Abstract
Background: Animal models have shown that adipose-derived palmitoleic acid may serve as a lipokine that contributes to resistance to diet-induced obesity. Studies in humans have evaluated only plasma palmitoleic acid concentrations, which reflect stearoyl-coenzyme A desaturase 1 (SCD1) activity in the liver and are associated with increased risk of obesity. These apparent opposite effects of palmitoleic acid deserve further research in humans. Because carbohydrate intake can increase hepatic SCD1 activity, it could be used as a stratifying variable to disentangle the effects of adipose tissue SCD1 compared with the effects of liver SCD1 activity on obesity.
Objective: We examined whether the effects of adipose tissue palmitoleic acid and SCD1 activity were associated with decreased obesity prevalence and whether this association was modified by carbohydrate intake.
Design: Prevalence ratios (PRs) of obesity [body mass index (in kg/m2) >30] were examined in a cross-sectional study in 1926 adults in Costa Rica. Two desaturation indexes (16:1/16:0 and 18:1/18:0) were used as surrogate measures of adipose tissue SCD1 activity.
Results: We observed a positive association between adipose tissue palmitoleic acid concentrations and obesity (PR for lowest compared with highest quintiles of palmitoleic acid: 2.27; 95% CI: 1.52, 3.38; P for trend < 0.0001). A significant association was also observed between obesity and adipose desaturation indexes. The association between adipose tissue palmitoleic acid concentrations and obesity was attenuated in persons with low carbohydrate intake.
Conclusions: There is no direct evidence that adipose tissue palmitoleic acid behaves as a lipokine to reduce obesity occurrence in humans. However, the attenuation of the association by low carbohydrate intake warrants further research on adipose-derived palmitoleic acid and obesity risk.
INTRODUCTION
Obesity is a worldwide health concern (1) and is caused by an imbalance between energy intake and energy expenditure that results in fat storage (2). It is known that stearoyl-coenzyme A desaturase (SCD) enzymes affect fat storage (3, 4). SCD enzymes are involved in de novo synthesis of monounsaturated fatty acids from saturated fatty acids (5). The major products of SCD are palmitoleic acid (16:1) and oleic acid (18:1), which are key substrates for the formation of triglycerides (5), from its preferred substrates, palmitic acid (16:0) and stearic acid (18:0). Two SCD isoforms have been observed in humans, SCD1 (primarily observed in adipose tissue and the liver) and SCD5 (observed in the brain and pancreas) (6–8).
Because of the difficulties of directly measuring SCD1 activity in humans, SCD1 activity is proxied by the use of its product:substrate ratios (ie, desaturation indexes) including the ratio of palmitoleic acid (16:1) to palmitic acid (16:0) concentrations and the ratio of oleic acid (18:1) to stearic acid (18:0) concentrations (9, 10). Results from animal and human studies suggest that SCD1 activity and its products may play an important role in obesity (9, 11–19). SCD1-defficent mice are protected against diet-induced obesity (11–13), and high SCD1 activity is related to obesity in animal models (14–17). Similarly, plasma palmitoleic acid concentrations and SCD1 activity are positively associated with hypertriglyceridemia, abdominal adiposity, and obesity in humans (9, 18, 19). In contrast, Cao et al (20) recently showed in their genetic mouse model (ie, mice deficient in adipose tissue lipid chaperones aP2 and mal1) that palmitoleic acid produced and released by adipose tissue may serve as a lipokine that contributes to resistance to diet-induced obesity and improvement in insulin sensitivity. Although findings in this specific mouse model may not be translated into humans, these apparent opposite effects of palmitoleic acid warrant further research in humans.
In previous human studies that have examined risk of obesity in relation to levels of SCD1 activity and palmitoleic acid concentrations, desaturation indexes and palmitoleic acid concentrations were derived by using blood samples, which cannot address the possible different roles that liver and adipose tissue play in the development of obesity. The purpose of this study was to use adipose tissue from a population-based study in Costa Rica to test the hypothesis that adipose tissue palmitoleic acid and desaturation indexes of SCD1 activity (16:1/16:0 and 18:1/18:0, respectively) are associated with a decreased prevalence of obesity. Compared with adipose tissue, the liver has a unique ability to dramatically increase fatty acid synthesis in response to changes in dietary carbohydrate intake (2), and carbohydrates can increase hepatic SCD1 activity (12, 21). Thus, we also examined whether carbohydrate intake modified the association between palmitoleic acid concentrations and obesity.
SUBJECTS AND METHODS
Study population
The participants in this study were control subjects from a case-control study conducted in the Central Valley in Costa Rica from 1994 to 2004 (22, 23). Control subjects were randomly selected by matching first cases of nonfatal acute myocardial infarction on the basis of age (±5 y), sex, and area of residence (county) from information available from the National Census and Statistics Bureau of Costa Rica. The final participation rate for control subjects was 88%. The general characteristics of the study population according to obesity status are shown in Table 1. All subjects provided written informed consent to participate in the study. The study was approved by the Human Subjects Committee of the Harvard School of Public Health and the University of Costa Rica.
