Abstract
Background: The amount and type of fat in the diet influence the development of obesity and related inflammatory activity. Knowledge of the possible influence of dietary habits on circulating adiponectin, a molecule with putative antiinflammatory properties, may be helpful in preventing atherosclerosis and type 2 diabetes.
Methods: The association between dietary fat, as inferred from plasma fatty acid composition (gas–liquid chromatography), and circulating adiponectin (RIA) was evaluated in 116 healthy individuals.
Results: The proportion of saturated fatty acids in plasma was significantly associated with circulating adiponectin concentration (r = −0.24; P = 0.01). Specifically, percentage of palmitic acid (C16:0) was significantly associated with lower adiponectin concentration (r = −0.28; P = 0.002), particularly among women (r = −0.37; P = 0.02) and nonsmokers (r = −0.30; P = 0.007). Percentage of myristic acid (C14:0) was also significantly associated with lower adiponectin among nonsmokers (r = −0.26; P = 0.02) and women (r = −0.39; P = 0.01). The other fatty acids were not significantly associated with adiponectin except for eicosanoic acid (C20:1 ω-9), which was significantly and positively associated with adiponectin in all individuals (r = 0.23; P = 0.01). This latter association was most significant in smokers (r = 0.43; P = 0.007). In a multivariate regression analysis to predict circulating adiponectin, after controlling for age, body mass index, waist-to-hip ratio, and the individual remaining fatty acids, the percentages of palmitic (P = 0.005) and eicosanoic acid (P = 0.03) contributed independently (6% and 3%, respectively) to adiponectin variance. Among nonsmokers, the percentages of palmitic acid (P = 0.01) and ω-3 fatty acids contributed 8% and 7%, respectively, to adiponectin variance. Among smokers, the percentage of eicosanoic acid (P = 0.03) contributed to 10% of adiponectin variance, independently of body mass index, age, waist-to hip ratio, and the remaining individual fatty acids.
Conclusions: Saturated and ω-3 fatty acids of dietary origin (as inferred from plasma fatty acid concentration) are associated with circulating adiponectin concentrations in healthy humans. The proportion of eicosanoic acid also appears to be positively associated with circulating adiponectin. The knowledge of how these interactions occur may be helpful in the planning of dietary measures aimed at the modulation of inflammatory activity.
Adipose tissue is now known to secrete a myriad of proteins that are increasingly recognized to influence the metabolism of carbohydrates, proteins, and lipids (1). Adiponectin (also called Acrp30 or adipoQ in mice) is a 244-amino acid protein synthesized and secreted exclusively by the adipose tissue (2)(3). Adiponectin circulates at relatively high (mg/L) concentrations, and its half-life is in the range of several hours (4). This rate of turnover would be consistent with a role for this protein as a regulator of metabolic processes, such as modulation of fat partitioning, among other actions. Administration of adiponectin and/or its globular head portion stimulates free fatty acid oxidation in muscle (5). High fat feeding leads to decreased Acrp30 secretion in rodents (6), whereas caloric restriction leads to higher circulating concentrations of adiponectin in mice and humans (7). Circulating adiponectin concentrations are also inversely correlated to triglyceride (8)(9) and intramyocellular lipid stores (10). It has been reported that plasma concentrations of adiponectin are lower in obesity and type 2 diabetes (8)(11) and that adipose tissue Acrp30/adiponectin mRNA expression is decreased in obese ob/ob mice and obese humans (12).
The mechanisms underlying the observed close association between plasma adiponectin concentration and lipid metabolism are being elucidated (13)(14). In vitro and in vivo studies in rodents have shown that adiponectin prevents lipid accumulation in skeletal muscles in parallel to lowering blood glucose and improving insulin action (14). Recent observations suggest that adiponectin could play a role in counteracting the development of diet-induced insulin resistance (6)(7). It is noteworthy that these actions appear to be independent of the presence of obesity: adiponectin-null mice showed diet-induced insulin resistance despite increases in body weight similar to those in control mice (15)(16).
