Abstract Body (Do not enter title and authors here):
Background
Obesity is associated with lower mortality in individuals with established ASCVD, a phenomenon known as the obesity paradox. Malnutrition is a robust and consistent predictor of poor outcomes in nearly all clinical settings and may often coexist with obesity. It is unknown if malnutrition modifies the survival advantage attributable to obesity in this population.

Research Question
We determine whether obesity and malnutrition independently predict all-cause mortality in adults with ASCVD, and how their coexistence alters mortality risk.

Methods
We analyzed adults with prevalent ASCVD from the National Health and Nutritional Examination Survey 2011-2018, with survey-weighted Cox regression analyses for nationally representative estimates. Obesity was defined using ethnic specific cut-offs of BMI, and malnutrition using the prognostic nutritional index (lowest quartile <48). We used interaction analyses to examine whether malnutrition attenuated the protective association of obesity with mortality, and whether the prognostic impact of malnutrition varied by obesity status.

Results
The final weighted sample included 17.1 million adults with ASCVD, of whom 1.9 million (11.5%) were both obese and malnourished. In fully adjusted models, malnutrition and obesity were independent predictors of all-cause mortality (Figure). Malnutrition was associated with increased hazards 1.73 [1.26–2.37], while obesity was associated with lower hazards 0.69 [0.49-0.96]. The effects of obesity on mortality differed significantly by nutritional status in interaction analyses (P-int = 0.004). Malnutrition was strongly associated with mortality among obese individuals 2.57 [1.53-4.33], but not among non-obese individuals 1.42 [0.93-2.16]. Conversely, the protective association of obesity was observed only among well-nourished individuals 0.61 [0.40–0.93], but was somewhat attenuated in malnourished 0.68 [0.38-1.24]. In a combined analysis, those who were both obese and malnourished did not demonstrate the protective association observed in the obese-only group.

Conclusion
Obesity was associated with lower mortality in ASCVD, but this benefit was absent in the presence of malnutrition. Malnutrition conferred higher risk among obese individuals than among the non-obese. The obesity paradox may at least partially reflect preserved nutritional reserve. Clinical obesity assessment should incorporate objective measures of nutritional status to accurately characterize risk.