TABLE 1.
Characteristics of the study population (n = 1926) according to obesity status1
| Obesity |
|||
| Variables | Present (n = 340) | Absent (n = 1586) | P |
| Age (y) | 58 ± 112 | 59 ± 11 | 0.130 |
| Female (%) | 40.9 | 24.0 | <0.001 |
| Urban residence (%) | 41.8 | 39.7 | 0.486 |
| Income ($/mo) | 627 ± 429 | 574 ± 410 | 0.034 |
| BMI (kg/m2) | 33.1 ± 2.9 | 24.9 ± 2.9 | <0.001 |
| Waist-to-hip ratio | 0.96 ± 0.08 | 0.94 ± 0.07 | 0.003 |
| Current smoker (%) | 12.9 | 23.1 | <0.001 |
| EE on daily activity (METs/d) | 32.1 ± 12.0 | 36.0 ± 15.8 | <0.001 |
| Dietary variables | |||
| Total energy intake (kcal/d) | 2357 ± 865 | 2451 ± 724 | 0.065 |
| Total fat (% of energy) | 32.1 ± 5.7 | 31.7 ± 5.8 | 0.249 |
| Saturated fat (% of energy) | 10.4 ± 2.6 | 10.4 ± 2.7 | 0.794 |
| Monounsaturated fat (% of energy) | 11.8 ± 3.6 | 11.7 ± 3.8 | 0.591 |
| Polyunsaturated fat (% of energy) | 6.4 ± 2.0 | 6.2 ± 2.1 | 0.122 |
| trans Fat (% of energy) | 1.26 ± 0.60 | 1.27 ± 0.59 | 0.864 |
| Carbohydrate (% of energy) | 54.7 ± 7.3 | 55.6 ± 7.3 | 0.064 |
| Protein (% of energy) | 13.4 ± 2.2 | 12.9 ± 2.1 | <0.001 |
| Cholesterol (mg/d) | 294 ± 193 | 290 ± 165 | 0.729 |
| Fatty acids in adipose tissue: (g/100 g total fatty acids) | |||
| Stearic acid (18:0) | 2.3 ± 0.8 | 2.9 ± 1.0 | <0.001 |
| Oleic acid (18:1n−9) | 42.5 ± 2.9 | 42.0 ± 3.1 | 0.008 |
| Palmitic acid (16:0) | 20.6 ± 2.6 | 21.5 ± 2.8 | <0.001 |
| Palmitoleic acid (16:1n−7) | 7.5 ± 2.1 | 6.4 ± 2.2 | <0.001 |
| α-Linolenic acid (18:3n−3) | 0.60 ± 0.18 | 0.66 ± 0.21 | <0.001 |
| Eicosapentaenoic acid (20:5n−3)3 | 0.04 ± 0.02 | 0.04 ± 0.02 | 0.105 |
| Docosahexaenoic acid (22:6n−3) | 0.15 ± 0.05 | 0.14 ± 0.05 | 0.002 |
| Linoleic acid (18:2n−6) | 15.3 ± 3.3 | 15.6 ± 3.9 | 0.139 |
| Arachidonic acid (20:4n−6) | 0.54 ± 0.13 | 0.46 ± 0.14 | <0.001 |
| SCD1 activity (18:1/18:0) | 20.8 ± 7.1 | 16.5 ± 6.3 | <0.001 |
| SCD1 activity (16:1/16:0) | 0.37 ± 0.13 | 0.31 ± 0.12 | <0.001 |
EE, energy expenditure; METs, metabolic equivalent tasks; SCD1, stearoyl-coenzyme A desaturase 1. Obesity was identified as BMI (in kg/m2) >30. The t test was used for continuous variables, and the chi-square test was used for categorical variables.
Mean ± SD (all such values).
Missing values for 348 subjects.
Data collection
Trained personnel visited all study participants at their homes for collecting information on their sociodemographic characteristics, medical histories, and smoking-related data by the use of a questionnaire with closed-ended questions. Anthropometric measurements were collected by fieldworkers with subjects wearing light clothing and without shoes. All measurements were performed in duplicate, and the average of measurements was used for all analyses. A steel anthropometer and a bathroom scale (Detecto, Webb City, MO) or a Seca Alpha Model 770 digital scale (Seca, Hanover, MD) accurate to 50 g were used to measure height and weight respectively. The 2 scales were calibrated biweekly. Body mass index (BMI) was calculated as weight in kilograms divided by the square of height in meters. Self-reported dietary intake was collected through a food-frequency questionnaire that was developed and specifically validated to assess fatty acid intake in the Costa Rican population (24). Biological specimens were collected in the morning at the homes of subjects after an overnight fast as described in detail elsewhere (23). A subcutaneous adipose tissue biopsy was collected from the upper buttock with a 16-gauge needle and disposable syringe following procedures previously described (25). Samples were stored in nitrogen tanks at −80°C.
