Table 4.

Influence of 6-year alcohol patterns on health-related quality of life over time among all participants and those in consistently good health

Model descriptor	Model 2	Model 3
	Pattern predictor	Consistently healthy group
Follow-up years	1998–2008	1994–2008
n	5,404	2,645
Variable	Estimates	Estimates
Effect of alcohol on HUI3 at initial follow-up
Persistent nonuser	−.059**	−.065**
Persistent former user	−.078***	−.023
Persistent infrequent user	−.020	−.034*
Persistent moderate user	Referent	Referent
Persistent heavy user	−.012	−.051
Decreasing user	−.108***	−.085***
Increasing user	−.014	−.011
U-shaped user	−.051**	−.027
Inverted U-shaped user	−.069***	−.066***
Effect of alcohol on HUI3 rate of change
Persistent nonuser	−.006	−.001
Persistent former user	−.007	−.009
Persistent infrequent user	−.005	.003
Persistent moderate user	Referent	Referent
Persistent heavy user	−.016	−.021
Decreasing user	−.025***	−.003
Increasing user	−.006	−.008
U-shaped user	.001	−.001
Inverted U-shaped user	−.008	.000
HUI3 at initial follow-up	.828***	.864***
HUI3 score rate of change	−.051***	−.046***
Variance of HUI3 at initial follow-up	.021***	.020***
Variance of HUI3 rate of change	.002***	.003***
RMSEA	.048	.038

Notes: Models were adjusted for age, gender, marital status, education, household income, smoking, chronic illness, body mass index, physical activities, and nonspecific psychological distress. HUI3 score = Health Utilities Index Mark 3; RMSEA = root mean square error of approximation.

^*p < .05;

^**p < .01;

^***p < .001.