Abstract
OBJECTIVE
We prospectively examined the relationship between lung function and risk of type-2 diabetes and fatal and nonfatal coronary heart disease (CHD) events and investigated the hypothesis that inflammation may underlie these associations.
RESEARCH DESIGN AND METHODS
A prospective study of 4,434 men aged 40–59 years with no history of cardiovascular disease (CHD or stroke) or diabetes drawn from general practices in 24 British towns and followed up for 20 years.
RESULTS
There were 680 major CHD events (276 fatal, 404 nonfatal) and 256 incident type 2 diabetes during the 20 years follow-up. Forced vital capacity (FVC) and forced expiratory volume in 1 s (FEV1) but not FEV1-to-FVC ratio were significantly and inversely associated with incident type 2 diabetes and fatal CHD events (not nonfatal events) after adjustment for age, potential confounders, and metabolic risk factors. The adjusted relative risk (RR) for type 2 diabetes (Quartile 1 vs. Quartile 4) were 1.59 (1.07–2.56) and 1.74 (1.16–2.61) for FVC and FEV1, respectively (P = 0.03 and P = 0.04 for trend). The corresponding RR for fatal CHD were 1.48 (1.00–2.21) and 1.81 (1.19–2.76) (P = 0.002 and P = 0.0003 for trend). Lung function was significantly and inversely associated with C-reactive protein and interleukin-6; the inverse associations with type 2 diabetes for FVC and FEV1 were attenuated after further adjustment for these factors (P = 0.14 and P = 0.11 for trend) but remained significant for fatal CHD (P = 0.03 and P = 0.01, respectively).
CONCLUSIONS
Restrictive rather than obstructive impairment of lung function is associated with incident type 2 diabetes (and fatal CHD) with both associations partially explained by traditional and metabolic risk factors and inflammation.
The association between impaired lung function and cardiovascular disease (CVD) morbidity and mortality is well established (1,2) although the mechanisms remain unclear. Reduced lung function is also one of the many clinical features associated with type 2 diabetes (3,4). Diabetes has been shown to be associated with impaired pulmonary function in a restrictive pattern (3), and prospective studies have suggested that reduced lung function may be associated with the development of type 2 diabetes (5–8). It has been suggested that inflammation may be a possible mechanism underlying the association between lung function and incident type 2 diabetes and CVD (6,7,9). Moreover, it has been suggested that lung function is more strongly associated with risk of fatal rather than nonfatal coronary heart disease (CHD) events and has been implicated in increased fatality following a cardiac event (10). This is of interest given that recent studies have suggested that markers of inflammation, which are associated with lung function (11), may be more strongly associated with risk for fatal than risk for nonfatal vascular events (12,13). While inflammatory markers have been shown to contribute to—but not fully explain—the increased CVD risk associated with reduced lung function (9), the contributing role of inflammation to the associations between lung function and diabetes and fatal and nonfatal CHD events separately has yet to be explored. We have examined the prospective relationship between measures of lung function (including forced vital capacity [FVC], forced expiratory volume in 1 s [FEV1], and FEV1-to-FVC ratio) and incident type 2 diabetes as well as with major CHD, fatal CHD, and nonfatal myocardial infarction (MI) separately in middle-aged men and assessed the contributing role of inflammation (C-reactive protein [CRP] and interleukin-6 [IL-6]) to these associations over and above a comprehensive panel of traditional and metabolic risk factors. The findings may also provide further insight into understanding the underlying associations between type 2 diabetes and CHD.
RESEARCH DESIGN AND METHODS
The British Regional Heart Study is a large prospective study of CVD comprising 7,735 men aged 40–59 years and drawn from general practices in each of 24 towns in England, Wales, and Scotland in 1978–1980. The criteria for selecting the town, the general practice, and the subjects, as well as the methods of data collection have been reported (14). Research nurses administered a standard questionnaire that included questions on lifestyle and medical history. Physical measurements including height and weight were made, and venous nonfasting blood samples were obtained to prepare serum for the measurement of biochemical and hematological variables. CRP and IL-6 measurements were only available for men in the 7th to 24th towns visited (n = 4,877). Details of lifestyle factor classifications, social class, and blood pressure and blood lipid measurements have been reported (14–16). The men were asked to recall a doctor diagnosis of CHD, stroke, diabetes, and a number of other disorders. Men with recall of a doctor's diagnosis of CHD or stroke, known diabetics at screening, and with asymptomatic hyperglycemia ≥11.1 mmol/l were excluded (n = 371). We also excluded men with missing spirometry function (n = 72). After these exclusions, data were available for a group of 4,434 men, who became the subjects of this study.
CRP
In 1,531 men who were included in an earlier case control analysis, CRP was measured using a sensitive enzyme immunoassay (17). In the remainder, CRP was assayed in Glasgow by ultrasensitive nephelometry (Dade Behring, Milton Keynes, U.K.); intraassay and interassay coefficients of variation were 4.7 and 8.3%, respectively. Using the results of a calibration study in 295 subjects whose samples were assayed using both methods, the results of the earlier CRP assays were adjusted to the Glasgow assay, for which the current CRP international standard was used by subtracting −0.128 from the earlier log CRP assay (87.9% of the original value). In a sensitivity analysis, we restricted the analyses to the 3,342 men with CRP assayed by the ultrasensitive nephelometric method. The main findings remained unchanged.