Fatty acid analysis
Fatty acids from adipose tissue were quantified by gas-liquid chromatography as described previously (26). Peak retention times and area percentages of total fatty acids were identified by injecting known standards (NuChek Prep; Nuchek, Elysian, MN) and analyzed with Agilent Technologies ChemStationA.08.03 software (Agilent Technologies, Santa Clara, CA). Twelve duplicate samples, which were indistinguishable from the others, were analyzed throughout the study. The CVs for palmitoleic acid, palmitic acid, oleic acid, and stearic acid were 5.7%, 2.9%, 2.4% and 7.7%, respectively.
Data analyses
The original sample size was composed of 2274 subjects. After eliminating control subjects with missing values on BMI, adipose tissue fatty acids, and potential confounders, a total of 1926 participants remained in the analyses. t Tests for continuous variables and chi-square tests for categorical variables were used to assess differences in means or distributions of lifestyle and other variables for subjects with obesity (BMI >30) and without obesity. Log-Poisson models with generalized estimating equations (27) were built to calculate the prevalence ratios (PRs) and 95% CIs of obesity according to quintiles of palmitoleic acid concentrations and SCD1 activity in adipose tissue. Two desaturation indexes were used to measure SCD1 activity in adipose tissue: the ratio of oleic acid to stearic acid concentrations in adipose tissue and the ratio of palmitoleic to palmitic acid concentrations in adipose tissue. Potential confounding factors, which were previously shown to be risk factors for obesity and were also correlated with adipose palmitoleic acid concentrations and desaturation indexes were entered in generalized linear models that included age, sex, annual household income, cigarette smoking [5 categories: never, past smoker, current smoker (<10 cigarettes/d), current smoker (≥10 and <20 cigarettes/d), current smoker (≥20 cigarettes/d), fatty acids in adipose tissue (eg, arachidonic acid, α-linolenic acid, and linoleic acid), physical activity, and total energy intake. Stratified analyses were carried out in the association between adipose tissue palmitoleic acid and the prevalence of obesity according to tertiles of glucose, starch, and carbohydrate intakes expressed separately as a percentage of total energy. Generalized score tests were computed to examine the interactive effects of carbohydrate intake and palmitoleic acid concentrations in adipose tissue on obesity. All analyses were carried out with SAS software (version 9.2; SAS Institute, Cary, NC).
RESULTS
A total of 340 subjects (17.7% of the study population) met the criterion for obesity. Obese subjects were more likely to be women but less likely to be smokers or to be physically active compared with nonobese subjects. In addition, obese subjects had significantly higher concentrations of palmitoleic acid and desaturation indexes in adipose tissue compared with those of individuals without obesity (all P values < 0.001) (Table 1).
As shown in Table 2, in nonobese individuals (n = 1586), percentages of women and concentrations of arachidonic acid in adipose tissue increased across quintiles of palmitoleic acid concentrations. Adipose tissue concentrations of α-linolenic acid and linoleic acid decreased across quintiles of palmitoleic acid concentrations. Similar findings were observed when we evaluated the role of potential confounders across quintiles of desaturation indexes (16:1/16:0 and 18:1/18:0) in adipose tissue (data not shown).
