TABLE 5.
Regression coefficient estimates and standard errors for associationa of categorized body mass index with systolic and diastolic blood pressure change (mmHg) in the HCHS/SOL target population using all compared methods
| Systolic blood pressure (mmHg) | Diastolic blood pressure (mmHg) | |||||||
|---|---|---|---|---|---|---|---|---|
| Overweightb | Obeseb | Overweightb | Obeseb | |||||
|
| ||||||||
| Method | Estimate | SE | Estimate | SE | Estimate | SE | Estimate | SE |
|
| ||||||||
| Original Analysisc | ||||||||
|
| ||||||||
| Observed | −0.252 | 0.440 | −0.036 | 0.467 | 1.312 | 0.326 | 1.722 | 0.391 |
| Exclude Treated | −0.465 | 0.448 | −0.122 | 0.459 | 1.017 | 0.340 | 1.506 | 0.401 |
| Add Constant | −0.272 | 0.442 | 0.303 | 0.465 | 1.144 | 0.326 | 1.699 | 0.384 |
| Add Random | −0.284 | 0.447 | 0.487 | 0.474 | 1.126 | 0.331 | 1.856 | 0.392 |
| Add Expected Effect | −0.330 | 0.439 | 0.073 | 0.462 | 1.244 | 0.324 | 1.651 | 0.386 |
| IPW Proposed | −0.503 | 0.364 | 0.093 | 0.389 | 0.994 | 0.294 | 1.601 | 0.343 |
| DR Proposed | −0.432 | 0.370 | 0.169 | 0.379 | 1.154 | 0.292 | 1.666 | 0.333 |
|
| ||||||||
| Sensitivity Analysisd | ||||||||
|
| ||||||||
| IPW Proposed | −0.533 | 0.366 | 0.130 | 0.391 | 0.976 | 0.293 | 1.610 | 0.346 |
| DR Proposed | −0.462 | 0.373 | 0.196 | 0.380 | 1.132 | 0.291 | 1.669 | 0.335 |
Abbreviations: SE, standard error.
NOTE: There were 5,301 participants in the analytic sample (weighted proportion 66%) without hypertension at follow-up, 2,114 participants (weighted proportion 22%) with untreated hypertension at follow-up, and 1,520 participants (weighted proportion 12%) with treated hypertension at follow-up. The average length of follow-up was 6.1 years (standard deviation = 0.9 years). The average systolic blood pressure change over follow-up was a decrease of 1.307 mmHg (standard deviation = 10.7) for the subgroup without hypertension at follow-up and an increase of 10.023 mmHg (standard deviation = 13.5) for the subgroup with untreated hypertension at follow-up. The average diastolic blood pressure change over follow-up was a decrease of 0.414 mmHg (standard deviation = 8.8) for the subgroup without hypertension at follow-up and an increase of 6.150 mmHg (standard deviation = 9.5) for the subgroup with untreated hypertension at follow-up.
Adjusted for baseline blood pressure, age, sex, time between study visits, and study field center.
Body mass index (BMI) was categorized as underweight or normal weight (BMI < 25 kg/m2), overweight (25 kg/m2 ≤ BMI < 30 kg/m2), or obese (BMI ≥ 30 kg/m2). The reference category was underweight or normal weight.
For the original analysis, time of treatment initiation was approximated as the time of the first survey in which the participant indicated treatment.
For the sensitivity analysis, time of treatment initiation was approximated as the mid-point of the time interval between the first survey in which the participant indicated treatment and the previous survey.