| |
Prospective study of physical fitness, adiposity, and inflammatory markers in healthy middle-aged men and women
|
| |
| |
Am J Clin Nutr 89: 85-89, 2009.
Vol. 89, No. 1, 85-89, January 2009
Mark Hamer and Andrew Steptoe
1 From the Department of Epidemiology and Public Health, University College London, London, United Kingdom.
2 Supported by the Medical Research Council, United Kingdom, and the British Heart Foundation, United Kingdom.
"In summary, changes in low-grade inflammation were positively associated with adiposity but not with fitness. Fit-overweight participants do, however, appear to have a lower risk of CVD than do their unfit counterparts. Because the excess risk of CVD events associated with obesity is partly explained by inflammation-sensitive plasma proteins, the antiinflammatory effects of exercise may partly contribute to a lower CVD risk in obesity.....
....The main findings from the present study indicate that markers of inflammation after 3 y of follow-up were largely associated with adiposity and not with physical fitness. However, the data suggest independent antiinflammatory effects of fitness in overweight-obese participants. Recent evidence suggests inflammatory and hemostatic biomarkers explain a notable (about 30%) proportion of the association between physical activity and CVD (25). Because the excess risk of CVD events associated with obesity is partly explained by inflammation-sensitive plasma proteins (26), the antiinflammatory effects of exercise may partly contribute to lower CVD risk in obesity.....
.....physical fitness may be an intermediate pathway in the association between weight gain and inflammation, although in the present study we did not observe any association between fitness and weight gain at follow-up. Our findings may have been limited by a relatively short follow-up period. In addition, we did not account for changes in fitness over time, and increased fitness levels over 2 y were previously related to attenuated weight gain in healthy middle-aged adults....
....Physical fitness is associated with CVD risk factors that may mediate the antiinflammatory effects, such as insulin sensitivity (32), HDL cholesterol (31), and vascular function (33). In addition, we have recently shown that less-fit participants have elevated IL-6 responses to psychosocial stressors, which over time may translate into greater basal levels of inflammation if these types of responses are elicited regularly (34). The reason that physical fitness appeared to have stronger antiinflammatory effects in overweight-obese participants may be related to differences in central adiposity. We found significant differences in WHR of overweight-obese participants that were physically fit compared with their unfit counterparts (0.87 ± 0.07 compared with 0.91 ± 0.08; P = 0.01). Previous reports also suggest that exercise training can result in selective loss of visceral adipose tissue even without weight reduction (35), and lower fitness was associated with more visceral adipose tissue after matching persons with similar BMI (36). Given that visceral adipose tissue is thought to be important in the production of adipocytokines (37), this may explain our findings."
ABSTRACT
Background: Physical fitness may provide cardiovascular benefits in the obese.
Objective: We prospectively examined the associations between inflammatory markers and fitness, body mass index, and central adiposity.
Design: Healthy men and women (n = 176) were recruited from the Whitehall II epidemiologic cohort. At baseline we measured physical fitness and adiposity, and blood was drawn for the assessment of inflammatory markers [C-reactive protein (CRP) and interleukin-6 (IL-6)]. We subsequently assessed inflammatory markers and adiposity at the 3-y follow-up visit.
Results: Body mass index, but not physical fitness, was independently associated with IL-6 and CRP at follow-up. Weight gain was also associated with CRP at follow-up. Compared with fit-lean participants, the unfit-overweight participants had significantly higher concentrations of CRP (adjusted _: 0.67; 95% CI, 0.31, 1.00) and IL-6 (adjusted _: 0.28; 95% CI: -0.06, 0.49) at follow-up. In contrast, the fit-overweight and unfit-lean participants did not differ significantly from the fit-lean participants after adjustments for age, sex, smoking, employment grade, and baseline inflammation.
Conclusions: In participants followed up for 3 y, changes in low-grade inflammation were positively associated with adiposity but not with fitness at baseline. Further attention should focus specifically on overweight-obese participants in relation to physical fitness and cardiovascular disease risk.
INTRODUCTION
Poor physical fitness is a cardiovascular disease (CVD) risk factor (1-3). Given that physically fit, habitually active persons are at lower risk of obesity and weight gain (4), the protective effects of fitness may be partly mediated through adiposity, although this issue remains controversial. Previous epidemiologic studies have suggested that fitness is an independent predictor of CVD risk (5-7), which suggests that persons who are obese yet fit may have a lower risk of CVD than do those who are obese and unfit. However, contrasting studies suggest that obesity is a stronger risk factor, independently from fitness (8, 9). The mechanisms for the potential protective effects of fitness in obesity are incompletely understood, although specific risk markers such as inflammation have gained attention (10).