IL-6
IL-6 was assayed using a high-sensitivity ELISA (R & D Systems, Oxford, U.K.). Intraassay and interassay coefficients of variation were 7.5 and 8.7%, respectively. IL-6 data were missing in 183 men.
Lung function
FEV1 and FVC were measured using a Vitalograph spirometer (model J49-B2) with the subject seated. Two consecutive readings were made 15 s apart and, as per convention, the maximum of these two readings was used. FVC is the maximum volume of air expired during forced expiration and is primarily an indicator of lung volume. FEV1 is the volume of air expired in the first second of forced expiration and is influenced by lung volume and airflow obstruction. Cole (18) has shown that dividing by the height2 is the most appropriate way of standardizing lung function for stature, and this yielded the best fit to the data (19). FEV1 and FVC were height standardized to the average height (1.73 m) in the study; thus, height standardization was FEV1 (FVC) = FEV1 (FVC) multiplied by (1.73/height)2. The FEV1-to-FVC ratio, an indicator of airflow obstruction, was defined as the ratio of raw FEV1 to FVC. While not presented in the main text, we also considered alternative analyses using percentage predicted FEV1 (raw FEV1 divided by predicted FEV1) and percentage predicted FVC (raw FVC divided by predicted FVC). Predicted FEV1 and FVC were calculated using equations from a reference population of nonsmokers that included age and height as suggested by the European Respiratory Society guidelines (20).
Preexisting undiagnosed CHD and presence of arrhythmia
The World Health Organization (WHO) Rose chest pain questionnaire was administered to all men at the initial examination, and a standard resting 3-lead electrocardiogram (ECG) was recorded at rest (15). A normal rhythm was defined as sinus rhythm, coronary sinus rhythm, or sinus arrhythmia. All other statements of rhythm were treated as an arrhythmia e.g., sinus rhythm with ventricular extrasystoles (15). Preexisting undiagnosed evidence of CHD included evidence of ischemic heart disease on the WHO (Rose) chest pain questionnaire or electrocardiographic evidence of myocardial ischemia or infarction in the absence of a doctor's diagnosis of CHD (15).
Follow-up
All the men have been followed up for all-cause mortality, cardiovascular morbidity, and development of type 2 diabetes from the initial screening in January 1978–July 1980 to June 2008, and follow-up has been achieved for 99% of the cohort (15,16). This analysis is based on a 20-year follow-up from initial screening for each man. Information on deaths was collected through the established “tagging” procedures provided by the National Health Service registers. Fatal CHD events were defined as death with CHD (ICD-9 codes 410–414) as the underlying code. A nonfatal MI was diagnosed according to WHO criteria. Patients who survived a first MI for more than 28 days and who died thereafter during follow-up were included in the nonfatal group. Case fatality is defined as the proportion of those major CHD events in which death occurred during the 28 days after the first event and in which the death certificate recorded CHD (ICD-9 codes 410–414). Evidence regarding diabetes was obtained by reports from general practitioners and by biennial reviews of the patients' notes (including hospital and clinic correspondence) through to the end of the study period. Cases are based on self-reported diagnoses confirmed by primary care records, an approach which has been validated in the present study.
Statistical methods
The men were divided by quartiles of FVC, FEV1, and FEV1-to-FVC ratio. Cox's proportional hazards model was used to assess the multivariate-adjusted relative risk (RR) for each quartile compared with the reference group (lowest quartile). In the adjustment, age, BMI, systolic blood pressure (sBP), HDL cholesterol, triglycerides, total cholesterol, blood glucose, CRP, and γ-glutamyl transferase (GGT) (which has been shown to be a strong predictor of type 2 diabetes) (16) were fitted as continuous variables; all other confounders were fitted as categorical variables. Smoking was fitted as five groups (never, former smokers, 1–19 cigarettes per day, 20–39 cigarettes per day, and 40+ cigarettes per day). Differences for associations between lung function measures and MI/CHD deaths and nonfatal events were evaluated using likelihood ratio tests based on methods of competing risk survival analysis (21). The likelihood ratio tests evaluated the hypothesis that the associations of lung function were the same for MI/CHD deaths and nonfatal MI.
RESULTS
In the 4,434 men with no prevalent diabetes or CHD, there were 256 incident cases of type 2 diabetes and 680 major CHD events (276 fatal, 404 nonfatal) during the 20-year follow-up period.
