Disparities in Prediabetes Prevalence, Awareness, and Behavior

1
Disparities in PrediabetesPrevalence, Awareness,
and Behavior
Gary He, PhD 1,3 Tetine Sentell, PhD 2 Dean
Schillinger, MD 1,4

• 1 California Diabetes Program
• California Dept of Public Health
• 2 UCSF Department of Community Health Nursing
• 3 UCSF Institute for Health and Aging
• 4 UCSF Center for Vulnerable Populations
• 2009 CDC Diabetes Conference
• April 21-24, 2009
• Long Beach, California

2
Background

• Disparities in prevalence, awareness, and
  behavior in diabetes are well-known, little is
  known in regarding to prediabetes.
• Prediabetes is a condition that raises the risk
  of developing type 2 diabetes, heart disease, and
  stroke.
• People with prediabetes have blood glucose levels
  higher than normal, but not high enough to be
  classified as diabetes.
• Previous study showed more than 57 million US
  adults 20 years of age or older (25) have
  prediabetes.
• Without intervention, about 25 of people have
  prediabetes will develop diabetes within 3-5
  years.
• People with prediabetes are 2-3 times more likely
  to develop type 2 diabetes compared to those who
  are obese.
• The NIH-funded Diabetes Prevention Program found
  that lifestyle changes can significantly reduce
  progression from prediabetes to diabetes

3
Methods

• Objectives of the study are to examine
  socio-demographic disparities in
• Prevalence of prediabetes
• Awareness of prediabetes
• Behaviors in people with prediabetes
• Cross-sectional study using NHANES 2005-2006 .
• NHANES lab measures including FPG and OGTT
• Classification of prediabetes
• Had IFG if FPG gt 100 125 mg/dL, and/or
• Had IGT if OGTT gt140 199 mg/dL
• From NHANES respondents age 18
• 1,887 participated both FPG and OGTT test
• 558 had prediabetes
• represents 29.0 of the U.S. adult population

4
Key Measures

• Prediabetes
• Prediabetes awareness
• Told have prediabetes
• Told at risk for diabetes
• Prediabetes behavior
• Losing weight
• Exercising
• Dieting
• Socio-demographic predictors
• Age, gender, race/ethnicity, education, language
  at home, income, heath insurance, food
  insecurity, smoking status, and body weight
• Mediating/modifying variables
• Report had blood glucose (BG) screening test in
  past three years
• Prediabetes awareness

5
Analysis Plan

• Analyze prediabetes disparities by examining odds
  ratio (OR) and adjusted odds ratio (AOR)
• Prevalence analysis
• In full sample (N1,887)
• Outcome prediabetes
• Predictor socio-demographics
• Awareness analysis
• In persons with prediabetes (N558)
• Outcome awareness
• Predictor socio-demographics
• Mediated model BG screening
• Behavior analysis
• In persons with prediabetes (N558)
• Outcome behavior
• Predictor socio-demographics
• Mediated model awareness
• Multivariate logistic regression models with
  survey design corrections

6
Prevalence of Prediabetes Iby Socio-Demographics
(n1,887)
7
Prevalence of Prediabetes IIby
Socio-Demographics (n1,887)
8
Odds Ratios of Prediabetes
In full sample (n1,887), using bivariate
models Only statistically significant ORs are
reported

• Older age, men, lesser education, limited English
  proficiency, former smokers and overweight/obese
  are at significant higher risk in prediabetes
• The statistical significance of prediabetes odds
  ratios for age and gender are independent of
  other socio-demographic factors in multivariate
  models

9
Odds Ratios of Awareness
In persons with prediabetes (n558) Awareness
defined as Told Have Prediabetes or Told At
Risk Only statistically significant odds ratios
(OR) are reported

• Lesser education and lower proficiency of English
  are significantly associated with lower awareness
  of prediabetes.
• BG screening has positive association with
  prediabetes awareness.
• Education and language are independently
  associated with prediabetes awareness, with or
  without controlling for BG screening.

10
Odds Ratios of Behavior
In persons with prediabetes (n558). Only
statistically significant ORs are reported

• Men are less likely losing weight, exercising,
  and dieting.
• People with less education are less likely losing
  weight or dieting.
• People with lower income are less likely
  exercising.
• Prediabetes awareness is positively associated
  with preventive awareness is significant in
  bivariate analysis (OR for losing weight 1.93,
  CI1.01, 3.69)
• Adding prediabetes awareness into the model dose
  not significantly alter the results

11
Limitations

• This is a cross-sectional study. Results are
  limited by temporal factors
• Association may not be causal
• Sample size may not be sufficient in some
  sub-group analyses
• Awareness and behavior are self-report measures
  that may be subject to reporting error

12
Conclusions

• Nearly 1/3 of US adults are at risk of
  progressing to diabetes.
• This risk is particularly high among males, those
  with less education, and those with limited
  English proficiency.
• Only 1 in 5 individuals with prediabetes is aware
  of the prediabetes status.
• Despite higher prevalence among those with less
  education and English proficiency, these
  individuals are least likely to be aware of their
  prediabetes status, regardless of blood glucose
  screening.
• Similarly, males and those with less education
  with prediabetes are less likely to practice
  prevention oriented behaviors, even after
  accounting for their prediabetes awareness.
• Results suggest that disparities in prediabetes
  are compounded across steps in the
  prevalence-awareness-action pathway.
• Multi-level interventions are needed that include
  health communication, self-management support,
  and policy changes to influence behavior through
  alternate pathways.