The Great BMI Debate

One of my RDNs posed this question recently related to BMI levels for older adults:

I've been seeing transfer notes from the hospital and other nursing homes with diet/nutrition histories where RDNs are charting that BMIs of less than 23 is underweight. For example, one note documented that a BMI of 21.3 was underweight "for age" for a man who was 92. State surveyors are also asking for a list of residents with BMI under 21 and wanting to see interventions on them. The MDS does not trigger for a low BMI until under 19. Do we need to adapt our practices?
The National Institute of Health classification of overweight and obesity by body mass index (BMI) is as follows:

Classification - Normalr
Obesity Class - None
BMI (kg/m2) - 18.4-24.9

Classification - Overweightr
Obesity Class - None
BMI (kg/m2) - 25.0-29.9

Classification - Obesityr
Obesity Class - IrnBMI (kg/m2) - 30.0-34.9

Classification - Obesityr
Obesity Class - IIrnBMI (kg/m2) - 35.0-39.9

Classification - Extreme Obesityr
Obesity Class - III
BMI (kg/m2) - > 40

BMI is interpreted based on age, health history, usual body weight, and weight history.

Adults should be assessed for indicators of nutritional status and decline using body mass index (BMI) as one of many factors. Data suggests that a higher BMI range may be protective in older adults and that the standards for ideal weight (BMI of 18.5 to 25) may be too restrictive in the elderly. A lower BMI may be considered detrimental to older adults due to association with declining nutrition status, potential pressure ulcers, infection and other complications. A BMI of 19 or less may indicate nutritional depletion, while a BMI of 30 or above indicates obesity.

In the literature, there is a lot of conversation about a BMI of 21-23 (rather than 18/19) as considered on the low side for older adults. At the same time, there is a lot of conversation about the “obesity paradox” saying a higher BMI might be protective against some diseases and death. There is still a lot of controversy regarding the efficacy of BMI for older adults, regardless of what is considered “too low” or “too high”.

To our knowledge, there are no firm recommendations from any source on BMI cutoffs for older adults. The MDS triggers a CAA if BMI is < 18.5, although as stated above a higher BMI can probably be considered too low for older adults.

In clinical practice, the BMI number is not as important as how it compares to an individual’s history. Monitoring changes over time is what is important.

If state surveyors question whether everyone with a low BMI needs an intervention, consider explaining that if a low BMI was normal for this person’s life history, then we would not attempt to correct it - although interventions might be put in place for other reasons (poor intake, weight loss, wounds, etc.). And for an older person with a high BMI of 35 who had been overweight their whole life, it is highly likely that lifestyle and habits are set and weight loss would probably not be necessary or successful in older age.

The new Academy/ASPEN criteria for diagnosing malnutrition does not use BMI – it uses unintended weight loss, body fat, muscle mass loss (as determined by nutrition focused physical assessment and/or handgrip strength in the case of severe malnutrition) and other factors. The National Quality Forum Measure #128 (NWF 0421) Preventive Care and Screening uses >23 and <30 for those over the age of 65.

There are several reference articles on BMI in the elderly which all suggest higher BMIs for those over 65:

-Flicker et al JAGS 2010; 68: 234.

-Bell et al JAMDA 2013; 14: 94-100.

-Winter J et al Am J Clin Nutri 2014; 99:875-890 Sorkin, J Am J Clin Nutri 2014; 99: 759-760.

-Winter J, MacInnis R, Wattanapenpaiboon N and Nowson C. BMI and all-cause mortality in older adults: a meta-analysis. First published January 22, 2014, doi: 10.3945/ajcn.113.068122. Am J Clin Nutr.