Table 1 shows baseline characteristics and levels of biological markers according to quartiles of FVC and the FEV1-to-FVC ratio. FVC was strongly and significantly associated (inversely) with age, current smoking, physical inactivity, manual social class, preexisting evidence of undiagnosed CHD (but not arrthymias), BMI, sBP, and metabolic risk factors (triglycerides, blood glucose), markers of inflammation (CRP, IL-6, white cell count), and GGT. Similar associations were seen for FEV1 with the exception of BMI and blood glucose, which showed little association with FEV1. By contrast, the FEV1-to-FVC ratio was positively associated with BMI and blood glucose; no association was seen with GGT.
Table 1.
Quartile |
P value | ||||
---|---|---|---|---|---|
1 | 2 | 3 | 4 | ||
FVC (l) | <3.97 | 3.97– | 4.43– | 4.87– | |
Age (years) | 53.1 | 51.4 | 49.4 | 47.6 | <0.0001 |
Current smokers (%) | 49.6 | 42.5 | 36.3 | 38.2 | <0.0001 |
Inactive (%) | 43.8 | 38.7 | 34.8 | 31.4 | <0.0001 |
Heavy drinkers (%) | 11.4 | 10.7 | 10.6 | 11.5 | 0.88 |
Manual (%) | 67.6 | 62.1 | 58.3 | 57.1 | <0.0001 |
Evidence of CHD (undiagnosed) (%) | 25.9 | 22.1 | 19.4 | 17.0 | <0.0001 |
Silent MI (ECG) (%) | 2.5 | 3.0 | 1.3 | 0.7 | 0.0002 |
Arrhythmia (%) | 5.9 | 6.2 | 6.0 | 5.4 | 0.87 |
BMI (mg/k2) | 25.6 | 25.5 | 25.5 | 25.0 | <0.0001 |
HDL cholesterol (mmol/l) | 1.14 | 1.15 | 1.15 | 1.17 | 0.11 |
Triglycerides (mmol/l) | 1.76 | 1.72 | 1.71 | 1.57 | <0.0001 |
Cholesterol (mmol/l) | 6.24 | 6.28 | 6.30 | 6.16 | 0.12 |
sBP (mmHg) | 148.1 | 145.6 | 145.4 | 143.1 | <0.0001 |
Glucose (mmol/l) | 5.53 | 5.47 | 5.47 | 5.42 | <0.0001 |
CRP (mg/l) | 1.97 | 1.31 | 1.11 | 0.91 | <0.0001 |
IL-6 (pg/ml) | 2.89 | 2.39 | 2.14 | 2.01 | <0.0001 |
WCC (×109/l) | 7.17 | 6.95 | 6.89 | 6.82 | <0.0001 |
GGT (IU/l) | 16.4 | 15.2 | 14.7 | 13.7 | <0.0001 |
FEV1-to-FVC ratio | <0.71 | 0.71– | 0.77– | 0.82– | |
Age (years) | 52.2 | 50.8 | 49.7 | 48.7 | <0.0001 |
Current smokers (%) | 57.9 | 46.8 | 37.0 | 25.2 | <0.0001 |
Inactive (%) | 41.1 | 42.3 | 33.1 | 32.2 | <0.0001 |
Heavy drinkers (%) | 15.2 | 12.3 | 9.4 | 9.1 | <0.0001 |
Manual (%) | 71.2 | 64.0 | 56.8 | 53.0 | <0.0001 |
Evidence of CHD (undiagnosed) (%) | 25.0 | 20.2 | 18.8 | 20.5 | 0.001 |
Silent MI (ECG) (%) | 2.3 | 1.5 | 2.0 | 1.8 | 0.57 |
Arrhythmia (%) | 6.3 | 5.5 | 5.8 | 5.7 | 0.87 |
BMI (kg/m2) | 24.7 | 25.4 | 25.5 | 26.0 | <0.0001 |
HDL cholesterol (mmol/l) | 1.17 | 1.15 | 1.14 | 1.14 | <0.0001 |
Triglycerides (mmol/l) | 1.60 | 1.75 | 1.69 | 1.72 | <0.0001 |
Cholesterol (mmol/l) | 6.09 | 6.32 | 6.27 | 6.31 | <0.0001 |
sBP (mmHg) | 146.8 | 146.3 | 144.7 | 144.4 | <0.0001 |
Glucose (mmol/l) | 5.42 | 5.42 | 5.48 | 5.53 | 0.001 |
CRP (mg/l) | 1.68 | 1.28 | 1.19 | 1.03 | <0.0001 |
IL-6 (pg/ml) | 2.80 | 2.32 | 2.23 | 2.08 | <0.0001 |
WCC (×109/l) | 7.32 | 7.03 | 6.89 | 6.62 | <0.0001 |
GGT (IU/l) | 15.0 | 15.2 | 14.6 | 14.9 | 0.77 |
Data are means unless otherwise stated. The numbers in bold indicate the range of FVC and FEV1-to-FVC ratio for the four quartiles. WCC, white cell count.
