Intervention studies on the Mediterranean Diet (MedDiet) have often led to weight loss, which may have contributed to the purported anti-inflammatory effects of the MedDiet. To investigate the impact of the MedDiet consumed under controlled feeding conditions before (−WL) and after weight loss (+WL) on markers of inflammation in men with metabolic syndrome (MetS).
Design and Methods:
Subjects (N = 26, male, 24–65 years) with MetS first consumed a North American control diet for 5 weeks followed by a MedDiet for 5 weeks both in isocaloric feeding conditions. After a 20-week weight loss period in free-living conditions (10 ± 3% reduction in body weight, P < 0.01), participants consumed the MedDiet again under isocaloric-controlled feeding condition for 5 weeks.
MedDiet − WL significantly reduced plasma C-reactive protein (CRP) concentrations (−26.1%, P = 0.02) and an arbitrary inflammatory score (−9.9%, P = 0.01) that included CRP, interleukin-6 (IL-6), IL-18, and tumor necrosis factor-α (TNF-α) compared with the control diet. The MedDiet + WL significantly reduced plasma IL-6 (−20.7%) and IL-18 (−15.6%, both P ≤ 0.02) concentrations compared with the control diet but had no further significant impact on plasma CRP concentration. Participants with a reduction in waist circumference ≥8.5 cm after MedDiet + WL showed significantly greater reductions in inflammation markers than those with a change in waist circumference <8.5 cm.
Thus, consuming MedDiet even in the absence of weight loss significantly reduces inflammation. However, the degree of waist circumference reduction with weight loss magnifies the impact of the MedDiet on other markers of inflammation associated with MetS in men.
Inflammation is now undisputedly recognized as being a key etiological factor in the pathogenesis of atherosclerosis and cardiovascular disease (CVD) (1). Although most studies show that plasma C-reactive protein (CRP) concentrations are positively associated with the risk of CVD (2,3), whether CRP has a direct causal impact on CVD outcomes remains a controversial issue (4). Other markers of inflammation such as interleukin-6 (IL-6), IL-18, and tumor necrosis factor-α (TNF-α) have also been associated with the risk of CVD (5-7). Metabolic syndrome (MetS) is associated with a proinflammatory state characterized by elevated plasma concentrations of several markers of inflammation including CRP, IL-6, IL-18, and TNF-α (8).
Reductions in plasma CRP and other proinflammatory cytokines concentrations have been observed following different weight loss regimens, particularly severe dietary restrictions leading to >10% body weight reductions (9,10). Although weight loss may represent the primary target to treating patients with MetS, improving the quality of the diet has also become increasingly recognized (8). A growing body of evidence suggests that a proportion of the cardio-protective effect of the Mediterranean type diet (MedDiet) may be explained by its anti-inflammatory properties (11,12). However, it is unclear how concurrent reductions in body weight frequently seen with the MedDiet (11) modulate its effects on CVD risk factors, including inflammation (13). Previous studies have attempted to dissect out the impact of the MedDiet and weight loss on proinflammatory markers through statistical analysis, with inherent limitation in such approaches. Based on such secondary analyses, it cannot be completely ruled out that weight loss indeed contributed to changes in inflammation with MedDiet, because these studies were not primarily designed to investigate this issue specifically.
The aim of this study was therefore to examine the impact of the MedDiet consumed under controlled feeding conditions in the absence of weight loss on markers of inflammation in men with MetS and to verify how weight loss modulates these effects.
Methods and Procedures
The study has been described in detail previously (14). Briefly, non-smoking men (18-65 years) with the MetS (National Cholesterol Education Program-Adult Treatment Panel III (NCEP-ATP III)) (8), but without previous history of CVD or type 2 diabetes and taking no lipid-lowering or hypertension medication, were recruited for the study. To be eligible, men had to have a stable weight for at least 6 months, to use no vitamin supplements and natural health products and to have no aversion for specific components of the MedDiet. All study procedures have been approved by the Research Ethics Committee of Laval University and written informed consent was obtained from all participants enrolled in the study.
