More on Body Fat Cutoff Points–Reply–I

We greatly appreciate the interest of Ho-Pham et al in our article¹ and in percentage of body fat (PBF) cutoff points in general. To provide as detailed a response as possible, Oreopoulos has allied herself with Lavie and Romero-Corral, who cowrote the accompanying editorial,² and Snitker, the author of a recent letter on a related topic³ and an unpublished correspondence with Mayo Clinic Proceedings criticizing the use of the World Health Organization (WHO) 1995 Technical Report⁴ to support specific cutoff points for PBF.

We enjoyed reading the historical account of the misattribution of the WHO 1995 Technical Report⁴ as recommending specific PBF cutoff points, which 3 of us (A.O., C.J.L., A.R.-C.) have inadvertently used as well.^1,5-8 We note that a guideline statement of the American Association of Clinical Endocrinology/American College of Endocrinology⁹ and an article by a recognized expert¹⁰ both define PBF cutoff points of 25% in men and 35% in women for obesity. One of these would probably be a better reference to use, although we admit that neither publication provides any rationale. Incidentally, these cutoff points are close to the means for PBF in the 13,601 adult participants in the Third National Health and Nutrition Examination Survey (NHANES III), which are 24.8% for men and 36.7% for women.⁷

The major contribution of the 1995 WHO Technical Report⁴ was to define the normal range of body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) in adults as between 18.5 and 24.9 (with additional thresholds at 30 and 40), later elevated to official standards of both the WHO¹¹ and the National Institutes of Health.¹² The consensus on these numbers provided a foundation for Gallagher et al¹³ to propose PBF cutoffs as the empirical age-, sex-, and race-specific PBF correlates of the now canonical BMI thresholds. According to Gallagher et al, a BMI between 25 and 29.9 corresponds to a PBF of 20% to 25% in men and of 32% to 38% in women, generally allowing for a higher PBF with advancing age and in Asians compared with others¹³; the studies of Romero-Corral et al⁷ and of Jackson et al¹⁴ provide PBF correlates of a BMI of 25 in the same range as Gallagher et al. We used the thresholds of Gallagher et al in our study of patients with chronic heart failure¹ as an example of the obesity paradox, ie, the observation that in some chronic conditions, a high BMI is associated with improved survival. We found that when body composition was quantified as its individual components, a high lean body mass and a low fat mass percentage were independently associated with advantageous prognostic factors; body fat thresholds are important because BMI misclassified body fatness status (in either direction) in a large proportion of our patients,¹ as also shown by Romero-Corral et al in the general population⁷ and in a cohort with coronary heart disease.⁸

A debatable aspect of the approach by Gallagher et al is the fact that it allows for a higher degree of obesity in Asians and the elderly. Such group differences are to be expected when a uniform BMI threshold is applied to groups that differ in their relation between PBF and BMI. The finding that health risks are evident at a lower BMI in Asians than in other populations¹⁴ begs the question of whether higher PBF cutoffs are indeed appropriate in this group. Only prospective studies of individuals in whom PBF has been measured can ascertain whether Asians and the elderly are particularly tolerant of a high PBF. Nevertheless, the Gallagher et al adjustments for demographics and age are small and do not detract from the soundness of the basic principle.

Using universal PBF cutoffs points of 25% in men and 35% in women, we have found that the obesity paradox in patients with coronary heart disease extends not only to BMI but also to PBF,⁵ thereby advancing the understanding of this phenomenon. In another study,¹⁶ we have defined normal weight obesity among 6171 individuals whose BMI was in the normal range (18.5-24.9) as those whose PBF was in the highest tertile, ie, greater than 23.1% in men and greater than 33.3% in women. Normal weight obesity was associated with a high prevalence of metabolic syndrome, similar to that observed in overweight individuals. More importantly, normal-weight obese women had more than a 2-fold increased risk of cardiovascular mortality.

In conclusion, our research has shown that PBF cutoffs in the 20% to 25% range in men and 30% to 38% in women are useful to identify individuals at risk of metabolic disease who are possibly “misclassified” by BMI and to provide insights into the obesity paradox as it applies to various conditions. It has not been within the scope of our research to determine whether a hypothetical “elbow” exists on the risk curve, to define actionable trigger points for clinical recommendations, or to examine how any of these might vary by age or ethnicity. We reiterate our call⁶ for research and guidelines to establish evidence-based cutoff points for PBF, as was done years ago for BMI.