All measures of lung function (FEV1, FVC, and FEV1-to-FVC ratio) were strongly (inversely) associated with inflammation markers even after adjustment for age, BMI, evidence of CHD, smoking, physical activity, social class, alcohol intake, blood glucose, triglycerides, sBP, and HDL cholesterol. This was seen in both smokers and nonsmokers (all P < 0.0001). The adjusted standardized regression coefficients for CRP for FEV1, FVC, and FEV1-to-FVC ratio were −0.21, −0.27, and −0.14 for smokers and −0.17, −0.14, and −0.04 for nonsmokers, respectively. For IL-6, the corresponding standardized regression coefficients were −0.11, −0.12, and −0.06 in smokers and −0.08, −0.10, and −0.06 in nonsmokers, respectively.
FVC and FEV1 were significantly and inversely associated with risk of type 2 diabetes after adjustment for lifestyle characteristics, BMI, sBP, metabolic risk factors (blood glucose, triglycerides, HDL cholesterol), and GGT (Table 2). The inverse associations seen with FEV1 and FVC were attenuated after further adjustment for CRP and IL-6 (P = 0.14 and P = 0.11 for trend, respectively) although the increased risks in those with low FEV1 remained significantly increased when compared with high FEV1 (Table 2). No association was seen between the FEV1-to-FVC ratio and incident type 2 diabetes.
Table 2.
FVC (l) |
P-value trend | ||||
---|---|---|---|---|---|
1 (<3.97) | 2 (3.97–) | 3 (4.43–) | 4 (4.87–) | ||
n | 1,095 | 1,106 | 1,116 | 1,117 | |
Rate/1,000 P-years | 4.7 | 3.2 | 3.2 | 2.1 | |
Model 1 | 2.54 (1.74–3.70) | 1.58 (1.07–2.34) | 1.56 (1.06–2.31) | 1.00 | <0.0001 |
Model 2 | 1.94 (1.32–2.86) | 1.32 (0.89–1.97) | 1.34 (0.90–1.97) | 1.00 | 0.001 |
Model 3 | 1.59 (1.07–2.56) | 1.14 (0.76–1.71) | 1.27 (0.85–1.88) | 1.00 | 0.03 |
Model 4 | 1.46 (0.97–2.20) | 1.21 (0.81–1.81) | 1.03 (0.68–1.55) | 1.00 | 0.14 |
FEV1 (l) |
P-value trend | ||||
---|---|---|---|---|---|
1 (<2.95) | 2 (2.95–) | 3 (3.41–) | 4 (3.81–) | ||
n | 1,108 | 1,107 | 1,111 | 1,108 | |
Rate/1,000 P-years | 3.9 | 3.4 | 3.5 | 2.2 | |
Model 1 | 2.05 (1.40–3.01) | 1.51 (1.03–2.23) | 1.56 (1.07–2.26) | 1.00 | 0.005 |
Model 2 | 1.92 (1.29–2.86) | 1.35 (0.91–1.99) | 1.52 (1.04–2.21) | 1.00 | 0.02 |
Model 3 | 1.74 (1.16–2.61) | 1.36 (0.92–2.03) | 1.35 (0.92–1.98) | 1.00 | 0.04 |
Model 4 | 1.67 (1.09–2.56) | 1.31 (0.87–1.97) | 1.40 (0.94–2.08) | 1.00 | 0.11 |
FEV1-to-FVC ratio |
P-value trend | ||||
---|---|---|---|---|---|
1 (<0.71) | 2 (0.71–) | 3 (0.77–) | 4 (0.82–) | ||
Rate/1,000 P-years | 2.5 | 3.7 | 3.2 | 3.6 | |
Age-adjusted | 0.67 (0.46–0.96) | 0.87 (0.63–1.21) | 0.77 (0.55–1.07) | 1.00 | 0.12 |
Model 1, age-adjusted; Model 2, Model 1 + adjusted for age, BMI, smoking, physical activity, alcohol intake, social class, evidence of CHD (undiagnosed); Model 3, Model 2 + triglycerides, sBP, HDL cholesterol, blood glucose, and GGT; and Model 4, Model 3 + CRP + IL-6. P-years, person-years.
When examined separately by BMI status (<28 kg/m2 and ≥28 kg/m2), the increased risk of type 2 diabetes associated with low FVC or low FEV1 was more apparent in the <28 kg/m2 BMI group. The adjusted RRs for FVC (Quartile 1 vs. Quartile 4) were 1.73 (95% CI 1.03–2.93) and 1.14 (0.58–2.22) for the two BMI groups, respectively. The corresponding RRs for FEV1 were 2.05 (1.17–3.61) and 1.11 (0.56–2.20). However, a formal test for interaction between FEV1 or FVC and BMI with incident type 2 diabetes was not significant. Although the numbers were small, the increased risk associated with low FEV1—and to a lesser extent FVC—was also seen in “never smokers” (Quartile 1 vs. Quartile 4: adjusted RR 1.21 [0.53–2.75] and 1.54 [0.63–3.80] for FVC and FEV1, respectively). No interaction was seen with smoking status.