The interaction among diet-induced insulin resistance, adiponectin concentrations, and lipid metabolism could be exerted at the level of the inflammatory cascade. The amount and quality of fat in the diet seem to be of importance for development of insulin resistance and related inflammatory activity (17). A low proportion of long-chain unsaturated fatty acids and a high proportion of saturated fatty acids in the diet have been associated with impaired insulin action (18). Adiponectin possesses antiinflammatory (19) and antiatherogenic properties (20)(21). On the other hand, highly unsaturated fatty acids, ω-3 fatty acids in particular, are also receiving increasing attention as potential antiinflammatory agents (22) because these dietary fatty acids appear to modulate the release of different cytokines (23)(24).
How dietary fat may impact on peripheral adiponectin concentration, or vice versa, is unknown. One potential pathway is activation of the nuclear receptor peroxisome proliferator-activated receptor-γ (PPAR-γ). 1 In fact, fatty acids activate PPAR-γ (25), and pharmacologic activation with PPAR-γ agonists leads to increased plasma adiponectin concentrations (26)(27).
Knowledge of the possible influence of dietary habits on circulating adiponectin concentrations could be helpful in delineating dietary measures aimed at preventing type 2 diabetes. However, studies of the associations between dietary habits in relation to the development of chronic diseases are hampered by several methodologic problems, including imprecision of dietary surveys (18). One way to monitor the type of fat in the diet is to record the fatty acid composition in plasma (18). We aimed to evaluate the association between dietary habits, as inferred from plasma fatty acid composition, and circulating adiponectin in healthy individuals.
Materials and Methods
participants
We evaluated 116 individuals as part of an ongoing epidemiologic study dealing with nonclassical cardiovascular risk factors in the general population. All participants were of Caucasian origin and reported that their body weight had been stable for at least 3 months before the study. None was taking any medication or had any evidence of metabolic disease other than obesity. Inclusion criteria for this group were (a) body mass index (BMI) <40 kg/m2, (b) absence of any systemic disease, and (c) absence of any infections in the previous month. The study was approved by the Hospital Ethics Committee, and informed consent was obtained from all participants.
Anthropometric and clinical measurements.
BMI and waist-to-hip ratio (WHR) were measured in all participants. BMI was calculated as weight (in kilograms) divided by height (in meters) squared, and participants’ waists were measured with a soft tape midway between the lowest rib and the iliac crest. The hip circumference was measured at the widest part of the gluteal region.
Smokers were defined as any person consuming at least one cigarette a day in the previous 6 months. Resting blood pressure was measured by mercury sphygmomanometer after participants had been in a sitting position for a minimum of 15 min. The blood pressure was read three times on the right arm by the same investigator. The mean of the three measurements was used for this study. All women were premenopausal, had regular menstrual cycles, and were studied in their follicular phase. Liver disease and thyroid dysfunction were specifically excluded by biochemical workup.
serum samples
Analysis of serum fatty acids.
Following the method of Lepage and Roy (28), we precisely weighed 100 μL of plasma obtained after a 12-h fast into glass tubes and diluted the plasma with methanol–benzene (4:1 by volume). We slowly added acetyl chloride (200 μL) over a period of 1 min. After transesterification, we dried the pooled solvent extracts under a gentle stream of nitrogen at room temperature. Residues were dissolved in 500 μL of hexane, and an aliquot was injected into the chromatograph. Fatty acids were chromatographed as methyl esters on a 30-m fused-silica column (0.25 mm i.d.). Analysis was performed on a Hewlett-Packard 5890 gas chromatograph equipped with a flame ionization detector. The column temperature was held at 80 °C for 3 min and then increased in a stepwise fashion to a plateau of 220 °C. The injection port and detector temperatures were 250 and 270 °C, respectively. Helium was used as carrier gas. An internal standard consisting of 50 μg of pentadecanoic acid (C15:0) was precisely weighed and added to the serum.
Adiponectin measurements.
Plasma adiponectin concentrations were measured by RIA (LINCO Research Inc.). Samples were diluted 500-fold (10 μL of plasma plus 4990 μL of assay buffer) before the assay. The detection limit of the method is 2 μg/L. The intra- and interassay CVs were <5%.
Measurements of other analytes.
The serum glucose concentration was measured in duplicate by the glucose oxidase method.
