|
|
Volume
8: No. 6, November 2011
ORIGINAL RESEARCH
A Latent Class Modeling Approach to Evaluate Behavioral Risk Factors and Health-Related Quality of Life
This diagram shows how the indicators, domains, outcomes, and sociodemographic characteristics are related. Twenty-four manifest dichotomous health indicators (represented by rectangles) are grouped into 4 domains (circles) that are related to 2 outcomes after adjustment for sociodemographic characteristics. The first 5 indicators measure the domain risky behaviors; the next 8 indicators, health conditions; the next 5 indicators, health care access; and the remaining
6 indicators, use of preventive services. The outcomes, frequent physical distress and frequent mental distress, are
then regressed on these 4 domains and on age group, sex, race/ethnicity, education level, and employment status.
Figure 1. Indicators, health domains, outcomes, and covariates for latent class modeling approach to analyze 2008 data from the Rhode Island Behavioral Risk Factor Surveillance System. Abbreviations:
BMI, body mass index; PSA, prostate-specific antigen; Pap, Papanicolaou.
Return to article
| Behavior |
Class 3: High Risk (9.1%
of Population), % |
Class 2: Intermediate Risk (31.0%
of Population), % |
Class 1: Low Risk (59.9%
of Population), % |
| Does not always wear seatbelt |
28.6 |
40.1 |
7.2 |
| Current smoking |
33.9 |
31.0 |
7.8 |
| Binge drinking |
99.5 |
17.4 |
5.1 |
| Heavy drinking |
53.8 |
2.5 |
2.1 |
| Drinking and driving |
29.5 |
3.4 |
0.4 |
Figure 2. Probabilities of risky behaviors by latent class, Rhode Island Behavioral Risk Factor Surveillance System, 2008. Behaviors refer to the past 30 days. Binge drinking defined as ≥5 drinks for men or ≥4 drinks for women on 1 occasion. Heavy drinking defined as ≥3 drinks/d for men or ≥2 drinks/d for women.
Return to article
| Condition |
Class 3: Multiple Conditions (5.7%
of Population), % |
Class 2: Intermediate Conditions (18.1%
of Population), % |
Class 1: Few Conditions (76.2%
of Population), % |
| Asthma |
15.6 |
20.0 |
8.0 |
| Diabetes |
33.0 |
20.4 |
2.4 |
| Obesity |
41.0 |
39.6 |
16.5 |
| Disability |
58.7 |
55.5 |
9.7 |
| ≥6 teeth removed |
54.4 |
37.5 |
4.9 |
| Angina |
63.3 |
2.1 |
0.8 |
| Heart attack |
64.3 |
0.0 |
0.4 |
| Stroke |
17.2 |
5.2 |
0.5 |
Figure 3. Probabilities of health conditions by latent class, Rhode Island Behavioral Risk Factor Surveillance System, 2008. Health conditions defined as having ever been told by a health professional that they had the condition. Obesity defined as body mass index ≥30 kg/m2. Return to article
| Access Indicator |
Class 3: Limited Access (10.2%
of Population), % |
Class 2: Intermediate Access (7.3%
of Population), % |
Class 1: Has Access (82.5%
of Population), % |
| No regular provider |
44.2 |
97.2 |
0.0 |
| No routine checkup in past year |
61.3 |
55.8 |
12.9 |
| No dental visit in past year |
58.3 |
22.1 |
16.2 |
| No health coverage |
68.3 |
19.2 |
2.9 |
| Does not see doctor due to cost |
73.7 |
0.1 |
4.8 |
Figure 4. Probabilities of health care access by latent class, Rhode Island Behavioral Risk Factor Surveillance System, 2008. Return to article
| Service |
Class 3: Lower Use (21.1%
of Population), % |
Class 2: Intermediate Use (51.6%
of Population), % |
Class 1: Higher Use (27.3%
of Population), % |
| No pneumonia vaccine ever |
85.3 |
94.0 |
28.5 |
| No flu vaccine in past year |
74.9 |
70.9 |
20.5 |
| No PSA test in past 2 y, ages ≥40 y |
98.8 |
33.3 |
11.5 |
| No sigmoidoscopy or colonoscopy ever, ages ≥50 y |
76.6 |
33.3 |
12.4 |
| No mammogram in past 2 y, ages ≥40 y |
81.4 |
3.2 |
6.0 |
| No Pap test in past 3 y |
48.4 |
0.1 |
13.6 |
Figure 5. Probabilities of use of preventive services
by latent class, Rhode Island Behavioral Risk Factor Surveillance System, 2008. The prostate-specific antigen (PSA) test question was only asked of men. The mammogram and Papanicolaou
(Pap) test questions were only asked of women. Return to article
|
|