Diets and study design
Figure 1 shows the study design (14). The subjects' diet was first standardized to a North American control diet that they consumed for 5 weeks under isocaloric, weight-maintaining conditions. This allowed us to minimize interindividual variations attributed to each participant's diet. Participants then consumed the MedDiet for 5 weeks again under isocaloric, weight-maintaining conditions. This first phase of the study in controlled feeding conditions (all foods provided to participants) allowed us to investigate the impact of the MedDiet per se on inflammation before weight loss, i.e., in the absence of clinically meaningful weight loss (referred to as MedDiet − WL). Participants subsequently underwent a 20-week weight loss period in free-living conditions. A registered dietitian was giving personalized advice on how to create a 500 kcal deficit in daily energy intake, in order to generate a minimum of 5% reduction in body weight. This was achieved mainly by recommending a reduction in portion size as well as by promoting low-energy density food choices. Finally, subjects consumed the MedDiet for a second time for 5 weeks under isocaloric feeding, weight-stabilizing conditions. This allowed us to investigate the impact of the MedDiet after weight loss (referred to as MedDiet + WL) on markers of inflammation compared to the control diet and the MedDiet without weight loss.
Diets before weight loss
The control diet was devised to reflect current macronutrient intake averages in North American men (15). The MedDiet was formulated to be concordant with characteristics of the traditional Mediterranean eating pattern (16). Standardized 7-day cyclic menus and daily servings of various food categories for the control diet and the MedDiet were developed for the study and have been previously published (14). The nutritional composition of the control diet and the MedDiet provided to participants is presented in Table 1. The main source of fat in the MedDiet was virgin olive oil.
Table Table 1. Mean nutritional composition of the control and the Mediterranean diet with and without weight loss
The control diet was first assigned using predetermined levels of energy intake estimated using 3-day food records, and revised according to body weight fluctuations measured daily during the isocaloric phases of the study. All meals, foods, and beverage including red wine were provided to participants at the clinical investigation unit, thereby optimizing control for energy and macronutrient intake. Men were instructed to consume only the meals provided and a checklist was provided to all participants to identify the prescribed foods that had not been consumed when eating outside of the clinical investigation unit (mostly dinners and breakfasts) or had been consumed in addition to the formulated diets. The use of vitamin supplements, anti-inflammatory medications (nonsteroidal anti-inflammatory drugs), and natural health products were strictly forbidden during the entire experimental period. Consumption of tea and coffee was allowed with a limit of 2 per day. Subjects were instructed to maintain their usual physical activity except for the 3 days that preceded blood sampling periods, during which they were asked to refrain from intense physical exercise. Information on physical activity was collected using a daily check-list and energy expenditure above resting was calculated as described previously (17).
Weight loss period and stabilization on the MedDiet
After the first isocaloric feeding phase, participants received instructions on how to reduce their usual energy intake by 500 kcal/day in free-living condition, with the objective to generate a minimum of 5-10% reduction in body weight over a period of 20 weeks. Unlike the first phase of the study, meals and foods were not provided to participants. Three 3-day food journals collected at weeks 2, 8, and 16 of the weight loss period were used to compute a dietary score reflecting overall “adherence” to the MedDiet as published in Goulet et al. (18). The MedDiet score obtained during the weight loss period indicated that participants were not adhering to the MedDiet principles during this period of the study (data not shown). To achieve our second objective, only participants who had lost at least 5% of their body weight were eligible for the last phase of the study, during which, they consumed the MedDiet for 5 weeks again under carefully isocaloric, weight-stabilizing conditions, with all foods provided to them as in the first phase.
Body weight and waist circumferences were measured according to standardized procedures (19). Serum concentrations of high sensitivity-CRP, IL-6, IL-18, and plasma TNF-α concentrations were measured using commercial enzyme-linked immunosorbent assay kits; CRP: BioCheck, Foster City, CA, coefficient of variation: 3.3%; IL-6 and TNF-α: R&D Systems, Minneapolis, MN, coefficient of variation: 7.8 and 8.4% respectively; IL-18: MBL International, Woburn, MA, coefficient of variation: 7.5%.
