Skip to main content
. Author manuscript; available in PMC: 2019 Jul 31.
Published in final edited form as: Eur Respir J. 2014 May 15;44(2):405–414. doi: 10.1183/09031936.00043614

Table 4:

Differences in two-year lipid changes between PAP adherent and non-users

Lipid BMI Group*, LS Mean±SE Change
p
Adherent Non-Users
Total Cholesterol Overall 6.5±2.2§ 8.8±2.8§ 0.528
<30 2.4±4.0 9.7±4.4§ 0.248
30–35 15.1±3.4§ 10.9±4.9§ 0.493
≥35 1.8±3.6 6.7±5.3 0.468

LDL Cholesterol Overall −0.2±1.9 2.2±2.5 0.475
<30 −5.7±3.8 1.6±4.1 0.210
30–35 6.4±3.0§ 2.6±4.3 0.481
≥35 −1.4±3.3 2.0±4.8 0.579

HDL Cholesterol Overall 6.7±0.7§ 7.3±0.9§ 0.619
<30 6.7±1.2§ 7.3±1.3§ 0.750
30–35 7.6±1.0§ 8.2±1.5§ 0.735
≥35 5.6±1.1§ 7.0±1.6§ 0.501

Triglycerides Overall −0.9±5.6 −7.1±7.4 0.517
<30 17.5±11.4 6.9±12.5 0.548
30–35 8.9±10.5 8.3±14.9 0.976
≥35 −26.4±7.3§ −29.1±10.8§ 0.844

Significant differences shown in bold

*

The propensity score sample included 199 adherent (56 with BMI<30; 65 with BMI 30–35; 78 with BMI≥35) and 118 non-users (47 with BMI<30; 33 with BMI 30–35; 38 with BMI≥35)

p-values for PAP by BMI group interaction: p=0.557 for TC, p=0.500 for LDL, p=0.615 for HDL, p=0.563 for TG

p-value from ANCOVA comparing adherent and non-users within PS designed study, adjusted for PS subclass and baseline lipid level

§

Within group estimate of lipid change significantly (p<0.05) different from zero

BMI: body mass index; LDL: low-density lipoprotein; HDL: high-density lipoprotein; LS: least squares; SE: standard error

HHS Vulnerability Disclosure