Improving Cardiometabolic Risk Parameters With Diet
Volunteers without any known medical condition (36 females, 8 men) were recruited through advertisement in local newspapers and were informed orally and in writing of the project disposal, emphasizing the non weight-loss character of the investigation. Inclusion criteria were age between 50 and 73, body mass index in the 25–33 kg/m range and fasting plasma glucose value ≤ 6.1 mmol/L. Only medications accepted were hormone replacement for thyroid problems (if constant during the whole trial; 1 female subject) and prescription-free painkillers without antinflammatory action. Baseline data collected at the time of the first clinical visit are shown in Table 1.
The study was designed as a randomized, controlled, crossover trial of the effect of a multifunctional AD on markers related to cardiometabolic risk. Participants were randomly assigned to one of two treatment orders starting with the AD or the CD, respectively. Each diet phase lasted 4 wk with a 4-wk washout period. The whole trial included four clinical visits, one before and one after each intervention period. During the first visit the participants underwent a physical examination, including auscultation of heart and lungs.
On each clinical visit, fasting body weight was recorded and BP was measured twice, while seating, in the non-dominant arm with a mercury sphyngomanometer. Venous blood was then drawn for the assessment of fasting blood glucose, insulin, HbA1c, cholesterol (total, LDL and HDL), triglycerides, high sensitivity CRP (hs-CRP), PAI-1, IL-6, TNF-α, FFA, apo A1, apo B.
Consumption of dietary supplements such as fish oil (9 subjects), probiotic-containing products (11 subjects) and herbal extracts (11 subjects), was stopped 2 weeks before the start of the trial. Before each dietary phase, the participants attended an introductory session with the nutritionist, who explained practical details of the upcoming diet period.
Volunteers were asked to maintain their normal physical activity regime during the study. A pretrial questionnaire indicated that 48% of the subjects had physical activity equivalent to 1 h/day or more, 30% had 30–60 min/day and 22% reported a low activity level (< 30 min/day). Subjects were also instructed to record their fasting body weight weekly. Variations larger than 1 Kg were reported to the nutritionist, who suggested compensatory dietary modifications. The intake of key components in AD was not affected by these modifications.
Except for almonds and fresh, frozen and smoked fish, the active food items required for the 4-wk AD were provided. Some of the foods included in the CD were also provided. In order to assess dietary compliance, participants completed a daily menu checklist covering each 4-wk diet period. They also completed a questionnaire investigating their experience with each diet. Coaching support was provided by the nutritionist, who contacted each participant at least once per diet period. The study was approved by the Regional Ethical Review Board, Lund, Sweden (Dnr 593/2008).
Participants ate their habitual diet before and between the experimental periods. A dietary habit questionnaire completed before randomization to the starting dietary treatment was used to help the participants to resume their eating habits during the washout period. To ensure good compliance the participants were provided with a detailed 2-wk rotating menu plan for each dietary period (AD and CD) with all foods ingredients expressed in weight and/or volume measures. Electronic self-taring scales were made available whenever needed. The menu plan also included recipes for preparing meals.
The nutritional profiles of CD and AD are shown in Table 2. Both diets were designed in close agreement with the Nordic Nutrition Recommendations and supplied 2,500–2,600 Kcal/d for men and 2,000–2,100 Kcal/d for women, combining foods from plant and animal origin. Both diets incorporated commercial foods available in food stores, but the AD also included prototypes. For a detailed list of products included in AD see additional file 1.
AD combined several functional concepts with potential to modulate different physiological variables related to the inflammatory tonus and cardiometabolic risk, including:
a) food items naturally rich in antioxidants which, in addition to the anti-inflammatory action of their antioxidants, contain phenolics that may improve BP and blood lipids.
b) omega-3 fatty acids, especially those long-chained present in oily fish, which have antiinflammatory and triglyceride-lowering properties. In addition, the overall fat quality was also a concept included in the diet design. Thus, the unsaturated-to-saturated fat ratio was larger in AD than in CD (3.6 vs 1.2, respectively; Table 2).
c) ingredients with potential to beneficially affect the gut microbiota: a probiotic strain (Lactobacillus plantarum Heal19, DSM 15313), and prebiotics, i.e. beta-glucans and resistant starch in intact barley kernels, whole kernel rye flour and isolated barley fiber, which were used for baking an experimental beta-glucan-rich bread. Additional sources of viscous fermentable dietary fibre were a prototype oat-based fiber drink, a rye/oat breakfast cereal and an oat-based muesli.