The serum insulin concentration was measured in duplicate by a monoclonal IRMA (Medgenix Diagnostics). The lower limit of detection for insulin was 4.0 mIU/L. The intraassay CV was 5.2% at a concentration of 10 mIU/L, and the interassay CV was 6.9% at 14 mIU/L. The fasting insulin resistance index [homeostasis model assessment (HOMA)] was calculated with the formula: HOMA = fasting glucose (mmol/L) × fasting insulin (mIU/L)/22.5. In our experience, HOMA correlates well with the SI calculated by the minimal model approach: r = 0.79; P <0.0001 (29).
Total serum cholesterol was measured by the reaction of cholesterol esterase/cholesterol oxidase/peroxidase. Total serum triglycerides were measured through the reaction of glycerol-3-phosphate oxidase and peroxidase.
statistical analysis
Descriptive results for continuous variables are expressed as the mean (SD). Before statistical analysis, gaussian distribution and homogeneity of the variances were tested. Variables that did not fulfill these tests (individual fatty acids and ratios, HOMA, adiponectin) were log-transformed. The relationships between variables were analyzed by simple correlation analysis (Pearson r). We set statistical significance at P ≤0.05 and have based our discussion on these relationships. Participants were divided into quintiles of circulating fatty acids. We compared adiponectin concentrations in individuals in the highest fatty acid quintiles vs the remaining participants with the Student t-test. We also constructed a stepwise multivariate linear regression analysis to predict adiponectin concentrations, taking into consideration those variables with statistical association of at least P <0.1 on univariate analysis. Because in a previous work we found that smoking affected the relationship between inflammatory markers and metabolic variables (30), we also examined whether smoking (defined as at least 1 cigarette/day in the previous 6 months) affected adiponectin relationships.
Results
The main characteristics and serum fatty acid composition of the study participants are shown in Tables 1⇓ and 2⇓ . The absolute fatty acid concentration did not differ significantly between men and women. Circulating adiponectin was significantly higher in women than men (P = 0.006; Table 1⇓ ). The proportion of saturated acids tended to be lower and the proportion of oleic acid (C18:1 ω-9) tended to be higher among women (Table 1⇓ ). The proportion of eicosanoic acid (C20:1 ω-9) was significantly higher in women than men (Table 1⇓ ). Plasma adiponectin correlated negatively with BMI (r = −0.23; P = 0.01), WHR (r = −0.28; P = 0.002), and fasting plasma triglycerides (r = −0.38; P <0.0001) and tended to be associated with HOMA value (r = −0.16; P = 0.08). Circulating adiponectin was not significantly associated with age, systolic or diastolic blood pressure, fasting glucose, or cholesterol in this series.
Anthropometric and biochemical variables.1
Plasma fatty acid profile according to smoking status.
The proportion of saturated fatty acids in plasma was significantly associated with circulating adiponectin concentrations (r = −0.24; P = 0.01; Table 3⇓ ). Specifically, the proportion of palmitic acid (C16:0) was significantly inversely associated with adiponectin concentration (r = −0.28; P = 0.002). This association was significant mainly among women (r = −0.37; P = 0.02) and in nonsmokers (r = −0.30; P = 0.007). The proportion of myristic acid (C14:0) was also significantly inversely associated with adiponectin among nonsmokers (r = −0.26; P = 0.02) and in women (r = −0.39; P = 0.01).
Correlations between serum fatty acids and adiponectin.1
The other individual fatty acids were not significantly associated with adiponectin except for eicosanoic acid (C20:1 ω-9), which was significantly and positively associated with adiponectin in all participants (r = 0.23; P = 0.01). This latter association was most significant in smokers (r = 0.43; P = 0.007). C20:1 ω-9 was the only fatty acid that was significantly increased in smokers compared with nonsmokers (Table 2⇑ ). Total monounsaturated ω-9 fatty acids were also significantly associated with adiponectin among smokers (Table 3⇑ ). The proportion of ω-3 fatty acids tended to be positively associated with adiponectin in nonsmokers (r = 0.21; P = 0.07).