FVC and FEV1 were significantly associated with major CHD events largely due to the association with fatal CHD events (Table 3). No association was seen for nonfatal events after adjustment for established risk factors. Comparisons of differing association between fatal and nonfatal events showed significant/marginally significant differences in the relationships between FEV1 and FVC and fatal CHD and nonfatal MI (P = 0.04 and P = 0.08, respectively). The inverse association between FVC and FEV1 with fatal CHD events was attenuated but remained significant after adjustment for CRP and IL-6 (P = 0.03 and P = 0.01 for trend) although the increased risk associated with low FVC was no longer significant compared with those with high FVC (Table 3). Exclusion of the men who developed type 2 diabetes during the follow-up period or men with silent MI on ECG (85 men) made little difference to the findings. Although the numbers were small, the inverse association between FVC and FEV1 and fatal CHD events after adjustment for CRP and IL-6 was seen in “never smokers” (Quartile 1 vs. Quartile 4; adjusted RR 2.86 [95% CI 0.77–10.66] and 2.41 [0.85–6.85] for FVC and FEV1, respectively). Case fatality decreased with increasing lung function after adjustment for age, and the trend became of marginal significance after adjustment for CRP and preexisting disease. The FEV1-to-FVC ratio was significantly, but weakly, inversely associated with fatal CHD events after adjustment for age (P = 0.02 for trend), but this was attenuated after adjustment for established CV risk factors (adjusted RR for the four quartiles 1.10 [0.76–1.60], 1.13 [0.79–1.63], 0.96 [0.66–1.40] and 1.00).
Table 3.
FVC (l) |
P for trend | ||||
---|---|---|---|---|---|
1 (<3.97) | 2 (3.97–) | 3 (4.43–) | 4 (4.87–) | ||
n | 1,095 | 1,106 | 1,116 | 1,117 | |
Major CHD | |||||
Rate/1,000 P-years | 11.9 (225) | 9.9 (189) | 7.9 (157) | 6.1 (123) | |
Model 1 | 1.56 (1.24–1.97) | 1.39 (1.11–1.76) | 1.19 (0.94–1.51) | 1.00 | <0.0001 |
Model 2 | 1.25 (0.98–1.59) | 1.26 (1.00–1.60) | 1.09 (0.86–1.39) | 1.00 | 0.007 |
Model 3 | 1.14 (0.88–1.47) | 1.23 (0.96–1.37) | 1.14 (0.88–1.47) | 1.00 | 0.16 |
Fatal CHD | |||||
Rate/1,000 P-years | 5.4 (102) | 4.3 (82) | 3.0 (59) | 1.9 (39) | |
Model 1 | 1.91 (1.30–2.81) | 1.70 (1.16–2.51) | 1.32 (0.88–1.98) | 1.00 | <0.0001 |
Model 2 | 1.48 (1.00–2.21) | 1.50 (1.01–2.22) | 1.15 (0.76–1.74) | 1.00 | 0.002 |
Model 3 | 1.41 (0.92–2.16) | 1.52 (1.00–2.32) | 1.28 (0.82–1.98) | 1.00 | 0.03 |
Nonfatal MI | |||||
Rate/1,000 P-years | 6.5 (123) | 5.6 (107) | 4.9 (98) | 4.2 (84) | |
Age-adjusted RR | 1.39 (1.04–1.86) | 1.25 (0.93–1.66) | 1.14 (0.85–1.53) | 1.00 | 0.04 |
Risk factor adjusted # | 1.12 (0.82–1.51) | 1.15 (0.86–1.53) | 1.07 (0.79–1.44) | 1.00 | 0.38 |
Case fatality (%) | 49.1 | 41.1 | 36.7 | 28.7 | |
Age-adjusted OR | 1.40 (0.86–2.27) | 1.40 (0.86–2.29) | 1.19 (0.72–1.98) | 1.00 | P = 0.03 |
+ CRP + preexisting CHD | 1.30 (0.80–2.14) | 1.36 (0.83–2.22) | 1.18 (0.71–1.96) | 1.00 | P = 0.07 |
FEV1 (l) |
|||||
---|---|---|---|---|---|
1 (<2.95) | 2 (2.95–) | 3 (3.41–) | 4 (3.81–) | ||
n | 1,108 | 1,107 | 1,111 | 1,108 | |
Major CHD | |||||
Rate/1,000 P-years | 12.1 (228) | 10.0 (190) | 8.2 (161) | 5.6 (115) | |
Model 1 | 1.66 (1.31–2.11) | 1.44 (1.14–1.83) | 1.32 (1.03–1.67) | 1.00 | <0.0001 |
Model 2 | 1.28 (1.00–1.65) | 1.18 (0.93–1.51) | 1.17 (0.92–1.49) | 1.00 | 0.002 |
Model 3 | 1.15 (0.88–1.50) | 1.12 (0.87–1.43) | 1.20 (0.93–1.55) | 1.00 | 0.12 |
Fatal CHD | |||||
Rate/1,000 P-years | 6.0 (112) | 4.1 (78) | 3.0 (59) | 1.6 (33) | |
Model 1 | 2.41 (1.60–3.63) | 1.80 (1.19–2.73) | 1.57 (1.02–2.40) | 1.00 | <0.0001 |
Model 2 | 1.81 (1.19–2.76) | 1.46 (0.96–2.23) | 1.37 (0.89–2.11) | 1.00 | 0.0003 |
Model 3 | 1.63 (1.03–2.67) | 1.46 (0.93–2.29) | 1.42 (0.90–2.25) | 1.00 | 0.01 |
Nonfatal MI | |||||
Rate/1,000 P-years | 6.1 (116) | 5.9 (112) | 5.2 (102) | 4.0 (82) | |