Several population studies have shown inverse associations between fitness and inflammatory markers independently of fatness (11-13), although others have suggested that obesity is a risk factor regardless of fitness (14-16). In addition, randomized controlled trials are equivocal. For example, in 152 female smokers, no changes were observed in C-reactive protein (CRP) or fibrinogen after a 12-wk exercise program that improved physical fitness, although body weight remained stable (16). In 193 sedentary, mildly obese men and women, 6-mo exercise training did not alter concentrations of CRP despite improvements in fitness and in visceral and subcutaneous adiposity (17). In contrast, significant reductions in inflammatory markers were observed after 3-6-mo exercise training without changes in body mass index (BMI; in kg/m2) or body fat in elderly participants (18, 19). In another trial that lasted for 18 mo, a dietary weight-loss intervention resulted in significant reductions in CRP, interleukin-6 (IL-6), and tumor necrosis factor- compared with the control group, although no effects were observed for an exercise intervention group (20). A number of reasons may explain these equivocal findings, including disparity in the exercise interventions, poor adherence levels, differences between characteristics of participants, and variable follow-up times. Indeed, the typical duration of most intervention studies is 3-6 mo, which may not be long enough to observe significant effects. We are unaware of any work to date that has studied the joint influence of fitness and adiposity on inflammatory markers with the use of a prospective design. The aim of the present study was therefore to examine the association between fitness, adiposity, and inflammatory markers prospectively over 3 y. We hypothesized that baseline fitness and adiposity would be independently related to inflammatory marker concentrations at follow-up.
RESULTS
Characteristics of the sample at the baseline assessment in relation to CRP concentrations at follow-up are shown in Table 1. The participants in the highest third CRP group had a higher heart rate and blood pressure during exercise testing, had greater BMIs, and were more likely to be of lower employment status. At baseline, an inverse association was observed between exercise heart rate and self-reported participation in vigorous exercise (r = -0.25, P = 0.001), which indicated that fitter participants were more vigorously active. Adiposity measures (BMI and WHR) were also associated with exercise heart rate in men and women (P < 0.05 for both), suggesting that fitness was inversely associated with adiposity at baseline. In addition, fitter participants were more likely to come from higher employment grades. Approximately 55% of the sample was classified as overweight or obese (BMI > 25.0) at baseline. At follow-up a large variation was observed in weight change, ranging from 18.4 kg weight loss to 14.7 kg weight gain (mean change: 0.13 ± 3.81 kg; P = 0.85). Forty-five participants gained >/=3% body weight, although weight gain was not associated with physical fitness or physical activity at baseline.
At baseline, a direct association was observed between exercise heart rate and both inflammatory markers in age- and sex-adjusted analyses, although after further adjustment for BMI this association only remained significant for CRP (_ = 0.25, P = 0.001). A significant increase in IL-6 was observed at follow-up compared with baseline (1.51 ± 0.73 pg/mL compared with 1.12 ± 0.61 pg/mL; P < 0.001, respectively). BMI, but not physical fitness, was independently associated with IL-6 and CRP at follow-up (Tables 2 and 3). Weight change was also a significant predictor of CRP at follow-up. A similar pattern of results emerged with the use of baseline WHR or change in WHR instead of BMI. For example, WHR was independently associated with IL-6 (_ = 0.35, P = 0.001) and CRP (_ = 0.21, P = 0.005) at follow-up. Exercise heart rate (fitness) was also retained in this model as a significant predictor of CRP (_ = 0.15, P = 0.028), although not for IL-6.
In further analyses we examined the interaction of fitness and BMI on inflammatory markers at follow-up (Table 4). The results show that in relation to fit-lean participants (n = 52), the unfit-overweight participants (n = 60) had significantly higher concentrations of CRP and IL-6 at follow-up. In contrast, the fit-overweight (n = 34) and unfit-lean (n = 30) participants did not differ significantly from the fit-lean participants.
