Association between cardiometabolic risk factors and body mass index based on diagnosis and treatment codes in an electronic medical record database

Diana Brixner, Sameer R. Ghate, Carrie McAdam-Marx, Rami Ben-Joseph, Qayyim Said

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

Background: Managed care organizations (MCOs) have access to treatment and diagnosis information from administrative claims data but generally have limited or no access to clinical information about laboratory values or biometric values such as body mass index (BMI) or waist circumference. Thus, MCOs are generally unable to identify overweight patients with cardiometabolic risk factors that put them at a high risk of poor outcomes. The National Heart, Lung, and Blood Institute defines normal body weight as a BMI (ratio of weight in kilograms to height in meters squared [kg/m2]) from 18.5 to 24.9 kg/m2, overweight as 25.0 to 29.9 kg/m2, and obesity as a BMI of 30 kg/m2 or greater. Current guidelines for weight-loss pharmacotherapy, including U.S. Food and Drug Administration-approved label indications, specify use in patients with a BMI of 30 kg/m2 or greater, or a BMI >27 kg/m2 and at least 1 concomitant cardiometabolic risk factor such as controlled hypertension, diabetes, or dyslipidemia. Objective: To evaluate the association of cardiometabolic risk factors with BMI as recorded in a database of electronic medical records (EMRs). Methods: Each patient had a minimum look-back observation period of 2 years from the last date of activity in the EMR. Patients with a BMI of 18 kg/m2 or greater recorded in the EMR at any time during the 10-year period from January 1996 through December 2005 were stratified into groups by the number of cardiometabolic risk factors and by individual cardiometabolic risk for those with just 1 risk factor. Cardiometabolic risk factors were identified from diagnoses and prescription orders in the EMR associated with high triglyceride levels, low high-density lipoprotein cholesterol (HDL-C) levels, type 2 diabetes, or hypertension. Unadjusted and adjusted odds ratios (ORs) of having a BMI >27 kg/m2 were calculated for each risk factor group and for patients with no risk factors. Using logistic regression analysis, ORs were adjusted for age, gender, insurance type, region, medications associated with weight gain or weight loss, and diseases that modify weight. Results: A total of 499,593 patients with a BMI of 18 kg/m2 or greater were identified; 56.4% (n = 281,988) had a BMI >27 kg/m2, whereas 43.6% (n = 217,605) had a BMI between 18 and 27 kg/m2. Compared with patients with no risk factors (n = 289,960), patients with 1-4 risk factors (n = 209,633) were significantly more likely to have a BMI >27 kg/m2; 48.4% of patients without cardiometabolic risk factors had a BMI >27 kg/m2, compared with 63.3%, 79.8%, 84.6%, and 88.5% for patients with 1-4 cardiometabolic risk factors, respectively (all comparisons P<0.001). Adjusted ORs for having a BMI >27 kg/m2 were 2.64 (95% confidence interval [CI]=2.51-2.77) for type 2 diabetes, 2.21 (95% CI = 2.05-2.37) for elevated triglycerides, 1.91 (95% CI = 1.88-1.94) for hypertension, and 1.45(95% CI = 1.29-1.63) for low HDL-C. Adjusted ORs for having a BMI >27 kg/m2 were 3.58(95% CI=3.47-3.69), 4.24 (95% CI = 3.93-4.59), and 5.07 (95% CI = 3.77-6.81) for patients with any 2, 3, and 4 risk factors respectively, relative to patients with no cardiometabolic risk factors. Conclusions: For patients with cardiometabolic risk factors, compared with patients with no risk factors, the odds of having a BMI >27 kg/m2 were multiplied by 1.45-5.07, depending on the type and number of risk factors. Diagnoses and treatment indicators for cardiometabolic risk factors are potential indicators of obesity. Copyright

Original languageEnglish (US)
Pages (from-to)756-767
Number of pages12
JournalJournal of Managed Care Pharmacy
Volume14
Issue number8
DOIs
StatePublished - Oct 2008
Externally publishedYes

ASJC Scopus subject areas

  • Pharmacy
  • Pharmaceutical Science
  • Health Policy

Fingerprint

Dive into the research topics of 'Association between cardiometabolic risk factors and body mass index based on diagnosis and treatment codes in an electronic medical record database'. Together they form a unique fingerprint.

Cite this