Table 3. Emphysema-like lungand all-cause mortality rates by age group, race/ethnicity, BMI, sex, CT scanner type, and site, in persons without airflow obstruction over six years of follow-up.
| Strata | Deaths | At risk | Person years | Mortality rate | Emphysema-like lung | |
|---|---|---|---|---|---|---|
| Fully-adjusted HR per ½ IQR increase (95% CI) | P-value for interaction | |||||
| Age Group | ||||||
| First tertile (44-56) | 17 | 935 | 5927 | 2.87 | 1.11 (0.89, 1.37) | 0.183 |
| Second tertile (57-67) | 44 | 992 | 6215 | 7.08 | 1.25 (1.10, 1.41) | |
| Third tertile (68-84) | 125 | 1038 | 6424 | 19.46 | 1.10 (0.98, 1.22) | |
| Race/ethnicity | ||||||
| White | 52 | 963 | 5890 | 8.83 | 1.16 (1.03, 1.31) | 0.28 |
| Black | 63 | 759 | 4711 | 13.37 | 1.13 (1.01, 1.26) | |
| Hispanic | 45 | 745 | 4721 | 9.53 | 0.97 (0.76, 1.24) | |
| Chinese | 26 | 498 | 3244 | 8.01 | 1.33 (1.00, 1.73) | |
| BMI Group | ||||||
| <25 (Normal or Underweight) | 51 | 832 | 5224 | 9.76 | 1.24 (1.08, 1.41) | 0.27 |
| 25-30 (Overweight) | 63 | 1184 | 7441 | 8.47 | 1.11 (0.98.1.26) | |
| 30+ (Obese) | 72 | 941 | 5852 | 12.30 | 1.09 (0.96, 1.24) | |
| Sex | ||||||
| Female | 86 | 1609 | 10102 | 8.51 | 0.95 (0.75, 1.21) | 0.108 |
| Male | 100 | 1356 | 8465 | 11.81 | 1.16 (1.06, 1.26) | |
| CT scanner type | ||||||
| Electron Beam | 111 | 1782 | 11235 | 9.88 | 1.18 (1.06, 1.31) | 0.26 |
| Multi-detector | 75 | 1183 | 7332 | 10.23 | 1.09 (0.96, 1.23) | |
| Site | ||||||
| Forsyth County, North Carolina | 23 | 408 | 2584 | 8.90 | 1.10 (1.00, 1.28) | 0.39 |
| Northern Manhattan & Bronx, New York | 41 | 549 | 3489 | 11.75 | 1.24 (1.00, 1.45) | |
| Baltimore, Maryland | 36 | 344 | 1999 | 18.01 | 1.17 (1.00, 1.41) | |
| St Paul, Minnesota | 16 | 431 | 2749 | 5.82 | 0.76 (0.40, 1.30) | |
| Chicago, Illinois | 27 | 556 | 3504 | 7.71 | 1.09 (1.00, 1.28) | |
| Los Angeles, California | 43 | 677 | 4267 | 10.08 | 1.21 (1.00, 1.43) | |
Abbreviations: BMI = body mass index. HR = hazard ratio. IQR = Interquartile range. CI = Confidence Interval. CT=computed tomography. Mortality rate is reported per 1,000 person-years of observation. Numbers may not sum to totals due to rounding.
Emphysema-like lung is parameterized as emphysematous voxels. BMI values are expressed in kilograms/meters-squared. One interquartile range is equivalent to 91809 voxels. Airflow obstruction is defined as an FEV1/FVC ratio < 0.70 on spirometry. With the exception of the FEV1, covariates were measured at baseline.
Models are adjusted for a multiplicative interaction term (emphysematous voxels*stratum), total imaged lung volume, scanner type and milliamperes, site, age, gender, race/ethnicity, education, height, weight, coronary artery calcium, cardiac risk factors, creatinine, cancer history, alcohol use, forced expiratory volume in one second, and high attenuation areas.