DISCUSSION
The main findings from the present study indicate that markers of inflammation after 3 y of follow-up were largely associated with adiposity and not with physical fitness. However, the data suggest independent antiinflammatory effects of fitness in overweight-obese participants. Recent evidence suggests inflammatory and hemostatic biomarkers explain a notable (30%) proportion of the association between physical activity and CVD (25). Because the excess risk of CVD events associated with obesity is partly explained by inflammation-sensitive plasma proteins (26), the antiinflammatory effects of exercise may partly contribute to lower CVD risk in obesity. More than 40 observational studies have examined associations between physical fitness or self-reported physical activity levels and inflammatory markers. Two-thirds of the observational studies have reported an inverse relation between inflammatory factors and fitness or physical activity after adjustment for adiposity, although evidence from randomized controlled trials is less consistent (10). The advantage of using the prospective design is that the potential biases that are present in cross-sectional analyses can be reduced, and the exposure effects can be studied over a greater length of time in comparison with most interventions.
We observed an association between weight gain and CRP at follow-up. This finding is consistent with data from recent prospective studies (27, 28). In the Finnish Birth Cohort study, weight gain between 14 and 31 y of age, as well as lower birth weight, was positively associated with CRP concentrations in adulthood (28). Other studies have shown that maintaining vigorous activity attenuated 7-y weight gain in 8340 runners (4), and higher maximal oxygen uptake was associated with lower odds of obesity at 15-y follow-up in the Canadian Physical Activity Longitudinal Study (29). Thus, physical fitness may be an intermediate pathway in the association between weight gain and inflammation, although in the present study we did not observe any association between fitness and weight gain at follow-up. Our findings may have been limited by a relatively short follow-up period. In addition, we did not account for changes in fitness over time, and increased fitness levels over 2 y were previously related to attenuated weight gain in healthy middle-aged adults (30).
A number of mechanisms may independently link both fitness and fatness with inflammatory pathways. The adipocytes produce a number of metabolically active hormones, such as cytokines, acute-phase reactants, leptin, and adiponectin, that have both antiinflammatory and proinflammatory properties (31). Physical fitness is associated with CVD risk factors that may mediate the antiinflammatory effects, such as insulin sensitivity (32), HDL cholesterol (31), and vascular function (33). In addition, we have recently shown that less-fit participants have elevated IL-6 responses to psychosocial stressors, which over time may translate into greater basal levels of inflammation if these types of responses are elicited regularly (34). The reason that physical fitness appeared to have stronger antiinflammatory effects in overweight-obese participants may be related to differences in central adiposity. We found significant differences in WHR of overweight-obese participants that were physically fit compared with their unfit counterparts (0.87 ± 0.07 compared with 0.91 ± 0.08; P = 0.01). Previous reports also suggest that exercise training can result in selective loss of visceral adipose tissue even without weight reduction (35), and lower fitness was associated with more visceral adipose tissue after matching persons with similar BMI (36). Given that visceral adipose tissue is thought to be important in the production of adipocytokines (37), this may explain our findings.
The limitations of this study should be recognized. We used a submaximal exercise test with a relatively light workload to assess fitness levels but did not obtain an indication of maximal aerobic capacity. The heart rate response to exercise is strongly predictive of physical fitness, and it was not feasible to perform tests of maximal capacity in the present sample of older, sedentary adults. In addition, maximal exercise tests also have limitations because of individual variability in exercise tolerance that can cause some participants to terminate a test before reaching volitional exhaustion (38). We did not measure fitness levels at follow-up; thus, we cannot account for the effect of changes in fitness over time. However, because exercise interventions generally show modest improvements in aerobic capacity (39), it is unlikely that small changes in physical activity would have caused significant changes in fitness during the 3-y observation period. This study had several strengths, including the careful selection of participants free of inflammatory diseases at the baseline assessment and control for factors such as smoking and social status, which can confound associations between fitness, adiposity, and inflammatory markers.
In summary, changes in low-grade inflammation were positively associated with adiposity but not with fitness. Fit-overweight participants do, however, appear to have a lower risk of CVD than do their unfit counterparts. Because the excess risk of CVD events associated with obesity is partly explained by inflammation-sensitive plasma proteins, the antiinflammatory effects of exercise may partly contribute to a lower CVD risk in obesity.