We next divided our population into quintiles for the different fatty acids. As anticipated from the correlation analysis, individuals in the two highest quintiles for percentage of palmitic acid had significantly decreased serum adiponectin concentrations compared with the remaining individuals [13.9 (4.8) vs 16.4 (5.28) mg/L; P = 0.006; Fig. 1⇓ ]. Not apparent from the correlation analysis were the following findings: individuals in the highest quintile of docosahexanoic acid (DHA; C22:6) had significantly increased circulating adiponectin concentrations compared with the remaining participants [18.02 (4.8) vs 14.76 (4.9) mg/L; P = 0.018; Fig. 1⇓ ]; and individuals in the highest quintile of γ-linolenic acid (C18:3 ω-6) had significantly decreased serum adiponectin compared with the remaining participants [12.1 (3.6) vs 15.4 (5) mg/L; P = 0.04].
95% confidence intervals (CI) for serum adiponectin in relation to quintiles for percentage of palmitic acid (top panel) and DHA (C22:6 ω-3; bottom panel).
As anticipated from the correlation analysis, individuals in the two highest quintiles of percentage of palmitic acid had significantly decreased serum adiponectin concentrations [13.9 (4.8) vs 16.4 (5.28) mg/L; P = 0.006]. Individuals in the highest quintile of DHA (C22:6) had significantly increased circulating adiponectin concentrations [18.02 (4.8) vs 14.76 (4.9) mg/L; P = 0.018].
The ratio of saturated/(ω-3) fatty acids correlated positively with both HOMA value (r = 0.19; P = 0.03) and circulating adiponectin concentrations (r = −0.19; P = 0.03).
We performed several multivariate regression analyses to predict circulating adiponectin concentrations. In these models, we considered as independent variables those individual fatty acids that showed a relationship with at least P <0.1 on univariate analysis. After controlling for age, BMI, WHR (which persisted in the model), and the individual remaining fatty acids, only the proportions of palmitic acid (C16:0; P = 0.005) and eicosanoic acid (C20:1 ω-9; P = 0.03) contributed independently to adiponectin variance (6% and 3%, respectively).
Among nonsmokers, and after again controlling for age, BMI, WHR (which persisted in the model), and the remaining individual fatty acids, only the proportions of palmitic acid (C16:0; P = 0.01) and ω-3 fatty acids contributed to adiponectin variance (8% and 7%, respectively).
Among smokers, only the proportion of eicosanoic acid (C20:1 ω-9; P = 0.03) contributed to adiponectin variance (10%), independently of BMI, age, WHR, and the remaining individual fatty acids. Of the remaining factors, only WHR persisted in the model.
Discussion
The proportions of fatty acids in plasma mirror the dietary fat composition. The relationships between the amount of polyunsaturated fatty acids in the diet and the corresponding proportions of the same fatty acids in plasma lipids are often strong (18). This is usually true for essential fatty acids, such as linoleic and α-linolenic acid. However, most other types of fatty acids can be synthesized by humans from precursors, particularly saturated fatty acids.
Adiponectin was negatively associated with the main saturated acid, palmitic acid (C16:0). In a recent in vitro study, saturated fatty acids induced activation of nuclear factor-κB, an important mediator in the production of several cytokines (31), whereas DHA counteracted these effects (31)(32). To our knowledge, the possible actions of saturated fatty acids on adiponectin production have not been evaluated. Our findings suggest that increased intake of saturated fatty acids and increased endogenous transformation of fatty acids that leads to increased concentrations of saturated fatty acids in plasma are associated with decreased circulating adiponectin concentrations. Cause and consequence cannot be derived from the present study. It could be hypothesized that decreased adiponectin concentrations amplify the proinflammatory action of saturated fatty acids.
Adiponectin was positively associated with the proportion of C20:1 ω-9. This association was especially significant in smokers (P = 0.007). Although BMI tended to be lower with increased proportions of C20:1 ω-9, the association remained significant after controlling for BMI (r = 0.25; P = 0.008). Interestingly, significantly increased products of delta-9 desaturation and significant increases in C20 elongation products have been observed in rats with reduced food intake (33). On the other hand, smoking has been found to be associated with a decreased proportion of essential fatty acids (34), and essential fatty acid deficiency leads to a characteristic increase in ω-9 fatty acids (35). Because caloric restriction can lead to higher circulating concentrations of adiponectin in mice and humans (7), it is tempting to speculate that reduced food intake leads to increased adiponectin and C20:1 ω-9 concentrations concomitantly.