TABLE 2.
Characteristics of subjects without obesity according to quintiles of palmitoleic acid concentrations (n = 1586)1
| Quintiles of palmitoleic acid (16:1n−7) in adipose tissue |
||||||
| Variables | 1 (n = 317) | 2 (n = 317) | 3 (n = 318) | 4 (n = 317) | 5 (n = 317) | P |
| Palmitoleic acid | 3.56 ± 0.712 | 5.16 ± 0.35 | 6.25 ± 0.29 | 7.43 ± 0.41 | 9.63 ± 1.34 | <0.001 |
| Age (y) | 58 ± 11 | 57 ± 11 | 58 ± 12 | 59 ± 12 | 60 ± 11 | 0.002 |
| Female (%) | 11.4 | 17.7 | 22.3 | 26.8 | 41.6 | <0.001 |
| Urban residence (%) | 42.9 | 42.3 | 40.3 | 36.6 | 36.6 | 0.305 |
| Income ($/mo) | 666 ± 466 | 604 ± 405 | 587 ± 400 | 523 ± 381 | 491 ± 370 | <0.001 |
| Current smoker (%) | 17.4 | 24.3 | 23.0 | 29.3 | 21.8 | 0.010 |
| EE on daily activity (METs/d) | 35.6 ± 14.7 | 36.5 ± 16.5 | 36.6 ± 16.4 | 36.7 ± 17.3 | 34.6 ± 13.8 | 0.378 |
| Dietary variables | ||||||
| Total energy intake (kcal/d) | 2463 ± 666 | 2481 ± 726 | 2472 ± 741 | 2441 ± 730 | 2396 ± 754 | 0.589 |
| Total fat (% of energy) | 32.5 ± 6.0 | 32.2 ± 5.7 | 32.1 ± 6.1 | 31.3 ± 5.4 | 30.3 ± 5.5 | <0.001 |
| Saturated fat (% of energy) | 10.1 ± 2.7 | 10.1 ± 2.5 | 10.4 ± 2.7 | 10.6 ± 2.8 | 10.6 ± 2.9 | 0.016 |
| Polyunsaturated fat (% of energy) | 6.8 ± 2.0 | 6.7 ± 2.2 | 6.2 ± 1.9 | 5.9 ± 2.0 | 5.5 ± 2.0 | <0.001 |
| trans Fat (% of energy) | 1.4 ± 0.6 | 1.3 ± 0.6 | 1.3 ± 0.6 | 1.2 ± 0.6 | 1.1 ± 0.5 | <0.001 |
| Carbohydrate (% of energy) | 55.1 ± 7.2 | 55.2 ± 7.1 | 55.1 ± 7.6 | 55.3 ± 7.4 | 57.0 ± 7.3 | 0.003 |
| Cholesterol (mg/d) | 297 ± 169 | 297 ± 174 | 299 ± 166 | 294 ± 158 | 265 ± 152 | 0.042 |
| Fatty acids in adipose tissue (g/100 g total fatty acids) | ||||||
| α-Linolenic acid (18:3n−3) | 0.75 ± 0.25 | 0.70 ± 0.22 | 0.67 ± 0.19 | 0.62 ± 0.19 | 0.57 ± 0.17 | <0.001 |
| Eicosapentaenoic acid (20:5n−3)3 | 0.04 ± 0.02 | 0.04 ± 0.02 | 0.04 ± 0.02 | 0.04 ± 0.02 | 0.04 ± 0.02 | <0.001 |
| Docosahexaenoic acid (22:6n−3) | 0.14 ± 0.06 | 0.14 ± 0.05 | 0.14 ± 0.06 | 0.14 ± 0.05 | 0.14 ± 0.04 | 0.200 |
| Linoleic acid (18:2n−6) | 17.8 ± 4.1 | 16.8 ± 3.8 | 15.5 ± 3.5 | 14.5 ± 3.3 | 13.4 ± 3.1 | <0.001 |
| Arachidonic acid (20:4n−6) | 0.39 ± 0.11 | 0.43 ± 0.13 | 0.45 ± 0.12 | 0.49 ± 0.13 | 0.53 ± 0.14 | <0.001 |
EE, energy expenditure; METs: metabolic equivalent tasks. ANOVA was used for continuous variables, and the chi-square test was used for categorical variables.
Mean ± SD (all such values).
Missing values for 277 subjects.
The associations between palmitoleic acid concentrations and SCD1 activity in adipose tissue and the prevalence of obesity are shown in Table 3. PRs of obesity increased across quintiles of palmitoleic acid concentrations (P for trend < 0.001). In the multivariate adjusted models, this association remained significant [PR of the lowest quintile compared with the highest quintile of palmitoleic acid concentrations: 2.27 (95% CI: 1.52, 3.38); P for trend < 0.0001]. A similar significant association between desaturation indexes and the prevalence of obesity was observed. Compared with subjects in the lowest quintile of the desaturation index 16:1/16:0 in adipose tissue, the PR for those in the highest quintile was 3.64 (95% CI: 2.54, 5.20; P for trend < 0.001), and this relation remained significant in the multivariable adjusted model (PR: 2.38, 95% CI: 1.60, 3.53; P for trend < 0.0001). Results were similar for the desaturation index 18:1/18:0 in adipose tissue (Table 3).
TABLE 3.