SUBJECTS AND METHODS
Participants
Participants were drawn from the Whitehall II epidemiologic cohort (21) for a substudy in 1999-2000, which served as the baseline assessment for the present study. Participants were invited for a follow-up assessment 3 y later. The criteria for entry into the study included no history or objective signs of coronary heart disease and no previous diagnosis or treatment of hypertension, inflammatory diseases, or allergies. Volunteers were of white European origin, were aged 45-59 y, lived in the London area, and were employed full-time, which was dictated by the overall characteristics of the main cohort. Selection was stratified by grade of employment to include participants of higher and lower socioeconomic status. The participation rate was high, approximately 92%. Participants declining invitation to take part were generally from lower work grades. Participants were prohibited from using any antihistamine or antiinflammatory medication 24 h before testing and were rescheduled if they reported colds or other infections on the day of testing. At baseline there were 214 participants, although follow-up data were unavailable in 38, leaving 176 volunteers in the final analysis (98 men, 78 women). Participants excluded did not differ significantly from those included in the final analyses. Participants gave full informed consent to participate in the study, and ethical approval was obtained from the University College London Hospital committee on the Ethics of Human Research.
Baseline assessment
Height and weight were recorded, while the subjects were wearing light clothing, for the calculation of BMI. Waist circumference was measured with a metal anthropometric tape midway between the lower rib margin and the iliac crest, and hip circumference was measured at the level of the great trochanters to calculate waist-to-hip ratio (WHR). After the insertion of a venous cannula, participants rested for 30 min before the collection of blood for the assessment of inflammatory markers. Peripheral blood was collected in EDTA-coated tubes and centrifuged at room temperature. All blood samples were frozen at -70C until assay. The analysis of plasma IL-6 and CRP concentrations was performed with the use of high-sensitivity enzyme-linked immunosorbent assay (R&D Systems, Oxford, United Kingdom), with intraassay and interassay CVs of <10%. Blood pressure was measured continuously with the use of a Portapres-2 device (Finapres Medical Systems, Amsterdam, Netherlands), and an average from the last 5 min of the rest period was calculated. Demographic details (age, education, grade of employment as an indicator of socioeconomic status) and information about health-related behaviors (units of alcohol intake per week, smoking, physical activity) were also recorded. For physical activity, participants reported how many times a week and for how long they engaged in activity that was mildly energetic (eg, gardening, general housework), moderately energetic (eg, brisk walking, golf, cycling), and vigorous (eg, running, tennis, squash, swimming). These measures were associated with cardiovascular and all-cause mortality in the original Whitehall study (22). Because cardiorespiratory fitness is most closely associated with participation in vigorous activity, we used this variable in the present analyses.
At the baseline assessment a submaximal exercise test was performed that was used as an indicator of physical fitness. This test was performed only as part of the substudy and not in the main cohort. Participants exercised on a cycle ergometer (model 864; Monark, Varberg, Sweden) for 8 min at a constant workload of 50 W. Heart rate and blood pressure were measured throughout the test, and the average heart rate during the last 2 min was used as a measure of physical fitness, which is widely regarded as a reliable indicator (23). Cycle ergometry testing is one of the most popular methods to assess physical fitness because there are relatively small individual variations in mechanical efficiency with this exercise method. A submaximal test with a relatively light workload was adopted in the present study because the sample consisted of older, mostly sedentary persons. Exercise tests were stopped if participants reported any contraindications such as chest pain, although no such events were recorded in the present study.
Follow-up assessment
At follow-up, blood was drawn for the measurement of IL-6 and CRP. In addition, height, weight, WHR, and blood pressure were assessed.
Statistical analysis
Log transformations were performed on IL-6 and CRP to normalize the data (24). Ten participants were excluded from IL-6 analyses because of missing data. Linear regression analyses were used to examine the relation between baseline fitness and adiposity with markers of inflammation at follow-up. Age, sex, and IL-6 or CRP concentrations at baseline were entered in step 1; followed by exercise heart rate, weight change, and baseline BMI in step 2; followed by employment grade, smoking, and blood pressure in step 3. Additional models were also tested with the use of WHR instead of BMI. Results of these regression analyses are presented with standardized β coefficients. We fitted various sex interaction terms (eg, sex x fitness and sex x BMI) into the models. However, because there was no evidence of any sex interaction, we present all analyses for men and women combined. To examine whether physical fitness and BMI interact, we created a categorical variable for the various combination of groups of BMI and physical fitness (fit-lean, unfit-lean, fit-overweight, unfit-overweight). The fitness x BMI groups were devised by taking a medium split of the sex-specific heart rate response to the fitness test (high-low fit), and lean or overweight was defined as BMI < or >/=25, respectively. This additive interaction term was fitted in a separate general linear model that included age, sex, baseline inflammatory markers, smoking, and employment grade as covariables.
|
|
| |
| |
|
|
|
