Article Text

FRI0578-HPR Why There Are so Few Obese Norwegians: the Validity of Self-Reported Body Mass Index in an Osteoarthritis Cohort
  1. K. Magnusson1,
  2. I.K. Haugen2,
  3. K.B. Hagen1
  1. 1National Resource Center for Rehabilitiaton in Rheumatology, Dept. of Rheumatology
  2. 2Dept. of Rheumatology, Diakonhjemmet Hospital, Oslo, Norway


Background Obesity is one of few modifiable important risk factors for osteoarthritis (OA). Body Mass Index (BMI) is frequently used as a measure of obesity, which is often calculated based on self-reported height and weight. However, to what extent self-reported BMI can be trusted in OA studies is unknown. Knowledge of this topic is important for the interpretation of results from existing studies as well as in the design of future epidemiological studies.

Objectives To investigate the validity of self-reported BMI in an OA cohort.

Methods In a population-based survey, persons aged 40-79 years (n=12155) were sent questionnaires in 2010 (the Musculoskeletal pain in Ullensaker study). Those who answered “yes” to the question “Have you been diagnosed with osteoarthritis in the knee, hip and/or hands by a doctor and/or on radiographs?” were invited to a clinical examination (n=1019), of whom 630 attended. In the analyses we included 449 participants who had clinical OA in the hips, knees and/or hands (according to the American College of Rheumatology criteria). Height and weight were self-reported on the initial questionnaire, and measured at the clinical examination.

Results See Table 1.

Table 1.

Difference between measured and self-reported height and weight according to weight category

Mean (SD) age was 65 (8.6) years and 129 (28%) were men. Mean (SD) self-reported BMI was 27.0 (4.5) kg/m2, while mean (SD) measured BMI was 28.3 (4.9) kg/m2. There was a dose-response relationship between a higher examined BMI and a greater overreporting of height, underreporting of weight, and consequently a lower reported than measured BMI (Table 1 and figure).

Conclusions OA patients with a high BMI were more likely to underreport their BMI compared to OA patients with a normal BMI. The dose-response relationship between actual BMI-category and underreporting of BMI should be taken into account when designing future studies of OA in which BMI is used as a measure of obesity.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.4040

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.