Data are reported as mean ± s.d. and percentage of changes from the control diet unless stated otherwise. Data were analyzed using the PROC MIXED procedure for repeated measures in SAS with diet as the main repeated effect (v9.2; SAS Institute, Cary, NC). The Tukey adjustment in the MIXED model was used to account for multiple comparisons of the three diets. The extent to which initial values (on the control diet) modified the change in inflammation markers in response to the MedDiet − WL and MedDiet + WL was also investigated using appropriate interaction terms. When interaction terms were found to be significant, subgroups were created using the median value for the selected parameter to better illustrate this interaction. The extent to which the magnitude of reduction in waist circumference after weight loss modified the impact of the MedDiet on markers of inflammation was tested using unpaired t-test between groups defined using median value of reduction in waist circumference achieved on MedDiet after weight loss (≥8.5 cm or <8.5 cm). The diet effect within groups of waist circumference reduction was assessed using paired t-tests. Correlation analyses were used to examine the contribution of various factors to the metabolic changes induced by the MedDiet before and after weight loss. An arbitrarily defined inflammation score was computed based on the sum of points for each tertile of CRP (<2.2, 2.2-4.9, ≥4.9 mg/l), IL-6 (<0.79, 0.79-1.35, ≥1.35 pg/ml), IL-18 (<232.6, 232.6-324.7, ≥324.7 pg/ml), and TNF-α (<0.72, 0.72-0.97, ≥0.97 pg/ml). This score therefore varied from 4 to 16 points. Variables with a skewed distribution were log-10 transformed before statistical analysis. Differences at P ≤ 0.05 (two-sided) were considered significant.
Of the 29 men recruited in the study, two subjects dropped out and one subject was excluded because of a lack of compliance to the dietary regimen. A total of 26 men with MetS completed the first phase investigating the impact of the MedDiet − WL on markers of inflammation. Their characteristics at screening are shown in Table 2. The overall compliance to the predetermined diets during the isocaloric feeding phases of the study calculated from the food checklist in completers was 98.0 ± 3.74%.
Table Table 2. Physical characteristics and plasma lipid profile of the 26 male subjects at screening
Plasma markers of inflammation measured after each dietary treatment are presented in Table 3. Despite every effort to maintain subjects' weight constant, there was a slight but significant body weight reduction after MedDiet − WL compared to the control diet (1.2 ± 0.9 kg, P < 0.001). However, body weight was constant over the last 3 weeks of the control and the MedDiet − WL. Statistical analysis for the MedDiet − WL were therefore adjusted for body weight variation in the mixed model (N = 26) even though adjustment for this small change in body weight had no impact on the differences in inflammatory markers between MedDiet − WL and the control diet. The MedDiet − WL significantly reduced plasma CRP concentrations (−26.1%, P = 0.02) and the arbitrary inflammation score (−9.9%, P = 0.01) compared with the control diet (N = 26). There was no significant change in plasma IL-6, IL-18, and TNF-α concentrations after MedDiet − WL compared with the control diet. There was also no correlation between change in plasma CRP (MedDiet-WL vs. the control diet) and concurrent variations in plasma IL-6, IL-18, and TNF-α (0.02 < r < 0.20, all P > 0.3). Variations in body weight or waist circumference after the MedDiet − WL vs. control were not correlated with changes in inflammatory markers, including CRP (data not shown).
Table Table 3. Plasma concentrations of pro-inflammatory markers at the end of each experimental diet
Seven participants were not able to achieve the predetermined 5% minimal reduction in body weight and were therefore excluded from the last dietary phase as well as for statistical analyses. Thus, comparisons including MedDiet + WL are based on N = 19 for all diets. Body weight was reduced by an average of 10.2 ± 2.9% and waist circumference by 8.6 ± 3.3 cm in these 19 subjects after the weight loss period and stabilization on the MedDiet for 5 weeks (N = 19, P < 0.001 for both) as shown previously (14). Body weight decreased by ∼1 kg during the first week of the controlled feeding period with the MedDiet after the weight loss period, and by 0.5 kg in average during the last 2 weeks of this feeding period. Therefore, subjects were not actively losing weight in the weeks that preceded the blood test. MedDiet + WL had no further significant impact on plasma CRP concentrations and the arbitrary inflammatory score compared with MedDiet − WL. On the other hand, MedDiet + WL significantly reduced plasma IL-18 concentrations compared with MedDiet − WL (−13.0%) and the control diet (−15.6%, both P < 0.001). MedDiet + WL was also associated with a significant reduction in plasma IL-6 concentrations (−20.7%, P = 0.020) compared with the control diet. Unlike after the MedDiet − WL, changes in plasma CRP concentrations on the MedDiet + WL from values on the control diet were positively correlated with changes in plasma IL-6, IL-18, and TNF-α concentrations (0.50 < r < 0.57, all P < 0.05).