Age-adjusted RR | 1.32 (0.97–1.79) | 1.30 (0.97–1.73) | 1.22 (0.91–1.63) | 1.00 | 0.02 |
Risk factor adjusted # | 1.02 (0.74–1.41) | 1.04 (0.77–1.40) | 1.09 (0.81–1.47) | 1.00 | 0.35 |
Case fatality (%) | 45.3 | 43.4 | 37.6 | 31.7 | |
Age-adjusted OR | 1.77 (1.06–2.95) | 1.36 (0.81–2.27) | 1.26 (0.74–2.13) | 1.00 | 0.03 |
+ CRP + preexisting CHD | 1.65 (0.97–2.79) | 1.31 (0.78–2.21) | 1.22 (0.72–2.07) | 1.00 | 0.06 |
Model 1, age adjusted; Model 2, Model 1 + adjusted for BMI, smoking, physical activity, alcohol intake, social class, preexisting evidence of CHD (undiagnosed), sBP, cholesterol, and HDL cholesterol; and Model 3, Model 2 + CRP + IL-6. OR, odds ratio.
The percent predicted FVC (FEV1) was highly correlated with height standardized FVC (FEV1) (r = 0.97), and models using continuous terms for percent predicted FVC and percent predicted FEV1 were consistent with the above findings using FVC/height2 and FEV1/height2 (data not shown).
CONCLUSIONS
In this study of British middle-aged men without diagnosis of CHD or type 2 diabetes, lung function as measured by FVC and FEV1 was significantly and inversely associated with risk of incident type 2 diabetes and fatal CHD after adjustment for a wide range of variables including smoking, physical activity, alcohol intake, social class, BMI, preexisting CHD (undiagnosed), blood pressure, total cholesterol, metabolic markers (blood glucose, triglycerides, HDL cholesterol), and hepatic function. Interestingly, inflammatory pathway (i.e., CRP, IL-6) adjustments further attenuated associations between FEV1 and FVC and type 2 diabetes and fatal CHD. Obstructive airways disease per se (as much as it may be reflected by the first quartile of the FEV1-to-FVC ratio [<0.71]) was related neither to type 2 diabetes nor independently with CHD events. Our findings confirm previous studies that have reported inverse associations between lung function and type 2 diabetes and CHD events (1,2,5–8, 22) and extend these findings further by examining the role of inflammation in explaining these associations and differentiating between fatal and nonfatal CHD events.
Lung function and type 2 diabetes
Our results are consistent with other prospective studies that have shown restrictive lung function (reduced FVC and FEV1) but not obstructive pulmonary function (FEV1-to-FVC ratio) is associated with the development of type 2 diabetes (5–8). Numerous cross-sectional studies have also suggested a restrictive pattern of deterioration in lung function in diabetics (3). Several explanations for the link between lung function and incident type 2 diabetes have been proposed including metabolic pathways, inflammatory processes, and early life influences (6,7). FVC in particular and FEV1 were associated with metabolic abnormalities and components of the insulin resistance syndrome, which is consistent with several previous studies that have reported associations between restrictive lung patterns with glucose metabolism and the metabolic syndrome (23). We did not have measures of fasting insulin but simultaneous adjustment for metabolic risk markers closely associated with insulin resistance (BMI, glucose, triglycerides, and HDL cholesterol) and GGT (a correlate of hepatic fat and insulin resistance) (16) did not fully explain the association. We observed a strong association between both restrictive and obstructive lung patterns and inflammatory markers CRP and IL-6. These factors contributed to the association between reduced lung function—in particular FVC—and type 2 diabetes, although the increased risk associated with low FEV1 compared with high FEV1 remained significant. Low birth weight has been linked to both type 2 diabetes and adult lung function (24). Thus, reduced lung function (FEV1) and risk of type 2 diabetes may also be partially determined by early growth.