Obesity according to quintiles of palmitoleic acid concentrations and stearoyl-coenzyme A desaturase 1 (SCD1) activity in adipose tissue of Costa Rican adults (n = 1926)1
| Quintiles |
||||||
| 1 | 2 | 3 | 4 | 5 | P for trend | |
| Palmitoleic acid concentrations in adipose tissue2 | 3.68 ± 0.73 | 5.33 ± 0.35 | 6.44 ± 0.31 | 7.68 ± 0.41 | 9.85 ± 1.25 | |
| Unadjusted model | 1.0 | 1.46 (0.97, 2.19) | 1.65 (1.11, 2.45) | 2.40 (1.66, 3.47) | 3.20 (2.25, 4.55) | <0.0001 |
| Adjusted model3 | 1.0 | 1.45 (0.97, 2.16) | 1.63 (1.10 2.43) | 2.39 (1.64, 3.47) | 3.08 (2.15, 4.43) | <0.0001 |
| Adjusted model4 | 1.0 | 1.27 (0.86, 1.90) | 1.41 (0.94, 2.10) | 1.82 (1.22, 2.72) | 2.27 (1.52, 3.38) | <0.0001 |
| SCD1 activity (16:1/16:0) | ||||||
| Unadjusted model | 1.0 | 1.36 (0.89, 2.09) | 1.75 (1.17, 2.62) | 2.55 (1.75, 3.71) | 3.64 (2.54, 5.20) | <0.0001 |
| Adjusted model3 | 1.0 | 1.39 (0.91, 2.13) | 1.77 (1.18, 2.65) | 2.43 (1.66, 3.55) | 3.46 (2.39, 5.00) | <0.0001 |
| Adjusted model4 | 1.0 | 1.23 (0.80, 1.87) | 1.48 (0.98, 2.22) | 1.80 (1.22, 2.67) | 2.38 (1.60, 3.53) | <0.0001 |
| SCD1 activity (18:1/18:0) | ||||||
| Unadjusted model | 1.0 | 1.46 (0.91, 2.36) | 2.38 (1.54, 3.68) | 3.08 (2.02, 4.68) | 5.15 (3.47, 7.65) | <0.0001 |
| Adjusted model3 | 1.0 | 1.40 (0.87, 2.27) | 2.26 (1.46, 3.48) | 2.83 (1.86, 4.31) | 4.71 (3.13, 7.10) | <0.0001 |
| Adjusted model4 | 1.0 | 1.24 (0.77, 2.00) | 1.86 (1.19, 2.89) | 2.16 (1.40, 3.33) | 3.29 (2.12, 5.11) | <0.0001 |
All model values are prevalence ratios (95% CIs) and were calculated by using log-Poisson models with generalized estimating equations. Tests for trends were derived from log-Poisson models with a single term that represented the medians of quintiles 1–5 of palmitoleic acid.
Values are means ± SDs.
Adjusted for age, sex, area of residence, income, smoking status, and total physical activity.
Adjusted for age, sex, area of residence, income, smoking status, total physical activity, adipose tissue α-linolenic acid, linoleic acid, and arachidonic acid, total energy intake, and carbohydrate intake.
We carried out a stratified analysis of the association between adipose tissue palmitoleic acid concentrations and the prevalence of obesity according to tertiles of carbohydrate intake (Table 4). We used the stratification by carbohydrate intake as an indirect way to disentangle the effects of adipose-derived compared with liver-derived palmitoleic acid in adipose tissue with the assumption that, at low intakes of carbohydrates, the proportion of liver-derived palmitoleic acid is lower in adipose tissue than at high intakes of carbohydrates. In individuals with a high carbohydrate intake, increasing palmitoleic acid concentrations in adipose tissue was associated with increasing PRs for obesity. However, the PRs for obesity with increasing palmitoleic acid concentrations in adipose tissue at low carbohydrate intakes were attenuated (generalized score test; P for interactive effects = 0.15) (Table 4). We also stratified by different types of carbohydrate intake (glucose and starch). The PRs for obesity with increasing palmitoleic acid concentrations in adipose tissue were also attenuated at low glucose intakes (generalized score test; P for interactive effects = 0.02) (Table 5). Results for starch show no significant interaction, although there was still an attenuation of the association in people in the low–starch intake group (generalized score test; P for interactive effects = 0.67).
TABLE 4.
Stratified analysis of the association between adipose tissue palmitoleic acid concentrations and presence of obesity by carbohydrate intake (n = 1926)1
| Quintiles of palmitoleic acid concentration in adipose tissue |
||||||
| Tertiles of median carbohydrate intake | 1 | 2 | 3 | 4 | 5 | P |
| 1 (49% of energy intake) | 1.0 | 0.99 (0.58, 1.68) | 1.00 (0.57, 1.75) | 1.00 (0.57, 1.75) | 1.78 (1.07, 2.98) | 0.152 |
| 2 (56% of energy intake) | 1.0 | 1.27 (0.62, 2.60) | 1.48 (0.75, 2.95) | 2.07 (1.07, 3.99) | 2.48 (1.32, 4.63) | — |
| 3 (62% of energy intake) | 1.0 | 2.53 (0.87, 7.37) | 3.06 (1.09, 8.55) | 4.70 (1.74, 12.7) | 4.26 (1.57, 11.6) | — |
All values are prevalence ratios (95% CIs) from log-Poisson models with generalized estimating equations. The analysis was adjusted for age, sex, area of residence, income, smoking status, total physical activity, total energy intake, and adipose tissue α-linolenic acid, linoleic acid, and arachidonic acid.
Calculated from the generalized score test for the interaction term of carbohydrate intake and adipose palmitoleic acid in log-Poisson models with generalized estimating equations.
TABLE 5.