d) low glycemic impact foods/meals were included for their association with reduced risk for the MetS and Type-2 diabetes, and its perceived ability to ameliorate the inflammatory tonus in healthy individuals. Low GI foods were represented by products with high beta-glucan content which, besides their prebiotic role, may improve glycemic regulation in a 10-h post-ingestion perspective, through mechanisms related to colonic fermentation and lowered inflammatory tonus. AD also included another high-fibre bread, baked from a wheat flour/guar gum blend: This item promotes a low glycemic response, with reduced peak and prolonged duration of net increment above fasting glucose levels (A. Nilsson, K. Radeborg et al., unpublished results). Additionally, AD included ingredients that lower the post-meal glycemic excursion, such as whey protein and vinegar.
e) ingredients with acknowledged abilities to normalize blood levels of both total and LDL cholesterol were also provided: different soybean products, a margarine enriched in stanol esters and dry almonds.
The mean daily quantities of the different functional ingredients in AD and the main functional properties considered for their selection are summarized in Table 3. None of the active ingredients were included in the CD, except for minor amounts of ω-3 fatty acids.
Participants were provided with a 14-day rotating menu plan. For representative 1-day menu plans for the CD and AD see additional file 2. A limited amount of alcohol-containing drinks were allowed during both dietary periods (30 and 37 g ethanol/wk for women and men, respectively). These limits, however, did not compel low-drinkers to increase their habitual alcohol consumption. Due to the low overall energy contribution from such contingent alcoholic drinks, they were not included in the estimation of the energy content of diets. The volunteers' coffee and tea drinking habits were not modified during the trial.
Routine blood tests were analyzed at the Clinical Chemistry Laboratory/Skåne University Hospital Lund, on fasting plasma (total and HDL cholesterol, triacylglycerols, apo A-1, apo B, high-sensitivity C-reactive protein), serum (insulin) or on total blood samples (HbA1c). LDL cholesterol concentrations were calculated. HOMA-IR was calculated from fasting blood plasma-glucose and serum insulin values.
Venous plasma glucose concentrations were measured immediately after bleeding (HemoCue®B-glucose, HemoCueAB, Ängelholm, Sweden). Serum concentrations of TNF-α and IL-6 were measured with a chemiluminescent immunometric assay, Immulite®/Immulite® 1000 TNF-α (Siemens, Deerfield, IL, USA) and a sensitive enzyme-linked immunosorbent assay, Quantikine® HS (RD Systems Inc, Abingdon, UK), respectively. Samples with high IL-6 concentration (> 10 ng/L) in the ELISA were also measured with the Immulite®/Immulite®1000 IL-6 assay (Siemens, Deerfield, IL, USA). Serum FFA were assessed with an enzymatic colorimetric assay kit (Wako Chemicals GmbH, Germany). Plasminogen activator inhibitor (PAI-1) activity in plasma was measured with TriniLize PAI-1 Activity kit (Trinity Biotech, Jamestown NJ, USA).
The diet compositions were analyzed using the 2009 food database from the Swedish National Food Administration and a computerized calculation program (Dietist XP 3.1; Kost och Näringsdata AB, Bromma, Sweden).
The results are expressed as means ± SEM. Data were evaluated by mixed model ANOVA with sequence and interaction between diet and start/end of treatment periods as fixed effects, and participants within sequence and visit as random effects. Least square means were estimated for the start (Active-week 0, Control-week 0) and end values (Active-wk 4, Control-wk 4) for each diet. Least square means and confidence intervals were calculated for the differences between Active/wk 0, Control/wk 0, Active/wk 4 and Control/wk 4, as well as for the net difference between the diets, i.e. (Active/wk 4 - Active/wk 0) - (Control/wk 4 - Control/wk 0). Another set of analyses with body weight included as a continuous covariate was performed. Analyses were carried out using SAS PROC Mixed (v. 8.2, SAS Institute Inc., Cary, NC, USA). Ten-year coronary heart disease risk was calculated both with the Framingham Study equation considering age, gender, total cholesterol, HDL-cholesterol, smoking and systolic BP values, and the Reynolds Risk Score, which also incorporates CRP values.
Primary outcome measure was change in plasma LDL cholesterol. Assuming a 0.5 mmol/L (approximately 10%) post-diet difference and a 0.97 SD, with α = 0.05 and 1-β = 0.8, a minimum of 30 participants were required. Among additional measures, changes in hs-CRP were also taken into account. Assuming a 0.44 mg/L (10%) post-diet difference and a SD of 1.0, the required sample size was 42 subjects. Data obtained from hypercholesterolemics with normal CRP levels confirmed that this n value provides sufficient power for assessing CRP changes.