Lung function and CHD
Although numerous studies have reported an inverse relationship between lung function (FEV1 or FVC) and risk of CHD, the independence of the association has been inconsistent (1,2,22); few studies have differentiated between fatal and nonfatal episodes. Our findings, based on a larger number of events, confirmed previous findings that reduced lung function is associated more closely with increased risk of fatal CHD than with nonfatal episodes (10) and predicts increased fatality in the event of a cardiac event (10). The association with fatal events was independent of cardiovascular risk factors and was present in “never smokers.” In most studies, fatal and nonfatal events are combined due to perceptions concerning common etiologies or limited power; this may contribute to the inconsistent findings between studies. The reason for increased incidence of fatal events with poor lung function, in particular FEV1, appears only partially explained by inflammation (CRP, IL-6). The increased risk of fatal CHD associated with low lung function may potentially be due to ventricular dysrhythmias or silent MIs, which are associated with poor lung function and increased fatality (25), or due to the increased risk of developing diabetes, which also increases case fatality (25). However, exclusion of the 85 men with silent MIs on ECGs or incident diabetics in the study and further adjustment for arrthymias made little difference to the findings.
Implications for diabetes and excess risk of fatal CHD
The finding from this and other studies suggest that reduced lung function not only preceded the onset of diabetes but also continued at an accelerated pace after the onset of diabetes (4). The increased risk of fatal events and increased case fatality associated with poor lung function suggest that reduced lung function may be another potential factor linking diabetes to increased risk of CHD and increased susceptibility to a fatal episode in the event of a cardiac event (15,25). Speculatively, this association may reflect an impaired respiratory buffering capacity needed in the context of a cardiac event to cope with reduced cardiac output. Further studies are warranted to see if physical training might improve cardiorespiratory function in type 2 diabetics and thus reduce risk of fatal CHD.
Strengths and limitations
The strengths and limitations of the present study require careful consideration. The study is restricted to almost exclusively white European men. However, the study population is socioeconomically representative of middle-aged men in the U.K. and follow-up rates are exceptionally high. Ascertainment of CHD death and MI is based on standard methods. We were able to take into account evidence of CHD (diagnosed and undiagnosed) with factors including ECG evidence of silent MIs and arrhythmias at baseline and a wide range of cardiovascular risk factors. Despite adjusting risk estimates for several potential confounders and mediators, we cannot rule out residual confounding. Although the associations persisted after adjustment for established risk factors, it is possible that this may be due to imprecision due to adjustment for baseline levels only. The current findings based on observational epidemiological data cannot assess causality.
CONCLUSIONS
In this study of middle-aged men free of diagnosed CHD stroke and diabetes, restrictive patterns of lung function (reduced FVC and FEV1) but not obstructive respiratory patterns (reduced FEV1-to-FVC ratio) were inversely associated with incident type 2 diabetes and fatal CHD events independent of established risk factors and metabolic risk factors. The association between reduced lung function and fatal CHD and type 2 diabetes in particular was to some extent associated with inflammatory pathways. Further studies are now needed to extend such novel observations. The association between reduced lung function and development of type 2 diabetes and fatal CHD events may provide another possible explanation for the increased risk of fatal CHD in individuals with type 2 diabetes.
Acknowledgments
The British Regional Heart Study is supported by program and project grants from the British Heart Foundation (RG/08/013/25942 and PG/08/098/26122).
No potential conflicts of interest relevant to this article were reported.
S.G.W. wrote the manuscript. S.G.W., A.G.S., N.S., and P.H.W. contributed to the discussion and reviewed and edited the manuscript. A.R. and G.D.L. researched the data and reviewed and edited the manuscript. M.C.T. researched the data.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
References
- 1.Schroeder EB, Welch VL, Couper D, Nieto FJ, Liao D, Rosamond WD, Heiss G: Lung function and incident coronary heart disease: the Atherosclerosis Risk in Communities Study. Am J Epidemiol 2003;158:1171–1181 [DOI] [PubMed] [Google Scholar]
- 2.Sin DD, Wu L, Man SF: The relationship between reduced lung function and cardiovascular mortality: a population-based study and a systematic review of the literature. Chest 2005;127:1952–1959 [DOI] [PubMed] [Google Scholar]
- 3.van den Borst B, Gosker HR, Zeegers MP, Schols AMWL: Pulmonary function in diabetes: a meta-analysis. Chest. 26March2010. [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
