Association between plasma fatty acids and inflammatory markers in patients with and without insulin resistance and in secondary prevention of cardiovascular disease, a cross-sectional study

Atherosclerotic cardiovascular disease (CVD) such as acute myocardial infarction and stroke remains the leading cause of death worldwide [1]. Therefore, the better understanding of its etiology and of strategies to prevent and to treat CVD is highly needed. It should be noted that inflammatory biomarkers have a positive association with insulin resistance, an important determinant of CVD [2–5]. Also, epidemiological and clinical studies have shown an association between inflammatory biomarkers, such as C-reactive protein (CRP) [6], tumor necrosis factor (TNF)-alpha [7] and interleukin (IL)-6 [8], and risk of cardiovascular events among primary and secondary prevention patients.

Dietary components, especially fatty acids, affect the expression and release of inflammatory biomarkers. Polyunsaturated fatty acids (PUFAs), such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), have a cardioprotective effect by reducing proinflammatory gene expression (TNF-alpha and IL-1ß [9, 10]), and by increasing gene expression of adiponectin [11, 12], which is protective against vascular dysfunction and insulin resistance [13, 14]. Indeed, clinical studies have shown that diets may have effect on inflammatory biomarkers levels and further effect on insulin sensibility [15, 16]. However, it has been poorly studied among secondary prevention patients, this population has an important characteristic which is they are at high cardiovascular risk and probably well treated (under many drug therapy) [17].

Given the association between inflammation and insulin resistance, and the role of fatty acids on modulating the inflammatory biomarkers levels on primary prevention patients, we aimed to understand if this association might be similar among CVD secondary prevention patients with insulin resistance. Such an association would potentially provide insights into the role of inflammatory biomarkers in different metabolic status of cardiovascular patients. Thus, this cross-sectional study has two objectives: (i) to compare inflammatory biomarker levels among CVD secondary prevention patients with and without insulin resistance; (ii) to evaluate if plasma fatty acid levels are associated with inflammatory biomarker levels among CVD secondary prevention patients with and without insulin resistance.

The volunteers are part of the Brazilian Cardioprotective Nutritional Program Trial (BALANCE Program Trial, which is funded by Hospital do Coração (HCor) as part of the “Hospitais de Excelência a Serviço do SUS (PROADI-SUS)” Program, in partnership with the Brazilian Ministry of Health. All eligibility criteria are reported in the study protocol [18]. Briefly, the BALANCE PROGRAM TRIAL included 2468 patients aged 45 years or older, who had experienced one or more indicators of established CVD in the preceding 10 years.

The initial population considered for this cross-sectional sub-study consisted of 400 volunteers, selected from the BALANCE PROGRAM TRIAL site at the Department of Clinical Nutrition in the Instituto Dante Pazzanese de Cardiologia, São Paulo, Brazil. Of these, 41 subjects were excluded due to lack of blood analysis, though the analysis was performed in 359 participants. STROBE-Nut checklist is provided as Additional file 1.

Regarding inflammatory biomarkers, we found that the plasma concentration of hs-CRP and IL-6 was significantly lower among the NIRG (p < 0.002) than the IRG. NIRG individuals had a lower plasma concentration of myristic fatty acid (p = 0.007), palmitoleic (p = 0.001), and a higher plasma concentration of arachidonic fatty acid (p = 0.02) and docosapentaenoic fatty acid (p = 0.02) than the IRG.

There was no significant association between plasma concentrations of hs-CRP and IL-12, and plasma fatty acids in either NIRG (Table 3) or IRG (Table 4). Among the NIRG (Table 3), eicosatrienoic fatty acid had a negative association with plasma adiponectin concentration (p = 0.024). This negative association was not found in the IRG, on the other hand, other four fatty acids showed association with adiponectin among IRG. Plasma adiponectin concentration was negatively associated with the eicosapentaenoic fatty acid (EPA) (p = 0.011) (Table 4), while the oleic and palmitoleic monounsaturated fatty acids and the arachidonic fatty acid were positively associated with the adiponectin plasma concentration (p = 0.015, p = 0.005, p = 0.047, respectively).

In the IRG, EPA and docosapentaenoic fatty acids were negatively associated with IL-6 plasma concentration (p = 0.015 and p = 0.007, respectively). For every one unit increase of plasma EPA level (1%), there was a reduction of 0.55 on the log(IL-6) mean values; in another perspective, the increase of 1% in EPA reduces on average 40% the IL-6 mean, independently of other variables (Table 3). In the NIRG IL-6 was positively associated with stearic fatty acid concentration (p = 0.032) and negatively associated with α-linolenic fatty acid (p = 0.008). In another perspective, in the NIRG the increase of 1% in stearic fatty acid increase on average 10% the IL-6 and the increase of 1% in alpha-linolenic fatty acid reduces on average 45% the IL-1 β.

