Does BMI Have Prognostic Value in Multiple Myeloma ? : A National Analysis

Introduction/Background/Significance: Obesity's relationship with cancer prognosis is often assumed to be uniformly unfavorable, but a well-documented "obesity paradox" exists in several malignancies, including reports in multiple myeloma (MM), where higher BMI is associated with improved survival. Proposed mechanisms include greater metabolic and energy reserve buffering against cachexia and catabolic stress during acute illness, though this potential benefit may be offset at higher BMI by increased cardiometabolic comorbidity burden. Our question was whether this survival benefit holds uniformly across the full weight spectrum, or whether it breaks down at its extremes.

Materials and Methods/Case Presentation/Objective: We analyzed the 2022-2023 National Inpatient Sample. Adult (≥18) hospitalizations with multiple myeloma (ICD-10-CM C90.0x) in any diagnosis position were categorized by weight status using ICD-10-CM BMI codes (Z68.1 underweight; Z68.25-29 overweight; Z68.30-34/35-39/41-45 obese class 1/2/3) and E66.x (obesity, unspecified), referenced against no BMI code documented - the largest group (~77% of the cohort), treated as its own category rather than assumed normal weight. Survey-weighted multivariable regression assessed in-hospital mortality, length of stay (LOS), total charges, mechanical ventilation, and non-home discharge among survivors, adjusted for age, admission year, and APR-DRG severity.

Results/Description/Main Outcome Measure(s): We identified 243,355 weighted MM hospitalizations (48,671 unweighted); only 23.8% carried a BMI/weight-status code. Mortality followed a U-shaped, non-linear pattern: underweight patients had higher adjusted mortality (aOR 1.26, 95% CI 1.07-1.48, p=0.004), while overweight (aOR 0.67, 95% CI 0.52-0.87, p=0.002) and obese class 1 (aOR 0.66, 95% CI 0.52-0.84, p< 0.001) patients had lower mortality; this protective association attenuated at obese class 2 (aOR 0.77, p=0.067) and disappeared at obese class 3 (aOR 0.94, p=0.62). Despite the mortality advantage at moderate BMI elevation, both extremes of the weight spectrum predicted worse discharge disposition: underweight (aOR 1.66, 95% CI 1.50-1.85) and obese class 3 (aOR 1.75, 95% CI 1.56-1.95) patients had significantly higher odds of non-home discharge among survivors (both p< 0.001), while overweight and obese class 1 did not. Total charges were higher across every BMI-coded category relative to the undocumented reference (+5.7% to +14.1%, all p< 0.05), and LOS did not differ significantly by category.

Conclusion(s).: BMI carries genuine, non-linear prognostic value in hospitalized MM patients: moderate excess weight is associated with an in-hospital mortality advantage consistent with an obesity paradox, but this benefit fades at severe obesity, and both underweight and severe obesity independently predict worse discharge disposition. A single "obesity is protective" narrative oversimplifies the data BMI's prognostic value in MM is U-shaped and outcome-specific, supporting weight-informed risk stratification. The elevated charges seen across all BMI-coded categories may partly reflect more intensive clinical documentation among patients with a recorded BMI, rather than a true cost effect of weight status. Limitations include reliance on ICD-10-CM BMI coding, present in only 23.8% of admissions, introducing potential documentation bias; cross-sectional inpatient data, which preclude long-term survival inference; and a race-adjusted sensitivity analysis restricted to the 2022 subset, as RACE is unavailable in 2023 NIS.

Trademarked Items:

References: