The mean age of the 7,536 participants was 51.7 years, 49.2% were men, and 41.7% were current or past smoker. The mean ± SD of BMI, systolic and diastolic BP, waist circumference, HDL cholesterol level, eGFR, hemoglobin level, and WBC count were 24.6 ± 3.1 kg/m2, 120.9 ± 18.0 mmHg, 80.2 ± 11.3 mmHg, 82.7 ± 8.7 cm, 44.6 ± 9.9 mL/dL, 94.1 ± 13.3 mL/min/1.73 m2, 13.6 ± 1.6 g/dL, and 6.5 ± 1.8 × 103/µL, respectively. Moreover, the median (IQR) of fasting glucose, HbA1c, triglyceride levels, and CRP levels were 82 (77–91) mg/dL, 5.6 (5.3–5.9) %, 135 (100–189) mg/dL, and 0.14 (0.07–0.24) mg/dL, respectively. At baseline, the mean ± SD values of FEV1, FVC, and FEV1/FVC ratio were 97.0 ± 13.8% predicted, 96.9 ± 13.0% predicted, and 80.1 ± 7.4%, respectively. The mean ± SD values of annual change in FEV1, FVC, and FEV1/FVC ratio were − 0.55 ± 2.54% predicted/year, -0.56 ± 2.65% predicted/year, and − 0.37 ± 1.44%/year, respectively.
The clinical characteristics of the study population according to HOMA-IR quartile are shown in Table 1. Although age was not associated with HOMA-IR quartile, increased HOMA-IR quartile was associated with decreased proportion of men. Although income and educational status were not associated with HOMA-IR quartile, increased HOMA-IR quartile was associated with a decreased rate of current or past smoking. With the increase of HOMA-IR quartile, BMI, systolic and diastolic BP, waist circumference, WBC count, and fasting glucose, HbA1c, triglyceride, hemoglobin, and CRP levels increased, whereas HDL cholesterol levels decreased. With the increase of HOMA-IR quartile, baseline percent-predicted FEV1 and percent-predicted FVC decreased, but FEV1/FVC increased.
Table 1
Clinical characteristics of the study population according to HOMA-IR quartile.
|
HOMA-IR groups (n = 7536)
|
|
|
1Q: < 1.05 (n = 1888)
|
2Q: 1.06–1.45 (n = 1887)
|
3 Q: 1.45–2.03 (n = 1882)
|
4 Q: ≥2.03 (n = 1879)
|
P-trend
|
Age(years)
|
52.09 ± 8.86
|
51.56 ± 8.73
|
51.43 ± 8.60
|
51.56 ± 8.63
|
0.099
|
Male, n (%)
|
1043 (55.24%)
|
930 (49.28%)*
|
852 (45.27%)*†
|
884 (47.05%)*
|
< 0.001
|
High income, n (%)
|
337 (17.85%)
|
343 (18.18%)
|
335 (17.80%)
|
376 (20.01%)
|
0.127
|
College graduate, n (%)
|
261 (13.82%)
|
263 (13.94%)
|
279 (14.82%)
|
283 (15.06%)
|
0.204
|
Current or past smoker, n (%)
|
865 (45.82%)
|
781 (41.39%)*
|
723 (38.42%)*
|
775 (41.25%)*
|
0.001
|
BMI (Kg/m2)
|
23.57 ± 2.98
|
23.96 ± 2.84*
|
24.84 ± 2.79*†
|
26.16 ± 3.08*†‡
|
< 0.001
|
Systolic BP (mmHg)
|
118.13 ± 17.31
|
118.91 ± 17.05
|
121.01 ± 17.96*†
|
125.50 ± 18.92*†‡
|
< 0.001
|
Diastolic BP (mmHg)
|
78.38 ± 11.14
|
79.00 ± 11.01
|
80.40 ± 11.14*†
|
83.10 ± 11.44*†‡
|
< 0.001
|
Waist circumference (cm)
|
79.92 ± 8.41
|
80.84 ± 8.04*
|
83.07 ± 8.15*†
|
86.89 ± 8.33*†‡
|
< 0.001
|
Fasting glucose (mg/dL)
|
79 (75–85)
|
82 (77–87)*
|
83 (78–90)*†
|
90 (82–101)*†‡
|
< 0.001
|
HbA1c (%)
|
5.5 (5.3–5.7)
|
5.5 (5.3–5.8)
|
5.6 (5.3–5.8)* †
|
