original contributions
nature publishing group
See reviewer commentary page 457
Influence of Leptin, Adiponectin, and Resistin on the Association Between Abdominal Adiposity and Arterial Stiffness B. Gwen Windham1, Michael E. Griswold2, S. Morteza Farasat3, Shari M. Ling4, Olga Carlson5, Josephine M. Egan5, Luigi Ferrucci4 and Samer S. Najjar6
Methods This is a cross-sectional analysis of data from the Baltimore Longitudinal Study of Aging (BLSA). Adiposity was measured as kilograms of abdominal adipose tissue using dual-energy X-ray absorptiometry (DXA). Arterial stiffness was assessed as carotid– femoral pulse wave velocity (PWV). Leptin, adiponectin, and resistin were assayed in fasting serum samples. The influence of adipokines on the relationship between adiposity and arterial stiffness by adipokines was examined using standard mediation pathway analysis. Results Among 749 participants ages 26–96 years (mean age 67, 52% men, 27% black), abdominal adiposity was positively associated with
Central arterial stiffness, a hallmark of arterial aging,1 is associated with increased risk of cardiovascular events, allcause and cardiovascular mortality.2,3 Obesity, an established risk factor for morbidity and mortality, has reached epidemic proportions in the United States, with an estimated prevalence of 34%, or over 72 million Americans ages ≥20 years.4 Obesity is an important risk factor for arterial stiffness,5 but the mechanisms for this association are poorly understood. Adipose tissues produce several hormones such as leptin, a diponectin, 1Department of Internal Medicine, Division of Geriatric Medicine, University of Mississippi Medical Center, Jackson, Mississippi, USA; 2Department of Biostatistics, University of Mississippi Medical Center, Jackson, Mississippi, USA; 3MedStar Research Institute, Baltimore, Maryland, USA; 4Clinical Research Branch, National Institute on Aging (NIA), NIH, Baltimore, Maryland, USA; 5Laboratory of Clinical Investigation, NIA, NIH, Baltimore, Maryland, USA; 6Laboratory of Cardiovascular Science, NIA, NIH, Baltimore, Maryland, USA. Correspondence: B. Gwen Windham (
[email protected])
Received 9 October 2009; first decision 7 November 2009; accepted 1 January 2010; advance online publication 11 February 2010. doi:10.1038/ajh.2010.8 © 2010 American Journal of Hypertension, Ltd.
PWV (relative ratio (RR) = 1.04, P = 0.02), after adjusting for potential confounders but was attenuated and no longer significant after adjusting for leptin (RR = 0.99, P = 0.77). The relationship between adiposity and PWV was not substantially influenced by adiponectin (RR = 1.03, P = 0.06) or resistin (RR = 1.05, P = 0.010). Leptin (RR = 1.02, P < 0.001), resistin (RR = 0.92, P < 0.0001), and adiponectin (RR = 0.97, P = 0.004), but not abdominal adiposity (RR = 1.00, P = 0.94), retained significant associations with PWV when adjusting for each other and confounders. Conclusions Our findings are consistent with the hypothesis that leptin explains, in part, the observed relationship between abdominal adiposity and arterial stiffness. Adiponectin, leptin, and resistin are independent correlates of PWV. Keywords: arteriosclerosis; blood pressure; elasticity; epidemiology; hypertension; obesity Am J Hypertens 2010; 23:501-507 © 2010 American Journal of Hypertension, Ltd.
and resistin. Leptin was initially identified as having important roles in regulating appetite and energy metabolism; however, leptin and other adipokines are increasingly being recognized as exerting pleiotropic effects in various organ systems, including the vasculature. For example, adipokines are associated with factors known to modulate arterial stiffness such as inflammation, sympathetic activity of the autonomic nervous system, hypertrophy, and proliferation of vascular smooth muscle cells, cell adhesion molecules, hyperglycemia, and hyperinsulinemia (see refs. 6,7). Indeed, generally small studies conducted in highly selected populations have reported associations between arterial stiffness, and leptin or adiponectin.8–10 We hypothesized that adipokines may explain the association between obesity and arterial stiffness. Therefore, the purpose of this study was to examine the influence of adipokines such as leptin, adiponectin, or resistin, independent of other factors known to affect vascular stiffness, on the relationship between abdominal adiposity and pulse wave velocity (PWV).
