BMI is a standard parameter in all clinical settings—both inpatient and outpatient. It is used in the pediatric population, in young and older adults, and in the chronically ill as a marker of health.
Using BMI, clinicians can determine whether a child is growing optimally or predict a patient’s risk of developing diabetes mellitus or cardiovascular disease. In the chronically ill, BMI has been correlated with mortality. Among those who treat hemodialysis (HD) patients, BMI has a critical role in the concept of “reverse epidemiology.”
Reverse epidemiology has been used heavily to indicate that unlike the healthy population, HD patients survive longer with higher BMIs. Kalantar-Zadeh et al presented the classic line graph depicting both HD and general populations (Kidney Int. 2003;63:793-808).
In this graph, relative risk of death in the general population rose to greater than 1.4 at BMI values between 32 and 34.9. In direct contrast, relative risk of death in HD patients decreased to less than 0.8 at the same BMI range. These startling data caused clinicians everywhere to question whether recommending weight loss to their obese HD patients was “best practice.” And in fact, in patients not eligible for transplant, weight loss appears only to increase the risk of death.
Multiple studies have confirmed Kalantar-Zadeh’s findings; however, the picture is not quite so clear-cut. Recently researchers have begun to examine the synergistic impact of abdominal obesity and BMI. Postorino et al studied end-stage renal disease patients on dialysis and found that waist circumference was an indirect predictor of all-cause mortality and cardiovascular mortality (J Am Coll Cardiol. 2009;53:1265-1272). In this study, a 10-cm increase in waist circumference equated with a 23% increase in risk of death. However, patients at highest risk were those with low BMI and large waist circumference.
Similarly, Cordeiro et al used a tool called the “conicity index” to measure the relationship of height, weight, and waist circumference (Nephrol Dial Transplant; published online ahead of print). Thus, a person who is shaped more like a double cone, with the points at the head and the feet and the base of the two cones meeting at the waist is at greater risk for death than a person whose body is shaped like a cylinder. Interestingly, the Cordeiro paper broke the conicity index down into tertiles: the lowest tertile group had a mean index of 1.3; the middle tertile group had a mean index of 1.41; and the highest tertile group had a mean index of 1.54.
These groups were then tested for differences in key nutritional and inflammatory markers. Presence of protein-energy wasting (as defined by the subjective global assessment), BMI, fat body mass, handgrip strength, interleukin-6, and C-reactive protein were all significantly different between conicity index groups. Given these data, one might hypothesize that in the HD population, having a high BMI and a lower waist circumference (i.e., being shaped like a cylinder) is better than having a low BMI and a large waist circumference (i.e., being shaped like a cone).
In the pre-dialysis CKD population, these relationships get even muddier. When in fact does reverse epidemiology kick in and make being larger better? The answer may lie in new data published by Obermayer et al (Nephrol Dial Transplant. 2009;24:2421–2428). These data come from the Vienna Health Screening Initiative, a longitudinal cohort study that ran from 1990 through 2006 and involved more than 49,000 volunteers.
The researchers identified subjects with CKD and categorized them by glomerular filtration rate (GFR). BMI and hazard ratios for cardiovascular death were then assessed according to mean GFR (in mL/min/1.73 m2): group 1 had a mean GFR of 105; group 2, a mean GFR of 75; and group 3, a mean GFR of 45. In groups 1 and 2, the hazard ratios were similar to that of the general population, and as the BMI rose from approximately 25, the risk of cardiovascular death also rose. However, in group 3, the opposite was true; the curve very clearly showed that the lowest risk of death was achieved at the highest BMI.
Promote healthy eating habits
These data are extremely valuable to the clinician working with overweight and obese patients as they progress through the stages of CKD. Patients who are in the higher GFR ranges probably will benefit from counseling on weight-loss techniques since lower BMI values are less predictive of cardiovascular death.
However, as the patients enter late stage 3 and stage 4 CKD, using weight loss as a treatment requires careful consideration. Instead, clinicians may want to promote healthy eating habits and exercise to try and prevent increased waist circumference without restricting calories. The impact of exercise on waist circumference and ultimately mortality has not been tested in a randomized clinical trial to date; however, extrapolating from the cross-sectional data available indicates that this type of intervention may provide more benefit than calorie restriction does.
Dr. Steiber is Coordinator of the Dietetic Internship/Master’s Degree Program at Case Western Reserve University in Cleveland.