As secondary findings, we found that those individuals with an increased proportion of C18:3 ω-6 had concomitantly decreased circulating adiponectin concentrations. The content of this fatty acid is typically very small in the diet. Rather, it reflects increased endogenous desaturation of linoleic acid by delta-6 desaturase, increasing the proportion of C18:3 ω-6, probably as a consequence of a low proportion of linoleic acid in the diet (18). Our findings suggest that low intake of this essential fatty acid could lead to increased C18:3 ω-6 and concomitant decreased adiponectin concentrations.
We also show that the proportions of ω-3 fatty acids in general, and of DHA in particular, were highest in those individuals with increased circulating adiponectin. In nonsmokers, this association persisted after controlling for age, BMI, WHR, and the remaining individual fatty acids. Interestingly, two putative antiinflammatory molecules, ω-3 fatty acids and adiponectin, seem to be linked. The same processes and dietary habits that increase the concentrations of ω-3 fatty acids lead to concomitant increased adiponectin concentrations. In recent years, novel functional sets of lipid-derived mediators with antiinflammatory actions generated from ω-3 fatty acids have been increasingly recognized (22). The association of the proportions of ω-3 and adiponectin described here seems particularly important in light of recent findings in American women: Increased intake of the ω-3 polyunsaturated fatty acids eicosapentanoic (C20:5 ω-3) and DHA (C22:6 ω-3) was associated with reduced risk of thrombotic diseases (26). In parallel with these observations, the authors of in vitro and in vivo studies in humans reported that supplementation with eicosapentanoic acid and DHA appeared to reduce cytokine production (23)(24)(31)(36)(37)(38). The above relationships were most likely to have a dietary explanation because food is the major source of these fatty acids (18). Adiponectin also has antiatherogenic properties and inhibits proliferation of vascular smooth muscle cell (16). Inflammation in the vessel wall plays an essential part in the initiation and progression of atherosclerosis (39)(40). Decreased adiponectin concentrations have been observed in patients with coronary artery disease (8). The findings of the present study suggest that all of these associations are interrelated events.
An association between concentrations of fasting plasma triglycerides and adiponectin has also been observed (9). Different studies have shown that adiponectin promotes lipid oxidation in humans, with a subsequent decrease in intracellular lipid content in human muscle (10). These results are consistent with animal data: Adiponectin was shown to enhance lipid oxidation and decrease the concentration of muscle triglycerides (5)(13).
One limitation of this study is its transversal design; thus, the findings shown here are hypothesis-generating. It will be necessary to demonstrate that adiponectin concentrations can be modulated by ingestion of certain fatty acids or by changing of serum fatty acid composition. In addition, fatty acids were measured in plasma, but it is well known that fatty acids are also a major component of the cell membranes. Fatty acids measured in cell membranes may better reflect longer-term dietary fatty acid intake because fatty acid turnover may be slower in cell membranes than in plasma.
In summary, saturated and ω-3 fatty acids of dietary origin (as inferred from plasma fatty acid concentration) are associated with circulating adiponectin concentrations in healthy humans. The proportion of C20:1 ω-9 also appears to be positively associated with circulating adiponectin. The knowledge of how these interactions occur will be helpful in the planning of dietary measures aimed at the modulation of inflammatory activity.
Acknowledgments
This work was partially supported by Grant 00/0024-01 from the Fondo de Investigaciones Sanitarias, National Health Institute of Spain, and by Grants RGDM G03/212 and RGTO G03/028 from the Instituto de Salud Carlos III, Madrid, Spain.
Footnotes
1 Values are the mean (SD).
2 NS, not significant.
1 Values are the mean (SD).
2 NS, not significant.
1 Bold font indicates fatty acids significantly correlated with adiponectin concentrations.
↵1 Nonstandard abbreviations: PPAR-γ, peroxisome proliferator-activated receptor-γ; BMI, body mass index; WHR, waist-to-hip ratio; HOMA, homeostasis model assessment; and DHA, docosahexanoic acid.
- © 2005 The American Association for Clinical Chemistry