AMERICAN JOURNAL OF HYPERTENSION | VOLUME 23 NUMBER 5 | 501-507 | May 2010
501
Downloaded from http://ajh.oxfordjournals.org/ at Rowland Medical Library on June 19, 2015
Background Adiposity is associated with arterial stiffness, and both adiposity and arterial stiffness independently predict morbidity and mortality. Because adipocytes account for most adipokine production, the objectives of this study were to examine the influence of adipokines such as leptin, adiponectin, and resistin on the relationship between abdominal adiposity and arterial stiffness.
original contributions Methods
MAP = diastolic blood pressure + (systolic blood pressure − diastolic blood pressure)/3. Total body DXA was performed using the Prodigy Scanner (General Electric) and analyzed with version 10.51.006 software (General Electric). Fat mass of the trunk was used as a measure of abdominal adiposity as used and validated in previous studies.13,14 502
Fasting serum samples were used to assay leptin and resistin by enzyme-linked immunosorbent assays (Linco Research, St Charles, MO; and Alpco Diagnostics, Salem, NH) with intra-assay variations of 1.09–4.98% and 2.8–3.4%, and interassay variations of 3.9–5.3% and 5.1–6.9%, respectively; adiponectin was assayed by a radioimmunoassay (Linco Research), intra-assay and interassay variations of 1.8–6.2% and 6.9–9.2%. Fasting lipids and creatinine were measured by standard clinical methods and glucose by glucose oxidase method (Beckman Instruments, Brea, CA). All participants provided informed consent. This study complied with the ethical rules for human experimentation as stated in the Declaration of Helsinki and was approved by the MedStar Research Institute Institutional Review Board. Statistical analysis. Analyses were conducted using generalized linear models with a gamma distribution and log-link to account for skewed distributions of the continuous outcome variables.15 The log-link leads to results presented as relative ratios (RRs), i.e., relative increases or decreases in the expected outcomes for each unit increase in the predictor, similar to relative risks for binary data. Robust variance estimates were constructed using the Huber–White sandwich estimator. Potential nonlinear relationships were examined through nonparametric smooth average response curves and spline formulations. Outliers were examined with “DFFITS” for influential points, Cook’s D statistic, and graphical displays such as residual and added-variable plots. To examine the influence of adipokines on the relationship between abdominal adiposity and PWV, we used a mediation pathway approach (see Supplementary Figure S1 online).16,17 A series of simple and multiple linear regression models were constructed. Multivariate models were adjusted for age, sex, race, mean arterial pressure, heart rate, high-density lipoprotein- and low-density lipoprotein-cholesterol, triglycerides, fasting glucose, smoking, diabetes mellitus, antihypertensive medication use, and serum creatinine. First, we examined the relationships between adiposity and each adipokine. Second, we examined the relationships between each adipokine and PWV. Third, we examined the relationship between abdominal adiposity and PWV. Fourth, the relationship between PWV and abdominal adiposity was then assessed adjusting for each adipokine and confounders. Interaction terms were included to assess race- or sex-based differences in the relationship between specific adipokines and arterial stiffness, and to assess modifying effects of adiposity on the relationship between each adipokine and PWV. Sex-specific cut points were used to create categorical variables for abdominal adiposity and adipokines. Statistical significance was defined as α < 0.05. We had full access to the data and take responsibility for its integrity. We have read and agreed to the article as written. Results
Sample characteristics of the population are shown in Table 1. Participants’ ages ranged from 26 to 96 years, with a mean age of 67 years; 52% were men and 27% were African Americans. May 2010 | VOLUME 23 NUMBER 5 | AMERICAN JOURNAL OF HYPERTENSION
Downloaded from http://ajh.oxfordjournals.org/ at Rowland Medical Library on June 19, 2015
