Obesity in Adults Is Associated With Reduced Lung Function in Metabolic Syndrome and Diabetes: The Strong Heart Study Open Access

The Strong Heart Study (SHS) is a multicenter, population-based, prospective study of cardiovascular disease (CVD) and its risk factors among AI adults. The study design, survey methods, and laboratory techniques have been described previously (11,12). The study population is composed of tribal members who reside in study communities. The present analysis was based on the second examination and the 4-year follow-up clinic visit—the third examination. The second examination included 3,638 participants, and the third included 3,197. Approval was obtained from relevant institutional review boards, and all participants gave written informed consent.

The following criteria were used in excluding participants from the analysis population: 1) >20 pack-year smoking history (n = 639), 2) any self-reported lung problems and taking asthma medications (n = 179), 3) having CVD (n = 430), and 4) missing data on DM, MS status, or spirometry (n = 268). The final study sample consisted of 2,396 individuals, including 483 adults without MS or DM (normal group), 729 adults without DM and with MS (MS group), and 1,184 adults with DM (DM group) at the second examination. These three groups of participants were mutually exclusive. MS-free (483 normal) and DM-free (483 normal and 729 MS) participants were used for the prediction of MS and DM, respectively.

Spirometry was performed by centrally trained and certified nurses and technicians. Normal reference values for the pulmonary function test (PFT) were derived from the SHS population; SHS-specific forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) were predicted using the equations developed by Marion et al. (13) for healthy SHS participants using the covariates of age, sex, and height. The prediction equations for normal lung function for men are as follows:

The prediction equations for normal lung function for women are as follows:

Before the analysis, crude data on FVC and FEV1 were divided by predicted FVC and FEV1, respectively, to yield FVC % predicted and FEV1 % predicted.

Individuals were classified as having DM according to the 1997 American Diabetes Association criteria; a fasting glucose level of at least 7.0 mmol/L (126 mg/dL); current use of antidiabetes medication; or on renal dialysis/kidney transplant with a positive response to the question, “Has a medical person ever told you that you had diabetes?” This group included both T1DM and T2DM; the majority of the participants were T2DM.

MS was defined according to the Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) guidelines (14) as having at least three of the following five conditions: abdominal obesity (waist circumference [WC] >102 cm in men and >88 cm in women), increased triglycerides (≥150 mg/dL), reduced HDL cholesterol (<40 mg/dL in men and <50 mg/dL in women), elevated blood pressure (≥130/≥85 mmHg), and high fasting glucose (100–125 mg/dL).

The definitions and methods used for other measurements (age, education level, cigarette smoking status and pack-years of smoking, physical activity index, height, BMI, and hypertension) have been reported previously (12,15). The methods used for the measurement of fibrinogen and C-reactive protein (CRP) were also reported before (16). The homeostasis model assessment for insulin resistance (HOMA-IR) was calculated according to the following formula: (fasting insulin in μU/mL × fasting glucose in mg/dL)/405.

Characteristics of normal, MS, and DM groups were compared using ANOVA for continuous variables and χ² tests for categorical variables. Kruskal-Wallis ANOVA by ranks was used to compare total triglycerides, plasminogen activator inhibitor-1, and CRP because of skewed distributions.

Multiple linear regression models were used to describe the cross-sectional relationships between lung function and metabolic disorders (MS and DM) after adjusting for potential confounding variables including age, sex, abdominal obesity, height, hypertension, physical activity index, education level, current smoking status, and SHS center. The same models were also fitted to describe the cross-sectional associations among lung function and duration of DM, type of antidiabetes medications, and tertiles of insulin resistance after adjusting for potential confounding variables.