Stratified analysis of the association between adipose tissue palmitoleic acid concentrations and presence of obesity by glucose and starch intakes (n = 1926)1
| Quintiles of palmitoleic acid concentrations in adipose tissue |
||||||
| 1 | 2 | 3 | 4 | 5 | P2 | |
| Tertiles of median glucose intake | ||||||
| 1 (9 g/d) | 1.0 | 1.29 (0.64, 2.57) | 1.12 (0.55, 2.26) | 1.68 (0.88, 3.19) | 1.98 (1.04, 3.79) | 0.02 |
| 2 (17 g/d) | 1.0 | 1.72 (0.95, 3.12) | 1.46 (0.77, 2.78) | 1.64 (0.89, 3.02) | 1.77 (0.96, 3.25) | — |
| 3 (29 g/d) | 1.0 | 0.64 (0.25, 1.59) | 1.68 (0.84, 3.38) | 2.13 (1.05, 4.34) | 3.05 (1.59, 5.88) | — |
| Tertiles of median starch intake | ||||||
| 1 (107 g/d) | 1.0 | 1.11 (0.61, 2.01) | 1.22 (0.69, 2.15) | 1.25 (0.71, 2.22) | 1.76 (1.02, 3.04) | 0.67 |
| 2 (157 g/d) | 1.0 | 1.96 (0.89, 4.28) | 1.76 (0.80, 3.90) | 2.86 (1.36, 6.01) | 2.86 (1.37, 5.99) | — |
| 3 (202 g/d) | 1.0 | 1.07 (0.51, 2.25) | 1.36 (0.64, 2.91) | 1.87 (0.92, 3.78) | 2.45 (1.25, 4.81) | — |
All values are prevalence ratios (95% CIs) from log-Poisson models with generalized estimating equations. The analysis was adjusted for age, sex, area of residence, income, smoking status, total physical activity, total energy intake, and adipose tissue α-linolenic acid, linoleic acid, and arachidonic acid.
From generalized score tests for the interaction terms between adipose palmitoleic acid and glucose and starch intakes in log-Poisson models with generalized estimating equations.
DISCUSSION
In this population-based cross-sectional study, we observed a strong positive association between adipose tissue concentrations of palmitoleic acid and the occurrence of obesity. A similar significant association was observed between obesity and the 2 desaturation indexes (16:1/16:0 and 18:1/18:0) in adipose tissue.
Our findings were consistent with the results of other investigations in humans. In a study of 134 healthy men, abdominal adiposity was positively associated with plasma concentrations of palmitoleic acid (18). In a study of 849 Swedish men and women, the desaturation index (16:1/16:0) derived from blood samples was associated with the prevalence of obesity (9). Therefore, adipose tissue concentrations of palmitoleic acid and SCD1 activity likely play a positive role in the development of obesity.
However, in their genetic mouse model, Cao et al (20) showed that palmitoleic acid produced and released by adipose tissue mediates the communication between adipose tissue and other tissues and inhibits SCD1 activity in the liver, which contributes to the resistance to diet-induced obesity and enhanced insulin sensitivity. Because of the relevance of this finding in an animal model, our goal was to examine whether adipose tissue palmitoleic acid would behave similarly in humans, even if many findings in animal models have little or no relevance to metabolic mechanisms in humans. Palmitoleic acid is virtually absent in commonly consumed foods and is primarily biosynthesized from palmitic acid by the SCD1 in liver and adipose tissue (18). In the liver, palmitoleic acid is used in the formation of triglycerides, packaged in very-low-density lipoprotein, and secreted into the blood (5). Adipose cells can take up these triglycerides from very-low-density lipoprotein by the action of lipoprotein-lipase on the lining cells of the capillaries in adipose tissue (5). Hence, palmitoleic acid and desaturation indexes in adipose tissue may reflect the metabolic activity of the liver and adipose tissue. Previous animal and human studies have reported a positive association between hepatic SCD1 activity and obesity (2, 3, 9). On the basis of our results, it is possible that adipose tissue palmitoleic acid may mainly reflect the palmitoleic acid that is synthesized and released by the liver instead of by adipose tissue, and the 2 adipose desaturation indexes may chiefly reflect SCD1 activity in the liver.
Compared with adipose tissue, the liver has a unique ability to dramatically increase fatty acid synthesis in response to changes in dietary carbohydrate intake (2). Studies have shown that dietary carbohydrates can robustly increase hepatic SCD1 activity (12, 21). Thus, it is possible that dietary carbohydrate intake may alter the proportion of liver-generated palmitoleic acid in adipose tissue, which may affect the association between adipose palmitoleic acid concentrations and obesity. In the stratified analysis of the association between adipose tissue palmitoleic acid concentrations and the prevalence of obesity by tertiles of carbohydrate intake (Table 4), we observed that the PRs of obesity for high compared with low palmitoleic acid concentrations were attenuated with decreased carbohydrate intake. Because the metabolic effects of carbohydrate intake are influenced by carbohydrate mass as well as carbohydrate components (28, 29), we also performed stratified analyses by glucose and starch intakes. The positive association of adipose tissue palmitoleic acid and obesity was attenuated in people in the lower tertile of glucose intake and to a lesser extent in the lower tertile of starch intake (Table 5). We speculate that a possible explanation for this phenomenon would be that the proportion of liver-generated palmitoleic acid in adipose tissue would be lower at low intakes of carbohydrates (and mostly because of glucose), and an increased proportion of adipose-generated palmitoleic acid would attenuate the PR of obesity. Our results suggest that it is not possible to completely disentangle the effects of liver and adipose tissue SCD1 activities only by measuring adipose tissue palmitoleic acid concentrations or desaturation indexes in observational studies. Consistent with these findings, a study performed in healthy Swedish men aged from 62 to 64 y showed that adipose tissue desaturation indexes of SCD1 (16:1/16:0 and 18:1/18:0) reflected SCD1 expression in that tissue but were associated with increased insulin resistance (10).
