Table 2.
Beta coefficients and 95% confidence intervals (CI) for multivariable linear regression models with CCA diameter as the outcome after adjustment for basic covariates and after adding atherosclerotic risk factors.* Both models included interaction terms allowing determination of effects of age in subsets of varying atherosclerosis severity†, the Atherosclerosis Risk in Communities (ARIC) Limited Data Set, 1987–1989
Basic Model N = 8163 | Full Model N = 8163 | |||
Beta (95% CI) | p-value | Beta (95% CI) | p-value | |
Main effects | ||||
Male gender | 0.492 (0.439, 0.545)‡ | <0.0001 | 0.435 (0.384, 0.486)‡ | <0.0001 |
Black race/ethnicity | 0.190 (0.149, 0.231)‡ | <0.0001 | 0.087 (0.044, 0.130)‡ | <0.0001 |
Height (cm) | 0.020 (0.018, 0.022)‡ | <0.0001 | 0.023 (0.021, 0.025)‡ | <0.0001 |
Current smoking | - | 0.154 (0.107, 0.201)‡ | <0.0001 | |
Current drinking | - | -0.038 (-0.075, -0.001)‡ | 0.0485 | |
Cigarette years of smoking (in 100's) | - | 0.019 (0.015, 0.023)‡ | <0.0001 | |
Ethanol (100 grams/week) | - | 0.039 (0.021, 0.057)‡ | <0.0001 | |
BMI | - | 0.025 (0.021, 0.029)‡ | <0.0001 | |
Systolic BP | - | 0.017 (0.015, 0.019)‡ | <0.0001 | |
Diastolic BP | - | -0.013 (-0.015, -0.011)‡ | <0.0001 | |
Fibrinogen | - | 0.0002 (0.0000, 0,0004)‡ | 0.0335 | |
Age (1 year of age) | 0.028 (0.020, 0.036)† | <0.0001 | 0.017 (0.009, 0.025)† | <0.0001 |
Additional diameter increase per year of age | ||||
Among Diseased Subset | 0.018 (0.004, 0.032)§ | 0.0139 | 0.013 (-0.001, 0.027)§ | 0.0611 |
Among Risk Factor Subset | 0.011 (0.003, 0.019)§§ | 0.0096 | 0.011 (0.003, 0.019)§§ | 0.0101 |
* Basic model independent variables = age, race, gender, height, dummy variables for disease severity, and interaction terms between age-disease severity dummy variables. Full model added variables = BMI, systolic and diastolic blood pressures, current smoker status, cigarette years of smoking in hundreds, current drinker status and grams of ethanol usually consumed per week in 100 g, fibrinogen value.
‡Beta indicates the expected difference in millimeters of CCA diameter associated with the presence of each categorical variable and for each unit increase of continuous variables.
†Beta indicates the expected difference in millimeters of CCA diameter for each year of older age in the low risk subset.
§-§§Betasindicate the expected difference in millimeters of CCA diameter associated with each year of age among the diseased (§) and risk factor or high risk (§§) subsets.