- 4.Yeh HC, Punjabi NM, Wang NY, Pankow JS, Duncan BB, Cox CE, Selvin E, Brancati FL: Cross-sectional and prospective study of lung function in adults with type 2 diabetes: the Atherosclerosis Risk in Communities (ARIC) study. Diabetes Care 2008;31:741–746 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Engström G, Janzon L: Risk of developing diabetes is inversely related to lung function: a population-based cohort study. Diabet Med 2002;19:167–170 [DOI] [PubMed] [Google Scholar]
- 6.Ford ES, Mannino DM: National Health and Nutrition Examination Survey Epidemiologic Follow-up Study Prospective association between lung function and the incidence of diabetes: findings from the National Health and Nutrition Examination Survey Epidemiologic Follow-Up Study. Diabetes Care 2004;27:2966–2970 [DOI] [PubMed] [Google Scholar]
- 7.Yeh HC, Punjabi NM, Wang NY, Pankow JS, Duncan BB, Brancati FL: Vital capacity as a predictor of incident type 2 diabetes: the Atherosclerosis Risk in Communities study. Diabetes Care 2005;28:1472–1479 [DOI] [PubMed] [Google Scholar]
- 8.Engström G, Hedblad B, Nilsson P, Wollmer P, Berglund G, Janzon L: Lung function, insulin resistance and incidence of cardiovascular disease: a longitudinal cohort study. J Intern Med 2003;253:574–581 [DOI] [PubMed] [Google Scholar]
- 9.Engström G, Lind P, Hedblad B, Wollmer P, Stavenow L, Janzon L, Lindgärde F: Lung function and cardiovascular risk: relationship with inflammation-sensitive plasma proteins. Circulation 2002;106:2555–2560 [DOI] [PubMed] [Google Scholar]
- 10.Engström G, Hedblad B, Janzon L: Reduced lung function predicts increased fatality in future cardiac events: a population-based study. J Intern Med 2006;260:560–567 [DOI] [PubMed] [Google Scholar]
- 11.Gan WQ, Man SF, Senthilselvan A, Sin DD: Association between chronic obstructive pulmonary disease and systemic inflammation: a systematic review and a meta-analysis. Thorax 2004:59:574–580 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Engström G, Hedblad B, Stavenow L, Tydén P, Lind P, Janzon L, Lindgärde F: Fatality of future coronary events is related to inflammation-sensitive plasma proteins: a population-based prospective cohort study. Circulation 2004;110:27–31 [DOI] [PubMed] [Google Scholar]
- 13.Sattar N, Murray HM, Welsh P, Blauw GJ, Buckley BM, Cobbe S, de Craen AJ, Lowe GD, Jukema JW, Macfarlane PW, Murphy MB, Stott DJ, Westendorp RG, Shepherd J, Ford I, Packard CJ: Prospective Study of Pravastatin in the Elderly at Risk (PROSPER) Study Group. Are markers of inflammation more strongly associated with risk for fatal than for nonfatal vascular events? PLoS Med. 23June2009. [Epub ahead of print] [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Shaper AG, Pocock SJ, Walker M, Cohen NM, Wale CJ, Thomson AG: British Regional Heart Study: cardiovascular risk factors in middle-aged men in 24 towns. BMJ 1981;283:179–186 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Wannamethee G, Whincup PH, Shaper AG, Walker M, MacFarlane PW: Factors determining case fatality in myocardial infarction “who dies in a heart attack.” Br Heart J 1995;74:324–331 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Wannamethee SG, Shaper AG, Lennon L, Whincup PH: Hepatic enzymes, the metabolic syndrome and the risk of type 2 diabetes in older men. Diabetes Care 2005;28:2913–2918 [DOI] [PubMed] [Google Scholar]
- 17.Danesh J, Whincup P, Walker M, Lennon L, Thomson A, Appleby P, Gallimore JR, Pepys MB: Low grade inflammation and coronary heart disease: prospective study and update meta-analyses. BMJ 2000;321:199–204 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Cole TJ: Linear and proportional regression models in the prediction of ventilatory function: with discussion. J R Stat Soc Ser A 1975;138:297–338 [Google Scholar]
- 19.Cook DG, Shaper AG: Breathlessness, lung function and the risk of heart attack. Eur Heart J 1988;9:1215–1222 [DOI] [PubMed] [Google Scholar]
- 20.Quanjer PH, Tammeling GJ, Cotes JE, Pedersen OF, Peslin R, Yernault JC: Lung volumes and forced ventilatory flows. Report working party standardization of lung function tests, European Community for Steel and Coal. Official Statement of the European Respiratory Society. Eur Respir J 1993;6(Suppl. 16):5–40 [PubMed] [Google Scholar]
- 21.Glynn RJ, Rosner B: Comparison of risk factors for the competing risks of coronary heart disease, stroke, and venous thromboembolism. Am J Epidemiol 2005;162:975–982 [DOI] [PubMed] [Google Scholar]
- 22.Marcus EB, Curb JD, MacLean CJ, Reed DM, Yano K: Pulmonary function as a predictor of coronary heart disease. Am J Epidemiol 1989;129:97–104 [DOI] [PubMed] [Google Scholar]
- 23.Fimognari FL, Pasqualetti P, Moro L, Franco A, Piccirillo G, Pastorelli R, Rossini PM, Incalzi RA: The association between metabolic syndrome and restrictive ventilatory dysfunction in older persons. J Gerontol A Biol Sci Med Sci 2007;62:760–765 [DOI] [PubMed] [Google Scholar]
- 24.Lawlor DA, Ebrahim S, Smith GD: Associations of measures of lung function with insulin resistance and type 2 diabetes: findings from the British Women's Heart and Health Study. Diabetologia 2004;47:195–203 [DOI] [PubMed] [Google Scholar]
- 25.Pearte CA, Furberg CD, O'Meara ES, Psaty BM, Kuller L, Powe NR, Manolio T: Characteristics and baseline clinical predictors of future fatal versus nonfatal coronary heart disease events in older adults: the Cardiovascular Health Study. Circulation 2006;113:2177–2185 [DOI] [PubMed] [Google Scholar]