The strengths of this study include its large sample size, the high participation rate, inclusion of a representative sample of the Costa Rican population, and the unique availability of adipose tissue measurements. Nevertheless, our observational study has several limitations. Palmitoleic acid concentrations in adipose tissue cannot be differentiated between adipose-derived and liver-derived palmitoleic acid. Because of the difficulty of directly measuring SCD1 activity, we inferred SCD1 activity by the use of product:substrate ratios as previously done (9, 18, 19). The surrogate estimates for SCD1 activity also could not distinguish between liver and adipose tissue SCD1 activity. Thus, methods that could be applied to measure adipose-derived palmitoleic acid concentrations and the exact adipose SCD1 activity need to be developed in the future to directly evaluate their effects on risk of obesity or other metabolic disorders.
In conclusion, on the basis of our data, there is no direct evidence that adipose tissue palmitoleic acid could behave as a lipokine that reduces the occurrence of obesity in humans. However, the attenuated correlation between adipose palmitoleic acid concentrations and obesity in the low dietary carbohydrate–intake group is of interest and warrants further research on the effects of dietary carbohydrates and adipose-derived palmitoleic acid on obesity.
Acknowledgments
The authors' responsibilities were as follows—JG: conducted data analyses and wrote the manuscript; HC: contributed to the interpretation of the data and proofread and edited the manuscript; SM and RG: proofread and edited the manuscript; ZW: provided assistance with the statistical methods used in analyses; and AB: supervised data analyses and main aspects of data interpretation and proofread and edited the manuscript. None of the authors reported a conflict of interest.
REFERENCES
- 1.National Institutes of Health NIH publication no. 98-4083. Bethesda, MD: National Institutes of Health, 1998 [Google Scholar]
- 2.Flowers MT, Ntambi JM. Stearoyl-CoA desaturase and its relation to high-carbohydrate diets and obesity. Biochim Biophys Acta 2009;1791:85–91 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Popeijus HE, Saris WH, Mensink RP. Role of stearoyl-CoA desaturases in obesity and the metabolic syndrome. Int J Obes (Lond) 2008;32:1076–82 [DOI] [PubMed] [Google Scholar]
- 4.Flowers MT, Ntambi JM. Role of stearoyl-coenzyme A desaturase in regulating lipid metabolism. Curr Opin Lipidol 2008;19:248–56 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Smith CM, Marks AD, Lieberman MA, Marks DB. Synthesis of fatty acids, triacylglycerols, and the major membrane lipids. : Marks DB, Lieberman MA, Smith CM, Marks' basic medical biochemistry: a clinical approach. 2nd ed Philadelphia. PA: Lippincott Williams & Wilkins, 2005:579–618 [Google Scholar]
- 6.Wang J, Yu L, Schmidt RE, et al. Characterization of HSCD5, a novel human stearoyl-CoA desaturase unique to primates. Biochem Biophys Res Commun 2005;332:735–42 [DOI] [PubMed] [Google Scholar]
- 7.Zhang S, Yang Y, Shi Y. Characterization of human SCD2, an oligomeric desaturase with improved stability and enzyme activity by cross-linking in intact cells. Biochem J 2005;388:135–42 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhang L, Ge L, Parimoo S, Stenn K, Prouty SM. Human stearoyl-CoA desaturase: alternative transcripts generated from a single gene by usage of tandem polyadenylation sites. Biochem J 1999;340:255–64 [PMC free article] [PubMed] [Google Scholar]
- 9.Warensjo E, Ohrvall M, Vessby B. Fatty acid composition and estimated desaturase activities are associated with obesity and lifestyle variables in men and women. Nutr Metab Cardiovasc Dis 2006;16:128–36 [DOI] [PubMed] [Google Scholar]
- 10.Sjogren P, Sierra-Johnson J, Gertow K, et al. Fatty acid desaturases in human adipose tissue: relationships between gene expression, desaturation indexes and insulin resistance. Diabetologia 2008;51:328–35 [DOI] [PubMed] [Google Scholar]