Multiple linear regression models were also used to describe the cross-sectional relationships between lung function and inflammatory markers (CRP and fibrinogen). For CRP analyses, participants with values >10 mg/L were excluded because CRP levels >10 mg/L may reflect an acute inflammatory process; CRP >3 mg/L was used as the high CRP cut point based on the American Heart Association/Centers for Disease Control and Prevention categories (17). For fibrinogen analyses, the lowest tertile of fibrinogen level in the normal group (≤350 mg/dL) was used as a control for the lung function comparisons. Multiple linear regression models were also carried out for investigation of the cross-sectional relationship between lung function and obesity. Cox proportional hazards models were used to analyze the association between DM/MS and pulmonary function, controlling for confounding variables. All tests of significance were two-tailed, with an α-level of 0.05. All analyses were performed using version 9.1 of the SAS statistical software package (SAS Institute, Cary, NC).

Characteristics of the three groups (normal, MS, and DM) are summarized in Table 1. Of the participants, 75.2% reported 100% AI heritage. There were no significant differences among these three groups for age. In general, participants with DM or MS were more likely to be hypertensive and smoked less than the normal group. They were also more likely to have a larger WC, higher triglycerides, lower HDL cholesterol, higher hemoglobin A_1c, presence of albuminuria, and an elevated concentration of inflammatory markers compared with the normal group. LDL cholesterol was higher in the MS group than in the normal and DM groups.

The clinical measurements of the excluded group because of missing DM, MS, or PFT status were similar to those of the study group, with the exception that they were more likely to have smaller WCs (data not shown).

Both percent predicted values for FVC and FEV1 were significantly lower in the participants with MS or DM compared with their normal counterparts (P < 0.0001) (Table 1), even after adjusting for age, sex, abdominal obesity, height, hypertension, sports activity index, education level, current smoking status, and SHS center (Table 2). An increased trend for MS and DM was observed for the FEV1-to-FVC ratio.

Significant relationships were found among pulmonary function and insulin resistance, duration of DM, and antidiabetes medications. Participants with higher HOMA-IR scores had greater reductions in both predicted FVC and FEV1 values (P_trend < 0.05) (Table 2). Subdividing the participants by duration of DM revealed that absolute and percent predicted FVC decreased with duration of DM (P_trend < 0.01). However, the reductions of FEV1 values were not different for durations <5 years vs. >10 years. Subdividing the participants by antidiabetes medications revealed that pulmonary function was significantly reduced in participants requiring insulin treatment compared with those on oral agents alone or no medication (P_trend < 0.01). The relatively greater reduction in FVC than in FEV1 in DM participants with longer duration or more severe DM was reflected in the FEV1-to-FVC ratio.

Partitioning normal, MS, and DM participants according to blood levels of the inflammatory markers CRP and fibrinogen revealed that pulmonary function decreased as marker concentration increased (Table 3). Compared with normal participants, MS and DM groups with elevated inflammatory markers had greater reductions in their PFT (FVC, FEV1, FVC % predicted, and FEV1 % predicted all P_trend < 0.01).

Among 1,212 participants who were DM-free at the SHS second examination, 129 developed DM during the 4 years of follow-up. By use of Cox proportional hazards models, in unadjusted analyses with FVC, FVC % predicted, FEV1, FEV1 % predicted, and FEV1-to-FVC ratio as continuous independent variables, FVC % predicted and FEV1-to-FVC ratio both predicted DM (Table 4, model 2). The risk of incident DM increased 3% for every 1% increase in FEV1-to-FVC ratio (hazard ratio 1.03 [95% CI 1.01–1.06]), and the risk of incident DM increased 2% for every 1% decrease in FVC % predicted (0.98 [0.97–0.99]). The same results were obtained when age, sex, and SHS center were added to the model as covariates (model 3). However, when more covariates (abdominal obesity, hypertension, physical activity index, education level, and pack-years of smoking) were added to the Cox proportional hazards model, pulmonary function did not predict DM (Table 4, model 4); abdominal obesity, as measured by WC, was retained in the final model as an independent predictor of the development of DM.

Similar analyses of data from participants who developed MS indicated that FEV1-to-FVC ratio predicted DM; however, neither FVC nor FEV1 alone predicted this syndrome. As before, abdominal obesity was retained in the final model as an independent predictor for MS (Table 4).