There were no associations between fatty acids and IL-1β and IL-8 in the NIRG. However, in the IRG, the IL-1β was negatively associated with oleic fatty acid (p = 0.039), the increase of 1% in oleic fatty acid reduces on average 5% the IL-1β. Furthermore, in the IRG, the increase of 1% in eicosatrienoic fatty acid reduces on average 15% the IL-8.

A negative association between IL-10 plasma concentration and alpha-linolenic fatty acid (p = 0.016) was seen in the NIRG. No association between IL-10 and fatty acid in the IRG was found. The TNF-a plasma concentration was positively associated with linoleic fatty acid (p = 0.021) and negatively with EPA (p = 0.045) in the IRG, which means that the increase of 1% in EPA reduces in average 25% the TNF-alpha. A negative association between TNF-a and alpha-linolenic fatty acid was observed among the NIRG (p = 0.027).

In the present cross-sectional study with subjects in secondary prevention for CVD, it can be seen that those who do not have insulin resistance have significantly lower anthropometric indices, as well as a lower plasma concentration of triacylglycerol, insulin and blood pressure. There is compelling evidence that CVD risk factors are present to a significantly greater degree in the subset of overweight/obese individuals that are also insulin resistant [22]. In the present study, we could also observe that patients on secondary prevention for CVD with insulin resistance have higher inflammatory biomarker concentrations than those patients on secondary prevention without insulin resistance.

This higher pro-inflammatory biomarkers concentration may be due to the excess of adipose tissue, especially visceral tissue, and high intakes of certain saturated fatty acids which can enhance the expression of genes encoding proinflammatory proteins such as TNF-alpha and IL-1β, which, in turn, impair the insulin signaling pathway [23].

Saturated fatty acid high plasma concentration, mostly palmitic fatty avid, has a proinflammatory and atherogenic role since it has an association with the toll-like receptor (TLR)-4 and the consequent activation of transcription factors such as NF-κB and AP-1, which induces the synthesis of pro-inflammatory cytokines such as IL-6 and TNF-α [24]. In addition, saturated fatty acid induces insulin resistance due to its antagonistic action on the PPAR-γ coactivator (PGC1-alpha), which promotes the expression of mitochondrial genes involved in oxidative phosphorylation and insulin-mediated glucose uptake [25, 26]. In the present study, in fact, individuals with insulin resistance had a higher plasma concentration of myristic-saturated fatty acid, a fatty acid typically found in dairy products. A systematic review with meta-analysis [27] showed that lauric fatty acid (C12: 0) has the greatest capacity to promote increased plasma concentration of LDL-c, followed by myristic acid (C14: 0) and palmitic acid (C16: 0).

The omega-3 PUFA family, including EPA and DHA, have an anti-inflammatory role and in the present study, a lower plasma concentration of PUFA was observed among IRG. Among the possible mechanisms for the anti-inflammatory effects of PUFA is the activation of peroxisome proliferator-activated receptors (PPAR), which affect the decrease of the NF-kB translocation into the nucleus, thus reducing the expression of pro-inflammatory genes [9, 10, 28]. In the present study, the plasma concentrations of omega-6 PUFAs were positively associated with the plasma concentration of TNF-alpha in the IRG. These fatty acids are considered proinflammatory because they are the main precursors of prostaglandins and leukotrienes with greater biological activity than those from the omega-3 series [29, 30]. However, subjects with high omega-6 intake do not appear to have elevated concentrations of inflammatory biomarkers; indeed, it seems there are not enough data to support the detrimental effect of low omega-6 intake on cardiovascular health [31]. Thus, further studies should be performed among individuals with and without insulin resistance for further clarification of this possible association.