5.7 (5.5–6.3) *†‡
|
< 0.001
|
Triglyceride (mg/dL)
|
122 (93–169)
|
124 (95–171)
|
136 (101–188)*†
|
162 (118–231)*†‡
|
< 0.001
|
HDL cholesterol (mg/dL)
|
46.06 ± 10.17
|
45.86 ± 10.13
|
44.10 ± 9.61*†
|
42.30 ± 9.00*†‡
|
< 0.001
|
eGFR (mL/min/1.73 m2)
|
94.06 ± 12.85
|
94.49 ± 13.26
|
94.17 ± 13.36
|
93.65 ± 13.63
|
0.278
|
Hemoglobin (g/dL)
|
13.68 ± 1.53
|
13.55 ± 1.54
|
13.55 ± 1.64
|
13.78 ± 1.55†‡
|
< 0.001
|
WBC count (× 103/µL)
|
6.43 ± 1.80
|
6.38 ± 1.71
|
6.46 ± 1.77
|
6.79 ± 1.81*†‡
|
< 0.001
|
CRP level (mg/dL)
|
0.12 (0.05–0.22)
|
0.13 (0.06–0.23)
|
0.14 (0.07–0.24)*†
|
0.17 (0.09–0.27)*†‡
|
< 0.001
|
Baseline FEV1 (L)
|
2.96 ± 0.68
|
2.96 ± 0.69
|
2.92 ± 0.70
|
2.91 ± 0.69
|
0.024
|
Baseline FVC (L)
|
3.74 ± 0.84
|
3.72 ± 0.88
|
3.66 ± 0.88*
|
3.63 ± 0.87*†
|
< 0.001
|
Baseline FEV1 (%-predicted)
|
96.82 ± 14.01
|
97.82 ± 13.84
|
97.38 ± 13.78
|
96.02 ± 13.42†‡
|
< 0.001
|
Baseline FVC (%-predicted)
|
97.45 ± 12.76
|
98.25 ± 13.25
|
97.08 ± 13.16†
|
94.61 ± 12.45*†‡
|
< 0.001
|
Baseline FEV1/FVC (%)
|
79.60 ± 7.91
|
79.97 ± 7.42
|
80.29 ± 7.26*
|
80.48 ± 6.89*
|
0.001
|
Values are expressed as mean ± standard deviation for normally distributed continuous variables, median and interquartile range for non-normally distributed variables and percentage for categorical variables. P-trend was analyzed normally distributed continuous variables by the linear regression, for non-normally distributed continuous variable by Jonckheere-Terpstra tests, and for categorical variables by Cochran-Armitage test for trend. *, †, and ‡ meant P < 0.05 when compared to < 1.05, 1.06–1.45, 1.45–2.03 groups of HOMA-IR, respectively, using Bonferroni post-hoc analysis of one-way ANOVA for normally distributed continuous variables, Mann-Whitney U tests for non-normally distributed continuous variable, and chi-square tests for categorical variables. Abbreviations: HOMA-IR, homeostatic model assessment for insulin resistance; BMI, body mass index; BP, blood pressure; HbA1c, hemoglobin A1c; HDL, high density lipoprotein; eGFR, estimated glomerular filtration rate; WBC, white blood cell; CRP, C-reactive protein; FEV1, forced expiratory volume in 1 second; FVC, functional vital capacity. |
We explored the association between HOMA-IR and annual changes in lung function test (Fig. 2). As HOMA-IR increased, annual FEV1 and FVC change decreased, whereas annual FEV1/FVC ratio change increased. Although the negative association between HOMA-IR and annual changes in FEV1 and FVC was statistically significant, the association between HOMA-IR and FEV1/FVC ratio was nullified in the multivariate linear regression analysis (Table 2). In the multivariate GAM plot analysis, we identified that the association between HOMA-IR and annual change in FEV1 and FVC was L-shaped, and the lowest AICs were found in HOMA-IR 2.4, in both annual FEV1 and FVC (Fig. 3).
Table 2
Association between log-transformed HOMA-IR and the annual lung function change.
