Algorithm for weight loss after gastric bypass surgery considering body mass index, gender, and age from the Bariatric Outcome Longitudinal Database (BOLD).

BACKGROUND: Insight into the effects of gender and age on bariatric weight loss can be disturbed by the well-known influence of initial body mass index (BMI) on excess weight loss (%EWL). Alternative metrics can be found that eliminate this influence. Their formulas can be used to construct an algorithm in which mean weight loss becomes a constant value, describing the effectiveness of the operation independent of the initial BMI. The objective of this study was to create an algorithm describing weight loss after LRYGB in search for a better outcome metric to demonstrate unequivocally the influence of patient characteristics on bariatric results.
METHODS: Nadir weight loss results of BOLD patients, grouped by gender and age (<40 yr and ≥40 yr), with ≥2 years follow-up after LRYGB and initial BMI ≥30 and<80 kg/m(2), are expressed in 26 different metrics with formula: 100%×(initial BMI - nadir BMI)/(initial BMI - a) with "reference BMI" a = 0-25 kg/m(2). For each subgroup, the "optimal reference BMI" (a) generating the smallest deviation and without significant difference in outcome between lighter and heavier patients is used to construct an algorithm (Mann-Whitney U test; P<.0002). Mean nadir relative weight loss results (b) are compared.
RESULTS: A total of 8945 patients met inclusion criteria (mean initial BMI, 47.7 kg/m(2); median age, 48 yr; 20.0% male). Both female subgroups had optimal reference BMI: a = 10 kg/m(2); both male subgroups: a = 17 kg/m(2). LRYGB effectiveness (b) was significantly higher for younger patients and for female patients. The %EWL metric rendered different significances.
CONCLUSIONS: Both genders have age-independent metrics for which nadir relative weight loss after LRYGB is not influenced by initial BMI. The resulting algorithm nadir BMI = a + (100%-b) × (initial BMI-a) consists of an inert part (a = 10-17 kg/m(2)) on which the bariatric effectiveness (b) does not act and an alterable part (initial BMI-a) on which it does. The proposed metric percentage alterable weight loss (%AWL) reduces results to constant values for bariatric effectiveness (b), facilitating research on the precise effect of patient characteristics and surgical variables on postoperative weight loss better than %EWL, a metric able to produce false conclusions. Women and younger patients had significantly more weight loss; initial BMI had no effect.