- 11.Ntambi JM, Miyazaki M, Stoehr JP, et al. Loss of stearoyl-CoA desaturase-1 function protects mice against adiposity. Proc Natl Acad Sci U S A 2002;99:11482–6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Miyazaki M, Kim YC, Ntambi JM. A lipogenic diet in mice with a disruption of the stearoyl-CoA desaturase 1 gene reveals a stringent requiremen of endogenous monounsaturated fatty acids for triglyceride synthesis. J Lipid Res 2001;42:1018–24 [PubMed] [Google Scholar]
- 13.Ntambi JM, Miyazaki M. Recent insights into stearoyl-CoA desaturase-1. Curr Opin Lipidol 2003;14:255–61 [DOI] [PubMed] [Google Scholar]
- 14.Park EI, Paisley EA, Mangian HJ, et al. Lipid level and type alter stearoyl CoA desaturase mRNA abundance differently in mice with distinct susceptibilities to diet-influenced diseases. J Nutr 1997;127:566–73 [DOI] [PubMed] [Google Scholar]
- 15.Smith SB, Mersmann HJ, Smith EO, Britain KG. Stearoyl-coenzyme A desaturase gene expression during growth in adipose tissue from obese and crossbred pigs. J Anim Sci 1999;77:1710–6 [DOI] [PubMed] [Google Scholar]
- 16.Enser M. Desaturation of stearic acid by liver and adipose tissue from obese-hyperglycaemic mice (ob/ob). Biochem J 1975;148:551–5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Enser M. The role of insulin in the regulation of stearic acid desaturase activity in liver and adipose tissue from obese–hyperglycaemic (ob/ob) and lean mice. Biochem J 1979;180:551–8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Paillard F, Catheline D, Duff FL, et al. Plasma palmitoleic acid, a product of stearoyl-coA desaturase activity, is an independent marker of triglyceridemia and abdominal adiposity. Nutr Metab Cardiovasc Dis 2008;18:436–40 [DOI] [PubMed] [Google Scholar]
- 19.Attie AD, Krauss RM, Gray-Keller MP, et al. Relationship between stearoyl-CoA desaturase activity and plasma triglycerides in human and mouse hypertriglyceridemia. J Lipid Res 2002;43:1899–907 [DOI] [PubMed] [Google Scholar]
- 20.Cao H, Gerhold K, Mayers JR, Wiest MM, Watkins SM, Hotamisligil GS. Identification of a lipokine, a lipid hormone linking adipose tissue to systemic metabolism. Cell 2008;134:933–44 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Miyazaki M, Dobrzyn A, Man WC, et al. Stearoyl-CoA desaturase 1 gene expression is necessary for fructose-mediated induction of lipogenic gene expression by sterol regulatory element-binding protein-1c-dependent and -independent mechanisms. J Biol Chem 2004;279:25164–71 [DOI] [PubMed] [Google Scholar]
- 22.Baylin A, Kabagambe EK, Ascherio A, Spiegelman D, Campos H. Adipose tissue alpha-linolenic acid and nonfatal acute myocardial infarction in Costa Rica. Circulation 2003;107:1586–91 [DOI] [PubMed] [Google Scholar]
- 23.Campos H, Siles X. Siesta and the risk of coronary heart disease: results from a population-based, case-control study in Costa Rica. Int J Epidemiol 2000;29:429–37 [PubMed] [Google Scholar]
- 24.Kabagambe EK, Baylin A, Allan DA, Siles X, Spiegelman D, Campos H. Application of the method of triads to evaluate the performance of food frequency questionnaires and biomarkers as indicators of long-term dietary intake. Am J Epidemiol 2001;154:1126–35 [DOI] [PubMed] [Google Scholar]
- 25.Beynen AC, Katan MB. Rapid sampling and long-term storage of subcutaneous adipose-tissue biopsies for determination of fatty acid composition. Am J Clin Nutr 1985;42:317–22 [DOI] [PubMed] [Google Scholar]
- 26.Baylin A, Kabagambe EK, Siles X, Campos H. Adipose tissue biomarkers of fatty acid intake. Am J Clin Nutr 2002;76:750–7 [DOI] [PubMed] [Google Scholar]
- 27.Spiegelman D, Hertzmark E. Easy SAS calculations for risk or prevalence ratios and differences. Am J Epidemiol 2005;162:199–200 [DOI] [PubMed] [Google Scholar]
- 28.Barclay AW, Petocz P, McMillan-Price J, et al. Glycemic index, glycemic load, and chronic disease risk–a meta-analysis of observational studies. Am J Clin Nutr 2008;87:627–37 [DOI] [PubMed] [Google Scholar]
- 29.Galgani J, Aguirre C, Diaz E. Acute effect of meal glycemic index and glycemic load on blood glucose and insulin responses in humans. Nutr J 2006;5:22. [DOI] [PMC free article] [PubMed] [Google Scholar]