Further investigation of obesity showed a significant reduction in pulmonary function in obese participants measured either by WC or by BMI (Table 5). Compared with normal participants, MS and DM adults with obesity had greater reductions in their PFT (FVC, FEV1, FVC % predicted, and FEV1 % predicted all P_trend < 0.05).

In this study, adult AIs with MS or DM had significantly lower FVC, FEV1, FVC % predicted, and FEV1 % predicted compared with normal AI participants. This relationship persisted after adjustment for multiple factors, including obesity, and was related to metabolic disorders and markers of inflammation. Major strengths of the current study are the inclusion of multiple measures of metabolic disorders and the consistency of the results for all these measurements. Our results are also consistent with those of other studies (7–9) that show restrictive lung function (reduced FVC and increased FEV1-to-FVC ratio), but not obstructive pulmonary function (decreased FEV1-to-FVC ratio), to be associated with MS and DM.

Participants with MS had significantly lower FVC, FEV1, FVC % predicted, and FEV1 % predicted compared with participants without DM or MS. These relationships were graded by insulin resistance. Our results are consistent with cross-sectional studies (7,8,18).

In patients with DM, the relationships were graded by DM severity and serum markers of inflammation after the adjustment for possible confounders. Our results support cross-sectional studies, which demonstrate lower FVC and FEV1 in adults with DM compared with their nondiabetic counterparts (1,19,20), especially when DM was of longer duration and subjects required medication treatment (1), had a higher HOMA score (19) and had higher levels of serum inflammatory markers (21).

Previous studies suggest that impaired lung function predicts the subsequent development of clinical DM (5,6); studies also show that WC predicts DM beyond commonly evaluated cardiometabolic risk factors (22,23). Yet few studies have assessed whether the relationship between lung function and DM is mediated by central obesity. Leone et al. (18) found that the relationship between lung function impairment and MS was predominantly due to abdominal obesity; our data also suggest that abdominal obesity is a significant factor that affects MS, DM, and PFT.

The underlying mechanisms relating this type of metabolic disorder to reduced lung function remain unclear; integration of inflammatory and metabolic pathways in MS or DM patients may be an important underlying mechanism relating the disorders to reduced lung function (24).

In the current study, there was a significant, graded, and inverse relationship between PFT and WC and/or BMI, indicating that obesity played a significant role in the relationship of reduced PFT and metabolic disorders. There was also a significant, graded, and inverse relationship between lung function and inflammatory markers, indicating that inflammation played a significant role in the relationship of reduced PFT and metabolic disorders. These observations seem to support the suggested mediatory mechanisms of inflammation and obesity.

The strengths of this study include the community-based sample, standardized spirometric techniques, extensive data on potential confounders, and a large sample size that increased precision and permitted multiple statistical adjustments. The study’s limitations include lack of generalizability of results to heavy/prolonged smokers and the lack of data on obesity-related inflammatory markers, which precluded more detailed investigations of the causal pathway.

The main conclusions from the cross-sectional analyses are that reduced lung function is independently associated with MS and DM and that obesity and inflammation are associated with reduced lung function in MS and DM; impaired lung function presents before the development of MS or DM in AIs. Further studies are needed to investigate how inflammation and obesity affect lung function in patients with MS and DM.

The SHS was supported by cooperative agreements U01-HL41642, U01-HL41652, and U01-HL41654 from the National Institutes of Health National Heart, Lung, and Blood Institute.

No potential conflicts of interest relevant to this article were reported.

F.Y. wrote the manuscript. A.E.D. contributed to discussion and reviewed and edited the manuscript. S.M. researched data and reviewed and edited the manuscript. C.S., Y.Z., L.G.B., and D.C. contributed to discussion and reviewed and edited the manuscript. E.R.R. researched data and reviewed and edited the manuscript. E.T.L. contributed to discussion and reviewed and edited the manuscript.

Parts of this study were presented in poster form at the 70th Scientific Sessions of the American Diabetes Association, Orlando, Florida, 25–29 June 2010.

The authors acknowledge the assistance and cooperation of the participating tribes and the Indian Health Service facilities serving those tribes. The authors also thank the study participants and the directors of the SHS clinics and their staff.