Self-perceived vs actual and desired weight and body mass index in adult ambulatory general internal medicine patients: a cross sectional study

Background: No study has compared patients’ self-reported heights and weights (and resultant self-reported body mass indexes [BMIs]) with their actual heights, weights, and BMIs; their self-perceived BMI categories; and their desired weights and BMIs and determined rates of clinicians’ documented diagnoses of overweight and obesity in affected patients in a single patient group. The objectives of this study were to make these comparisons, determine patient factors associated with accurate self-perceived BMI categorization, and determine the frequency of clinicians’ documented diagnoses of overweight and obesity in affected patients.

Results: A total of 508 consecutive adult general internal medicine outpatients (257 women, 251 men; mean age, 62.9 ± 14.9 years) seen at Mayo Clinic in Rochester, Minnesota, between November 9 and 20, 2009, completed a questionnaire in which they reported their heights, weights, self-perceived BMI categories (“underweight,” “about right,” “overweight,” or “obese”), and desired weights. These self-reported data were compared to actual heights, actual weights, and actual BMI categories (measured after the questionnaire was completed). Overall, 70% of the patients were overweight or obese. The average self-reported weight was significantly lower than the average actual weight (80.3 ± 20.1 kg vs 81.9 ± 21.1 kg; P < .001). The average self-reported BMI was significantly lower than the average actual BMI (27.6 ± 5.7 kg/m2 vs 28.3 ± 6.1 kg/m2P < .001). Overall, 32% of patients had obesity; however, only 6% perceived they were obese. Accuracy of self-perceived BMI category decreased with higher actual BMI category (P < .001 for trend). Female sex, higher education level, smoking status, and lower BMI were associated with higher accuracy of self-perceived BMI category. Desired weight loss increased with higher self-perceived and actual BMI categories (P < .001 for trends). Of the 165 patients who actually were obese, only 40 (24%) had obesity documented as a diagnosis in their medical records by their clinicians. Statistical tests used were the paired t test, the Pearson χ2 test, the Cochrane-Armitage trend test, the Wald test of marginal homogeneity, analysis of variance, and univariate and multivariate logistic regression.

Conclusions: Many obese patients inaccurately perceive their BMI categories; accuracy decreases with increasing BMI. Clinicians should inform patients of their BMIs and prescribe treatment plans for those with overweight and obesity.

Free Full-text

Alex’s Notes: More than one-third of US adults are obese and yet many do not perceive themselves this way. This weight problem has progressed enough to have the American Medical Association (AMA) classify obesity as a disease that requires medical intervention, but this view is not held by many persons and what was once considered overweight has now become viewed as “just right.” The aim of the current study was to compare actual and self-perceived BMIs and weight categories, and determine factors associated with accurate BMI perception.

To this end, 508 outpatients from the Mayo Clinic in Rochester, Minnesota, USA completed a four-item questionnaire, and then had actual weights and heights measured. The four questions were:

  1. What is your height? (How tall are you?)
  2. What is your weight? (How much do you weight?)
  3. Which of the following best describes your weight?
    1. Underweight
    2. About right
    3. Overweight
    4. Obese
  4. What is your preferred weight? (How much would you like to weigh?)

Actual BMI was calculated from actual weight and height, self-reported BMI was calculated from the listed height and weight, self-perceived BMI was based on the third question where “about right” was considered normal BMI, and desired BMI was calculated using the self-reported height and desired weight. Data from the participants’ medical records were also reviewed for clinician’s diagnosis of obesity and other lifestyle factors that may influence self-perceived weight status.

The Truth

There was a nearly even split of men and women, and the average age of the participants was 62.9 years. The majority (91%) was never smokers and 72% had at least some college education.

Seventy percent of the participants were overweight or obese, but the average self-reported BMI was significantly lower than the actual BMI thanks to the self-reported height being significantly taller and the self-reported weight being significantly lower than reality. Even so, the desired weight and BMI were significantly lower than the actual and self-reported values, and the average desired weight loss averaged 9.3kg (20 lbs). Moreover, only 24% of obese patients were documented as obese by their clinicians.

A solid 32% of participants were obese, yet only 6% perceived themselves as obese, and the accuracy of self-perceived BMI decreased as BMI increased. For instance, 71% of the normal-weight persons accurately perceived themselves as being normal weight, but only 58% of overweight and 20% of obese persons perceived themselves to be overweight and obese, respectively. Even more surprising, of the 42% inaccurate overweight persons, 98% perceived themselves to be normal-weight and 2% to be underweight. Of the 80% inaccurate obese persons, 90% perceived they were merely overweight and the remainder normal weight.

The above speaks to the notion that society’s view of weight has changed. So which came first? It appears that the average weight and the perceived normal weight have increased, so was it societies acceptance and accommodations to being overweight that led to it being viewed as the norm? Or was it becoming the norm that led to society’s acceptance and accommodations? Using multivariate analysis, the female sex, a higher education, being a never smoker, and having a lower actual BMI increased the odds of accurate self-perceived BMI.

More truth

One interesting take-away from this study is that although self-reported heights and weights may be useful for epidemiologic studies, the current results suggest that they may not be accurate. Nearly all the inaccurate participants believed they were in a lower BMI category than they actually were, and this inaccuracy increased with actual BMI.

Another interesting point is the clinicians’ position. Only a quarter of the current study’s clinicians documented their patient’s obesity, and yet these health care professionals are in a very unique position to help these persons recognize their unhealthy weight. According to the Health Belief Model of behavior change, only when an obese person recognizes their unhealthy weight and its associated health risks will they consider modifying lifestyle behaviors and losing weight. This raises the question of why the clinicians are hesitant to diagnose overweight and obesity.

Some simply don’t consider obesity a disease. Others may consider counseling a futile effort. Still others may have had no financial incentive, as Medicare did not reimburse for obesity counselling until recently. According to unpublished data from a Mayo Clinic survey of perceived barriers among their clinicians, lack of time to discuss patients’ weights, other clinical priorities, perceived lack of effective treatments, provider unpreparedness to discuss obesity, and patient sensitivity to the term “obesity” are all seen as obstacles. Other factors include patients’ concerns about stigma and “anti-fat bias” by clinicians. Finally, some clinicians simply put the burden on the patients and assume the barriers to weight loss belong to them, even though the patients may not see the need without clinician intervention.

It must be mentioned that the average age was old at 63 years, and ethnic and socioeconomic demographics were not included. The participants were also confined to the general area of the Mayo Clinic. Moreover, some persons with “overweight” BMIs may simply have more muscle and less body fat. The study would have been strengthened with measurements of waist circumference and some form of body fat analysis (e.g. skinfold measurement). Despite these limitations, this study does provide further evidence of disconnect between actual and self-perceived weight status.


Network Affiliates

Quick Links I

Our Location

2908 Brownsboro Rd
Suite 103
Louisville, KY 40206
(502) 690-2200

SHR Newsletter

Subscribe to our FREE newsletter
to receive the latest updates in your inbox!
SHR Newsletter
Internet Radio