We further examined the extent to which reductions in waist circumference on MedDiet + WL modulated the changes in markers of inflammation. Subjects were separated based on the median value of waist circumference reduction achieved during the weight loss phase (Figure 2). Despite a mean reduction in body weight of 8.0 ± 1.9% and in waist circumference of 6.1 ± 1.9 cm, subjects with a reduction in waist circumference <8.5 cm (N = 9), showed a significant reduction only in plasma IL-18 concentrations after the MedDiet + WL compared to the control diet. Subjects with a reduction in waist circumference ≥8.5 cm (N = 10) had an average weight loss of 12.2 ± 2.2% and waist circumference reduction of 10.8 ± 2.7 cm, and they showed significant reductions in plasma CRP (−54.3%, P < 0.01), IL-6 (−29.5%, P = 0.03), IL-18 (−21.1%, P < 0.01), and TNF-α (−15.8%, P = 0.07) concentrations after the MedDiet + WL compared with the control diet.
Plasma CRP concentrations on the control diet predicted the CRP response to the MedDiet + WL and MedDiet − WL (Figure 3). Only participants with plasma CRP concentrations above the median after the control diet showed significant reductions in plasma CRP concentrations on the MedDiet − WL and MedDiet + WL compared to the control diet (P for interaction = 0.01).
Individuals with MetS frequently manifest a proinflammatory state characterized by elevated plasma cytokines concentrations (IL-6, IL-18, and TNF-α) and acute phase reactants such as CRP. Inflammation plays an important role in atherosclerosis process, from the initial recruitment of monocyte into the intima to the rupture of the plaque. Several inflammatory biomarkers such as CRP, cytokines, and interleukins have been associated with MetS features, including insulin resistance and obesity, as well as with incidence of CVD (9,20). We believe this is the first nutritional intervention in controlled feeding conditions specifically designed to dissect out the impact of the MedDiet and weight loss on CVD risk factors including inflammatory markers.
First, the small difference in weight (1.3%) between the control and the MedDiet − WL is unlikely to be of clinical relevance, being too small to induce a metabolic response. In our study, the MedDiet in the absence of clinically meaningful weight loss significantly reduced plasma CRP concentrations and an arbitrary inflammation score in men with MetS. Changes in plasma CRP concentrations in our study were comparable in magnitude to those obtained from two intervention studies in which adherence to the MedDiet was assessed by a diet score. In these studies, participants with higher MedDiet score had 20-24% lower plasma CRP and 16-17% lower plasma IL-6 concentrations compared to those with a lower MedDiet score (21,22). Mena et al. (23) have also shown that a 3-month MedDiet supplemented with virgin olive oil was associated with significant reduction in plasma CRP (30-40%) and IL-6 (10-20%) concentrations. Finally, adherence to MedDiet for 3 years reduced plasma IL-18 concentrations compared to baseline values (24).
