Skip to main content

Table 1 Description of the outcome and explanatory variables from NHANES and BRFSS and the corresponding odds ratios (OR) and 95% confidence intervals (95% CI).

From: Diabetes prevalence and diagnosis in US states: analysis of health surveys

Variable

Reason for inclusion in analysis

Possible values

OR

(95% CI)

Outcome (dependent) variables for the regression

Undiagnosed diabetes (available in NHANES; predicted in BRFSS)

Outcome variable to estimate undiagnosed diabetes

0 (FPG < 126 mg/dL)

1 (FPG ≥ 126 mg/dL)

-

-

Explanatory (independent) variables

Sex

Predictor of diabetes, possibly because of differences in lifestyle and health care determinants

Male

1.0

-

  

Female

0.47

0.29, 0.77

Age (years)

Predictor of diabetes

30-39

1.0

-

  

40-49

1.07

0.41, 2.81

  

50-59

3.25

1.41, 7.50

  

60-69

7.49

3.55, 15.82

  

≥ 70

7.02

3.23, 15.26

Race *

Predictor of diabetes and health care access

Non-Hispanic white

1.0

-

  

Non-Hispanic black

1.11

0.63, 1.96

  

Hispanic

2.03

1.07, 3.83

  

Other

0.26

0.03, 1.91

Doctor visit (have you seen a doctor in the past year?) †

Indicator for diabetes knowledge and control

No

0.49

0.23, 1.05

  

Yes

1.0

-

Insurance status (do you currently have health insurance?)

Indicator for diabetes knowledge and control

No

1.58

0.83, 3.02

  

Yes

1.0

 
  

Missing

0.76

0.07, 8.42

BMI (kg/m2) ‡

Determinant of diabetes and indicator for selected lifestyle factors such as diet and physical activity

< 25

1.0

-

  

25-29

1.85

0.93, 3.67

  

≥ 30

4.29

2.25, 8.17

Smoking (do you now smoke cigarettes?)

Indicator for lifestyle factors

Yes ("everyday" or "some days")

1.0

-

  

No ("not at all")

1.27

0.65, 2.49

  

Missing

0.86

0.51, 1.45

  1. * We combined Mexican Americans and other Hispanics in NHANES to match the race categories in BRFSS.
  2. † This variable was defined as a composite of multiple questions about physician contact for specific conditions in the BRFSS.
  3. ‡ BMI was corrected for bias in self-reported height and weight in telephone interviews, using methods described elsewhere [2].