Prevalence and Associated Risk Factors of Falls among Rural Community-Dwelling Older People: A Cross-Sectional Study from Southern Sri Lanka
Table 5
Logistic regression analysis of risk factors for falls (n=300).
Risk factor
Univariate analysis
Multivariate analysis
Crude OR
95% CI OR
P value
Adjusted OR
95% CI OR
P value
Biological factors
Age
0.1
0.1-0.3
<0.001
0.1
0.0-0.3
<0.001
Gender
1.9
1.2-3.2
0.008
3.0
1.5- 5.9
0.002
Presence of at least one chronic medical condition
3.4
1.9-5.9
<0.001
-
-
-
Presence of more than one chronic medical condition
2.3
1.4-4.0
0.001
-
-
-
Presence of more than two chronic medical conditions
3.2
1.3-7.8
0.001
-
-
-
Diabetes mellitus
2.8
1.5-5.1
0.001
2.7
1.2- 6.3
0.01
Ischemic heart diseases
3.1
1.5-6.5
0.002
-
-
-
Hypertension
2.2
1.3-3.6
0.001
-
-
-
Vertigo
3.5
1.0-12.3
0.04
-
-
-
Self-reported vision problem affecting day-to-day life
4.0
1.9-8.3
<0.001
-
-
-
Self-reported balance/gait problems or foot abnormality
7.3
4.3-12.6
<0.001
4.2
2.0- 8.4
< 0.001
Self-reported sleeping problems
0.2
0.1-0.4
<0.001
-
-
-
Behavioral factors
Usage of at least one long term medication
2.6
1.6-4.3
<0.001
3.5
1.2- 10.4
0.02
Usage of more than one long term medication
2.5
1.4-4.4
0.002
-
-
-
Antihypertensive
1.7
1.0-2.8
0.03
0.2
0.0- 0.8
0.02
Hypoglycemic drugs
3.1
1.6-5.8
<0.001
-
-
-
Cardiac/Lipid lowering drugs
2.2
1.2-4.3
0.01
-
-
-
Exposure to sun
0.2
0.1-0.4
<0.001
-
-
-
Nutritional status
0.5
0.3-0.8
0.007
-
-
-
Usage of walking aids
0.1
0.0-0.2
<0.001
-
-
-
Environmental factors
Floor material
11.6
2.5-53.6
0.002
-
-
-
Light source
8.2
1.7-39.4
0.009
-
-
-
Water source
3.0
1.5-5.7
0.001
-
-
-
Socioeconomic factors
Level of education
0.5
0.3-0.8
0.006
OR=Odds Ratio; CI=Confidence Interval. The variables which are significant in the chi square or fisher’s exact test with falls prevalence were used for univariate logistics regression analyses and were presented in this table. The variables which were significant with the falls prevalence in univariate logistics regression were further evaluated with multivariate logistic regression and only the significant variables were presented in the table.