Participants and Methods
Participants
Volunteers without any known medical condition (36 females, 8 men) were recruited through advertisement in local newspapers and were informed orally and in writing of the project disposal, emphasizing the non weight-loss character of the investigation. Inclusion criteria were age between 50 and 73, body mass index in the 25–33 kg/m range and fasting plasma glucose value ≤ 6.1 mmol/L. Only medications accepted were hormone replacement for thyroid problems (if constant during the whole trial; 1 female subject) and prescription-free painkillers without antinflammatory action. Baseline data collected at the time of the first clinical visit are shown in Table 1.
Study Protocol
The study was designed as a randomized, controlled, crossover trial of the effect of a multifunctional AD on markers related to cardiometabolic risk. Participants were randomly assigned to one of two treatment orders starting with the AD or the CD, respectively. Each diet phase lasted 4 wk with a 4-wk washout period. The whole trial included four clinical visits, one before and one after each intervention period. During the first visit the participants underwent a physical examination, including auscultation of heart and lungs.
On each clinical visit, fasting body weight was recorded and BP was measured twice, while seating, in the non-dominant arm with a mercury sphyngomanometer. Venous blood was then drawn for the assessment of fasting blood glucose, insulin, HbA1c, cholesterol (total, LDL and HDL), triglycerides, high sensitivity CRP (hs-CRP), PAI-1, IL-6, TNF-α, FFA, apo A1, apo B.
Consumption of dietary supplements such as fish oil (9 subjects), probiotic-containing products (11 subjects) and herbal extracts (11 subjects), was stopped 2 weeks before the start of the trial. Before each dietary phase, the participants attended an introductory session with the nutritionist, who explained practical details of the upcoming diet period.
Volunteers were asked to maintain their normal physical activity regime during the study. A pretrial questionnaire indicated that 48% of the subjects had physical activity equivalent to 1 h/day or more, 30% had 30–60 min/day and 22% reported a low activity level (< 30 min/day). Subjects were also instructed to record their fasting body weight weekly. Variations larger than 1 Kg were reported to the nutritionist, who suggested compensatory dietary modifications. The intake of key components in AD was not affected by these modifications.
Except for almonds and fresh, frozen and smoked fish, the active food items required for the 4-wk AD were provided. Some of the foods included in the CD were also provided. In order to assess dietary compliance, participants completed a daily menu checklist covering each 4-wk diet period. They also completed a questionnaire investigating their experience with each diet. Coaching support was provided by the nutritionist, who contacted each participant at least once per diet period. The study was approved by the Regional Ethical Review Board, Lund, Sweden (Dnr 593/2008).
Diets
Participants ate their habitual diet before and between the experimental periods. A dietary habit questionnaire completed before randomization to the starting dietary treatment was used to help the participants to resume their eating habits during the washout period. To ensure good compliance the participants were provided with a detailed 2-wk rotating menu plan for each dietary period (AD and CD) with all foods ingredients expressed in weight and/or volume measures. Electronic self-taring scales were made available whenever needed. The menu plan also included recipes for preparing meals.
The nutritional profiles of CD and AD are shown in Table 2. Both diets were designed in close agreement with the Nordic Nutrition Recommendations and supplied 2,500–2,600 Kcal/d for men and 2,000–2,100 Kcal/d for women, combining foods from plant and animal origin. Both diets incorporated commercial foods available in food stores, but the AD also included prototypes. For a detailed list of products included in AD see additional file 1.
AD combined several functional concepts with potential to modulate different physiological variables related to the inflammatory tonus and cardiometabolic risk, including:
a) food items naturally rich in antioxidants which, in addition to the anti-inflammatory action of their antioxidants, contain phenolics that may improve BP and blood lipids.
b) omega-3 fatty acids, especially those long-chained present in oily fish, which have antiinflammatory and triglyceride-lowering properties. In addition, the overall fat quality was also a concept included in the diet design. Thus, the unsaturated-to-saturated fat ratio was larger in AD than in CD (3.6 vs 1.2, respectively; Table 2).
c) ingredients with potential to beneficially affect the gut microbiota: a probiotic strain (Lactobacillus plantarum Heal19, DSM 15313), and prebiotics, i.e. beta-glucans and resistant starch in intact barley kernels, whole kernel rye flour and isolated barley fiber, which were used for baking an experimental beta-glucan-rich bread. Additional sources of viscous fermentable dietary fibre were a prototype oat-based fiber drink, a rye/oat breakfast cereal and an oat-based muesli.
