Table 4.

Logistic regression models comparing current smokers with noncurrent smokers by the best 10% and the worst 10% exercise performances, respectively.

	OR	95% CI	p value
Top 10% performance levels
Model 1
2 min push-up ≥60 times	0.590	0.464 to 0.751	<0.0001
2 min sit-up ≥58 times	0.582	0.450 to 0.752	<0.0001
3000 m run ≤783 sec	1.024	0.853 to 1.229	0.80
Model 2
2 min push-up ≥60 times	0.578	0.454 to 0.737	<0.0001
2 min sit-up ≥58 times	0.581	0.449 to 0.752	<0.0001
3000 m run ≤783 sec	1.030	0.857 to 1.237	0.76
Bottom 10% performance levels
Model 1
2 min push-up ≤37 times	1.107	0.878 to 0.878	0.39
2 min sit-up ≤40 times	1.392	1.093 to 1.774	0.0074
3000 m run ≥934 sec	1.386	1.084 to 1.772	0.0092
Model 2
2 min push-up ≤37 times	1.122	0.885 to 1.423	0.34
2 min sit-up ≤40 times	1.420	1.110 to 1.817	0.0053
3000 m run ≥934 sec	1.428	1.110 to 1.837	0.0056

OR, odds ratio; CI, confidence intervals. Multiple logistic regression analysis was used to determine the smoking association with the best 10% and the worst 10% exercise performance. Model 1 was adjusted for age, heart rate, specialty, and alcohol drinking. Model 2 was adjusted for age, body mass index, heart rate, service specialty, hypertension, alcohol drinking, and exercise frequency.