The Baltimore Longitudinal Study of Aging (BLSA) is an observational study conducted by the Intramural Research Program of the National Institutes of Health, National Institute on Aging that began in 1958 to study normative aging in a volunteer cohort of healthy persons at ≥17 years of age at study entry. Participants are enrolled if they are healthy at baseline (e.g., no evidence of diabetes, stroke, heart disease, or heart failure) but remain in the study if disease develops. Currently, ~1,100 men and women actively participate in the BLSA and are invited to visit the National Institute on Aging Clinical Research Branch every 1–4 years depending on their age, with participants who are ≥80 years invited back annually. Most reside in the Eastern United States, predominantly the Baltimore–Washington DC area. In 2003, new standardized protocols were implemented to measure the adipokines such as leptin, resistin, and adiponectin, and PWV along with concurrent measurements of dual-energy X-ray absorptiometry (DXA), fasting lipids, serum creatinine, insulin, and glucose. Current data include participants’ most recent visit through September 2007. Participants underwent a standardized medical examination, and medical and socio demographic interviews. All invited participants did not return for their scheduled visits. Exclusions for blood tests, PWV, or DXA were refusal, unstable medical conditions, or for DXA, bilateral hip replacements. Cross-sectional data inclusive of requisite data for this analysis were available on 749 participants. Carotid–femoral PWV is the gold standard for noninvasive measurement of central arterial stiffness.11 Subjects rested in the supine position in a quiet room for at least 10 min. Blood pressure was then measured with an oscillometric device (Dash 4000; General Electric, New York, NY) using appropriately sized cuffs. Next, pressure waveforms in the right common carotid artery and right femoral artery were acquired, and the pulse transit time between these two sites was automatically determined by the Complior SP device (Artech Medical, Pantin, France), as previously described and validated.12 The distance traveled by the pulse wave was estimated from body surface measurements, and PWV was calculated by dividing the estimated distance between the two arterial sampling sites by the pulse transit time. Measurements were made in triplicate and averaged for the analyses. Because PWV is influenced by distending pressures, which are indexed by mean arterial pressure (MAP), MAP was calculated for inclusion in the adjusted models. The following formula was used to calculate MAP from the single blood pressure measurements acquired immediately prior to the PWV assessment using the equation:
Adipokines, Adipose, Arterial Stiffness
original contributions
Adipokines, Adipose, Arterial Stiffness
For the relationships between adiposity and the adipokines (Table 2), each additional 10 kg of abdominal adiposity was associated with 2.9-fold higher leptin (RR = 2.90, P < 0.0001), 16% lower adiponectin (RR = 0.84, P < 0.001), and 10% higher resistin (RR = 1.10, P = 0.053), after adjusting for covariates. The relationships between adipokines and PWV (Table 3) revealed positive associations between PWV and leptin (RR = 1.02, P ≤ 0.0001) and negative associations between PWV and both resistin (RR = 0.90, P < 0.0001) and adiponectin (RR = 0.97, P < 0.001). When the effects of adiposity on PWV were examined without considering adipokine levels, each additional 10 kg increase in adiposity was associated with ~7% higher PWV
Mean (s.d.), median (interquartile range), or N (%)
Variable Age (years)
66.9 (14.3)
Abdominal adiposity (kg)
14.1 (5.5)
Male sex
389 (52%)
Black race
200 (27%)
Body mass index (kg/m2)
26.6 (4.4)
Mean arterial pressure (mm Hg)
87 (11.0)
Systolic blood pressure (mm Hg)
127 (18)
Diastolic blood pressure (mm Hg)
66 (9)
Heart rate (bpm)
65 (9.8)
Leptin (ng/ml)
14.9 (7.1–32.8)
Adiponectin (µg/ml)
10.7 (6.4–17.4)
Resistin (ng/ml)
4.7 (3.2–7.3)
Glucose (mg/dl)
93.8 (16.0)
Total cholesterol (mg/dl)
192.9 (38.2)
HDL-cholesterol (mg/dl)
59.2 (16.8)
LDL-cholesterol (mg/dl)
113.6 (34.6)
Triglycerides (mg/dl)
100.6 (50.2)
Creatinine (mg/dl)
1.03 (0.31)
Never smoked
397 (53%)
Blood pressure medication
335 (45%)
Diabetes mellitus
58 (8%)
bpm, beats/min; HDL, high-density lipoprotein; LDL, low-density lipoprotein.
Table 2 | Relative increase in adipokines per 10 kg increase in abdominal adiposity Dependent variable
Abdominal adipose (per 10 kg)
Leptin
Adiponectin
Resistin
Unadjusted RR (P value) (95% CI)
Adjusted for confounders RR (P value) (95% CI)
Unadjusted RR (P value) (95% CI)
Adjusted for confounders RR (P value) (95% CI)
Unadjusted RR (P value) (95% CI)
Adjusted for confounders RR (P value) (95% CI)
2.56 (
