Skip to main content

Table 6 Age-adjusted analysis the association between a number of baseline co-variables and incident DR amongst those DR free at baseline in the Nakuru eye disease cohort study

From: The incidence of diabetes mellitus and diabetic retinopathy in a population-based cohort study of people age 50 years and over in Nakuru, Kenya

 

Study sample, n = 1421

 

No at risk of diabetic retinopathy

Incident diabetic retinopathy

Risk per 1000/6 years (95% CI)

Age adjusted risk ratio (95% CI)

Age

 50–59

691

12

22.0 (11.0–43.4)

Baseline

 60–69

466

6

13.3 (5.4–32.8)

0.6 (0.2–2.0)

 70–79

212

2

8.6 (2.1–34.8)

0.4 (0.1–1.9)

 80+

52

0

Gender

 Male

686

12

20.5 (10.9–38.2)

Baseline

 Female

735

8

11.5 (4.8–27.1)

0.5 (0.2–1.5)

BMI (2 missing values)

 Underweight

166

1

5.3 (0.7–36.6)

Baseline

 Normal

727

5

8.1 (3.3–19.5)

1.5 (0.2–13.0)

Overweight

336

11

39.6 (20.3–75.9)

6.4 (0.8–53.6)

 Obese

190

3

12.9 (4.1–40.2)

2.0(0.2–21.9)

Location

 Rural

1053

12

11.6 (6.6–20.4)

Baseline

 Urban

368

8

23.5 (10.3–52.7)

1.8 (0.7–4.4)

SES Quartile (6 missing values)

 Lower

309

1

3.0 (0.4–21.8)

Baseline

 Lower middle

394

2

4.9 (1.2–19.4)

1.6 (0.4–6.9)

 Upper middle

380

9

25.1(13.1–47.5)

7.5 (1.0–58.1)

 Upper

332

8

29.3 (13.3–63.3)

8.3 (1.0–66.9)

Smoker

 Never

964

16

18.3 (10.0–33.1)

Baseline

 Former

113

0

 Current

344

4

14.3 (5.4–37.5)

0.8 (0.3–2.6)

Hypertension (2 missing values)

 No

764

8

15.0 (6.0–36.7)

Baseline

 Yes

655

12

16.9 (9.3–30.3)

1.2 (0.4–3.6)

Alcohol (1 missing value)

 Never

580

8

15.3 (6.5–35.6)

Baseline

 Former

611

9

15.0 (7.4–30.1)

1.1 (0.4–3.3)

 Current

229

3

19.0 (4.1–83.5)

1.3 (0.2–7.9)

Ethnic group

 Kikuyu

895

14

16.9 (10.3–27.8)

Baseline

 Kalenjin

357

3

8.5 (1.9–36.9)

0.5 (0.1–2.3)

 Other

169

3

22.6 (7.2–69.1)

1.1 (0.4–3.0)

Education level

 No education

123

4

42.8 (15.3–114.2)

Baseline

 Primary

422

0

 Secondary

726

14

21.7(11.9–39.0)

0.6(0.2–1.9)

 College/Uni

150

2

12.6(3.0–51.6)

0.3(0.0–1.8)

  1. For multivariable analysis, an initial model was fitted that included those variables shown to be associated with outcome in age-adjusted analysis (using a Wald test threshold p-value of <0.05 to indicate association). A backward stepwise approach was then applied in order to obtain a final multivariable model, removing variables with p > 0.05 one-by-one