In addition to the increase in saturated fatty acid concentration, IRG also showed an increase in the concentration of hs-CRP and IL-6, which may indicate an increase in cardiovascular risk, given the role that these cytokines play in atherogenesis. IL-6 acts in the chemotaxis and mitogenesis of smooth muscle cells [32], being the main cytokine to stimulate the hepatic synthesis of positive acute phase proteins, such as CRP, which is now known to be an independent, major risk marker of cardiovascular complications [33]. In addition, IL-6 has recently been proposed to play a central role in the link between obesity, inflammation and coronary heart disease [8, 34]. IL-6 production by adipose tissue could directly affect liver metabolism by inducing very low-density lipoprotein (VLDL) secretion and hypertriglyceridemia since the visceral adipose tissue is closely connected to the liver by the venous portal system [35]. Studies have suggested that IL-6 could be involved in insulin resistance and its complications [36, 37]. Considering the results from the comparative analysis between plasma fatty acid concentration and inflammatory biomarkers among the IRG and NIRG groups, we evaluated whether concentrations of plasma fatty acids were associated with an increase or reduction of inflammatory biomarkers in these two groups. We emphasize that significant associations between these variables were most frequently observed among the IRG group, which may demonstrate the importance of the role and impact of nutritional management to reduce cardiovascular risk among these individuals. We noted that, although the IRG group had higher plasma concentration of myristic fatty acid, hs-CRP, and IL-6, there was no association between these variables in either the NIRG or IRG. On the other hand, PUFAs were inversely associated with the plasma concentration of IL-6 in both groups. Furthermore, in the IRG group, plasma stearic saturated fatty acid concentration was positively associated with the plasma concentration of IL-6.

There was no difference in adiponectin concentrations between the IRG and the NIRG. Possibly, this result is associated with the fact that 67.0% of individuals without insulin resistance are already overweight. In addition to the increase in pro-inflammatory protein expression, the visceral adipose tissue increase is related to a reduction in adiponectin expression and synthesis [38]. On the other hand, weight loss and increased PUFA intake are associated with increased adiponectin concentration [39, 40]. The increase in this adipocytokine is associated with an improvement in insulin sensitivity in liver, muscle and adipose tissue [41], as it inhibits the NF-kB signaling pathway [42].

In the present study, a higher plasma concentration of PUFAs was observed in the NIRG group. In the same group, these fatty acids were negatively associated with both inflammatory biomarkers such as IL-6 and TNF-alpha and anti-inflammatory markers such as adiponectin and IL-10. Previously published studies highlight the absence of an association between polyunsaturated fatty acid and adiponectin in patients in secondary prevention for CVD [43, 44]. Why some studies have not found this association is not clear, but it seems to be linked to the fact that these patients are well medicated with drugs that may affect inflammatory biomarker activities [45, 46].

To our knowledge, this is the first study to analyze the difference in characteristics between individuals with and without insulin resistance in secondary prevention for cardiovascular disease. It should be borne in mind that these individuals are at high cardiovascular risk and probably well treated. However, in the present study, we observed that even when medicated, subjects in secondary prevention with insulin resistance presented uncontrolled cardiovascular risk factor. The BALANCE Program nutritional prescription follows the Brazilian guidelines for the treatment of CVD The BALANCE Program diet is designed to meet the nutritional recommendations proposed by the Brazilian Cardiovascular Guidelines, which, in turn, are guided by the nutritional composition of diets such as the Mediterranean diet and the DASH diet. The proposal of is to ensure adequate nutrient intake with a locally appropriate diet, ie, one consisting of foods that are consumed in Brazil, not necessarily the foods consumed as part of the Mediterranean diet. We highly recommend that a subgroup analysis should be realized in the end of BALANCE Trial, regarding the effect of this intervention among those with and without insulin resistance.

However, some limitations deserve consideration. Due to the cross-sectional design, the possibility of residual confounding (such as unknown confounders) cannot be excluded. Even though we have adjusted for confounder by including it in our model, it may still have residual confounding. Only randomization can eliminate unknown confounders. However, once it is well known the effect of fatty acids, especially polyunsaturated fatty acids, on cytokines levels among primary prevention patients for cardiovascular disease [47], we suggest it is not a source of false positive result, it would not bias the estimates toward the null. In addition, we have analyzed 12 different fatty acids, if possible other fatty acids should also be studied. Another limitation would be the lack of dietary intake information in the present study. In the BALANCE Trial, baseline food intake data were obtained by two 24-h dietary recall. However, the patients analyzed in the present cross-sectional study have answered to only one 24-h dietary recall. Since for a better food intake analysis, we must use tools such as a food frequency questionnaire and/or several 24-h reminders, we preferred to exclude this data for this manuscript analysis. Therefore, information such as the ingestion of antioxidants and other bioactive compounds that may have anti-inflammatory activity was not evaluated and therefore could not be included in the analysis. Another important limitation to highlight is that in this study, the significance level was 0.05 without adjustment for multiple comparisons. Because of this, all analyses should be interpreted as exploratory.

Subjects in secondary prevention for cardiovascular disease with insulin resistance have a higher concentration of hs-CRP and IL-6 than individuals without insulin resistance, and these inflammatory biomarkers are positively associated with saturated fatty acids and negatively associated with unsaturated fatty acids.