MedDiet in the absence of clinically meaningful weight loss had little impact on adipose tissue-derived cytokines such as IL-6, IL-18, and TNF-α, suggesting that diet-induced reduction in CRP with MedDiet was independent of variations in these adipose tissue-derived proinflammatory cytokines (20). Consistent with this hypothesis is the lack of correlation between MedDiet-induced change in plasma CRP and concurrent variations in IL-6, IL-18 or TNF-α in the absence of weight loss. IL-6 has been shown to increase the de novo synthesis of CRP primarily by upregulating CRP mRNA transcription while IL-1β is thought to regulate CRP synthesis through post-translational processes (25,26). Consumption of a MedDiet for 4 weeks has been associated with reduced IL-1β gene expression in mononuclear cells in fasting and postprandial states (27). It is therefore possible that the reduction in plasma CRP with MedDiet - WL may be due to a downregulated translation of the protein through attenuation of IL-1β gene expression, independent of changes in IL-6 expression. Consumption of a MedDiet for 4 weeks has also been associated with reduced monocyte chemoattractant protein 1 (MCP-1) postprandial gene expression in mononuclear cells, suggestive of attenuated monocyte recruitment in adipose tissue (28). These mechanisms may partly explain the reduced plasma CRP concentrations with MedDiet in the absence of weight loss even when other proinflammatory cytokines are not significantly altered. Although in our study plasma IL-18 concentrations on the control diet were comparable to those obtained in other populations with MetS, plasma IL-6 and TNF-α concentrations were much lower (about 40-50%) and this may also account for the lack of effect of the MedDiet without weight loss on these proinflammatory markers (11,29,30). Finally, we cannot exclude the possibility that the small reduction observed in plasma IL-6, IL-18, and TNF-α would have been significant with a greater number of subjects.
MedDiet and weight loss
The study also allowed us to investigate the combined effect of weight loss and the MedDiet on markers of inflammation. The MedDiet combined with a large reduction in waist circumference had greater anti-inflammatory effect than the MedDiet − WL as reflected by more pronounced reductions in CRP, IL-6, IL-18, and TNF-α concentrations compared with the control diet. Nevertheless, whereas MedDiet + WL improved IL-6 and IL-18 concentrations compared to the control diet, it had no further effect on the arbitrary inflammatory score and on plasma CRP concentrations in men with MetS compared to the MedDiet − WL. Moreover, the magnitude of the reduction in plasma CRP concentrations after the MedDiet + WL compared with the control diet was positively correlated with concurrent reductions in plasma concentrations of CRP-modulating cytokines such as IL-6, IL-18, and TNF-α.
Although combining weight loss to the MedDiet had no further impact on plasma CRP concentrations, our data suggest that the extent of the reduction in waist circumference magnifies the impact of the MedDiet on inflammation. Indeed, only the subgroup of men with waist circumference reduction >8.5 cm (>10% body weight loss) showed further significant reductions in plasma CRP, IL-6, and TNF-α concentrations with the MedDiet. Changes in most markers of inflammation were also positively associated with changes in body weight (data not shown). This is consistent with data from several studies, which have shown that weight loss, particularly when greater than 10% of initial body weight, is generally associated with beneficial changes in most biomarkers of inflammation (9,10). Taken together, these results indicate that large reductions in body weight and waist circumference are required to further impact inflammation while on the MedDiet.
There was a significant interaction between baseline CRP values and CRP response to the MedDiet with and without weight loss (Figure 3). Specifically, MetS patients characterized by a more pronounced proinflammatory state at baseline (CRP >3.8 mg/l) showed significant reduction in plasma CRP on the MedDiet even in the absence of weight loss compared with the control diet. This is further evidence indicating that the MedDiet may reduce inflammation even in the absence of weight loss, particularly among men with a more pronounced proinflammatory state (31).
In summary, this study shows for the first time that the MedDiet attenuates inflammation in men with MetS independent of weight loss. Data also confirm that important reductions in body weight and waist circumference magnify the anti-inflammatory effect of the MedDiet.
We are thankful to Provigo-Loblaws which provided the foods used in the present study through their support of the Chair in Nutrition and Cardiovascular Health. We also thank the staff of the metabolic kitchen from the Institute of Nutraceuticals and Functional Foods of Laval University for their dedicated work during the study. We are grateful to the nurses and the laboratory staff of the Institute of Nutraceuticals and Functional Food for their technical assistance and the expert care provided to the participants. We also express our gratitude to the participants, without whom the study would not have been possible. ClinicalTrial.gov registration number: NCT00988650. B.L. is a Chair in Nutrition and Cardiovascular Health. S.D. is a Canadian Institutes of Health Research (CIHR) New Investigator and a Fonds de la recherche en santé du Québec (FRSQ) Junior 1 Scholar. C.R. is recipient of doctoral scholarship from CIHR and FRSQ. This study was supported by an operating grant from the Canadian Institutes for Health Research (MOP-68866). Provigo/Loblaws donated the foods used in this study.