|
Per 1 unit increase of Log-HOMA-IR
|
|
|
Univariate
|
Model 1
|
Model 2
|
Outcomes
|
Beta (95% CI)
|
P
|
Beta (95% CI)
|
P
|
Beta (95% CI)
|
P
|
∆ FEV1 (%-predicted/year)
|
-0.15 (-0.24 to -0.06)
|
< 0.001
|
-0.15 (-0.24 to -0.07)
|
< 0.001
|
-0.23 (-0.32 to -0.13)
|
< 0.001
|
∆ FVC (%-predicted/year)
|
-0.20 (-0.29 to -0.11)
|
< 0.001
|
-0.20 (-0.29 to -0.11)
|
< 0.001
|
-0.29 (-0.39 to -0.19)
|
< 0.001
|
∆ FEV1/FVC (%/year)
|
0.07 (0.02 to 0.12)
|
0.004
|
0.06 (0.02 to 0.11)
|
0.010
|
0.05 (0.00 to 0.11)
|
0.051
|
In model 1, age and sex were added as covariates. Model 2 included high income, college graduate, smoking status, BMI, waist circumference, systolic and diastolic BP, triglyceride, HDL cholesterol, eGFR, WBC count, hemoglobin and CRP levels in addition to the covariates in the model 1. Abbreviations: HOMA-IR, homeostatic model assessment for insulin resistance; CI, confidence interval; FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity. |
There were no subgroups showing the effect modification between HOMA-IR and annual FVC change. However, participant age significantly modified the effect of HOMA-IR on annual FEV1 change (P for interaction < 0.001). Although HOMA-IR was not associated with annual FEV1 change in people aged < 50 years, increased HOMA-IR was significantly associated with decreased annual FEV1 change in individuals aged ≥ 50 years (Table 3). We also identified similar effect modification of age on the association between HOMA-IR and annual FEV1 change in the LOWESS regression plot with interaction (Fig. 4).
Table 3
Subgroup analysis for the effects of HOMA-IR on the annual rates of changes in FEV1 and FVC.
|
Per 1 unit increase of Log-HOMA-IR
|
|
|
|
Adjusted beta (95% CI, P)
|
P for interaction
|
Subgroup
|
No. of people
|
∆ FEV1 (%-predicted/year)
|
∆ FVC (%-predicted/year)
|
∆ FEV1
|
∆ FVC
|
Age (years)
|
< 50 (n = 3,755)
|
-0.03 (-0.16 to 0.09, 0.597)
|
-0.14 (-0.27 to 0.02, 0.026)
|
< 0.001
|
0.144
|
|
≥ 50 (n = 3,781)
|
-0.37 (-0.51 to -0.23, < 0.001)
|
-0.38 (-0.53 to -0.24, < 0.001)
|
|
|
Sex
|
Male (n = 3,709)
|
-0.25 (-0.37 to -0.12, < 0.001)
|
-0.26 (-0.39 to -0.12, < 0.001)
|
0.585
|
0.306
|
|
Female (n = 3,827)
|
-0.23 (-0.37 to -0.09, 0.001)
|
-0.33 (-0.47 to -0.19, < 0.001)
|
|
|
Current or past smoker
|
No (n = 4,392)
|
-0.20 (-0.32 to -0.07, 0.002)
|
-0.31 (-0.44 to -0.18, < 0.001)
|
0.912
|
0.309
|
|
Yes (n = 3,144)
|
-0.29 (-0.44 to -0.15, < 0.001)
|
-0.27 (-0.42 to -0.12, < 0.001)
|
|
|
BMI (kg/m2)
|
< 25 (n = 4,277)
|
-0.25 (-0.38 to -0.13, < 0.001)
|
-0.30 (-0.43 to -0.17, < 0.001)
|
0.289
|
0.704
|
|
≥ 25 (n = 3,259)
|
-0.20 (-0.34 to -0.05, 0.008)
|
-0.27 (-0.42 to -0.12, < 0.001)
|
|
|
WBC count (x103/µL)
|
< 6.3 (n = 3,693)
|
-0.19 (-0.34 to -0.04, 0.011)
|
-0.29 (-0.43 to -0.14, < 0.001)
|
0.683
|
0.772
|
|
≥ 6.3 (n = 3,843)
|
-0.26 (-0.39 to -0.14, < 0.001)
|
-0.29 (-0.43 to -0.16, < 0.001)
|
|
|
Hemoglobin (mg/dL)
|
< 13.6 (n = 3,668)
|
-0.23 (-0.38 to -0.08, 0.002)
|
-0.31 (-0.47 to -0.16, < 0.001)
|
0.442
|
0.389
|
|
≥ 13.6 (n = 3,449)
|
-0.23 (-0.35 to -0.11, < 0.001)
|
-0.27 (-0.40 to -0.15, < 0.001)
|
|
|
Adjusted beta and 95% CIs were analyzed using the multivariate linear regression. All covariates of model 2 shown in Table 2 were included for adjustment. The variable used to divide subgroup was excluded. Abbreviations: HOMA-IR, homeostatic model assessment for insulin resistance; CI, confidence interval; BMI, body mass index; WBC, white blood cell; FEV1, forced expiratory volume in 1 second; FVC, functional vital capacity. |