will need to have adequate follow up to ensure that their blood sugar values are within the recommended range. The Mediterranean diet is a good recommendation as a healthy dietary pattern, and in conjunction with proper portion control and exercise, it can be safely recommended to most patients.
References
- Olubukola A, et al. Systematic review and meta-analysis of different dietary approaches to the management of type 2 diabetes. Am J Clin Nutr 2013;97:505-516.
- Hoffman R, Gerber M. Evaluating and adapting the Mediterranean diet for non-Mediterranean populations: A critical appraisal. Nutr Rev 2013;71:573-584.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content
Already have an account? Log in