d) low glycemic impact foods/meals were included for their association with reduced risk for the MetS and Type-2 diabetes, and its perceived ability to ameliorate the inflammatory tonus in healthy individuals. Low GI foods were represented by products with high beta-glucan content which, besides their prebiotic role, may improve glycemic regulation in a 10-h post-ingestion perspective, through mechanisms related to colonic fermentation and lowered inflammatory tonus. AD also included another high-fibre bread, baked from a wheat flour/guar gum blend: This item promotes a low glycemic response, with reduced peak and prolonged duration of net increment above fasting glucose levels (A. Nilsson, K. Radeborg et al., unpublished results). Additionally, AD included ingredients that lower the post-meal glycemic excursion, such as whey protein and vinegar.
e) ingredients with acknowledged abilities to normalize blood levels of both total and LDL cholesterol were also provided: different soybean products, a margarine enriched in stanol esters and dry almonds.
The mean daily quantities of the different functional ingredients in AD and the main functional properties considered for their selection are summarized in Table 3. None of the active ingredients were included in the CD, except for minor amounts of ω-3 fatty acids.
Participants were provided with a 14-day rotating menu plan. For representative 1-day menu plans for the CD and AD see additional file 2. A limited amount of alcohol-containing drinks were allowed during both dietary periods (30 and 37 g ethanol/wk for women and men, respectively). These limits, however, did not compel low-drinkers to increase their habitual alcohol consumption. Due to the low overall energy contribution from such contingent alcoholic drinks, they were not included in the estimation of the energy content of diets. The volunteers' coffee and tea drinking habits were not modified during the trial.
Analyses
Routine blood tests were analyzed at the Clinical Chemistry Laboratory/Skåne University Hospital Lund, on fasting plasma (total and HDL cholesterol, triacylglycerols, apo A-1, apo B, high-sensitivity C-reactive protein), serum (insulin) or on total blood samples (HbA1c). LDL cholesterol concentrations were calculated. HOMA-IR was calculated from fasting blood plasma-glucose and serum insulin values.
Venous plasma glucose concentrations were measured immediately after bleeding (HemoCue®B-glucose, HemoCueAB, Ängelholm, Sweden). Serum concentrations of TNF-α and IL-6 were measured with a chemiluminescent immunometric assay, Immulite®/Immulite® 1000 TNF-α (Siemens, Deerfield, IL, USA) and a sensitive enzyme-linked immunosorbent assay, Quantikine® HS (RD Systems Inc, Abingdon, UK), respectively. Samples with high IL-6 concentration (> 10 ng/L) in the ELISA were also measured with the Immulite®/Immulite®1000 IL-6 assay (Siemens, Deerfield, IL, USA). Serum FFA were assessed with an enzymatic colorimetric assay kit (Wako Chemicals GmbH, Germany). Plasminogen activator inhibitor (PAI-1) activity in plasma was measured with TriniLize PAI-1 Activity kit (Trinity Biotech, Jamestown NJ, USA).
The diet compositions were analyzed using the 2009 food database from the Swedish National Food Administration and a computerized calculation program (Dietist XP 3.1; Kost och Näringsdata AB, Bromma, Sweden).
Calculations and Statistical Analysis
The results are expressed as means ± SEM. Data were evaluated by mixed model ANOVA with sequence and interaction between diet and start/end of treatment periods as fixed effects, and participants within sequence and visit as random effects. Least square means were estimated for the start (Active-week 0, Control-week 0) and end values (Active-wk 4, Control-wk 4) for each diet. Least square means and confidence intervals were calculated for the differences between Active/wk 0, Control/wk 0, Active/wk 4 and Control/wk 4, as well as for the net difference between the diets, i.e. (Active/wk 4 - Active/wk 0) - (Control/wk 4 - Control/wk 0). Another set of analyses with body weight included as a continuous covariate was performed. Analyses were carried out using SAS PROC Mixed (v. 8.2, SAS Institute Inc., Cary, NC, USA). Ten-year coronary heart disease risk was calculated both with the Framingham Study equation considering age, gender, total cholesterol, HDL-cholesterol, smoking and systolic BP values, and the Reynolds Risk Score, which also incorporates CRP values.
Power Calculation
Primary outcome measure was change in plasma LDL cholesterol. Assuming a 0.5 mmol/L (approximately 10%) post-diet difference and a 0.97 SD, with α = 0.05 and 1-β = 0.8, a minimum of 30 participants were required. Among additional measures, changes in hs-CRP were also taken into account. Assuming a 0.44 mg/L (10%) post-diet difference and a SD of 1.0, the required sample size was 42 subjects. Data obtained from hypercholesterolemics with normal CRP levels confirmed that this n value provides sufficient power for assessing CRP changes.
SHARE
