Clinical Nutrition Markers as Indicators of Colorectal Cancer (CRC) Disease Stage in a Public Hospital in New York City

OBJECTIVES: The interrelationships between clinical nutrition markers and colorectal cancer (CRC), including the stage at disease onset, are not well defined. We evaluate nutritional indicators as a function of CRC disease stage. We seek to estimate the nutrition markers’ relative importance in the identification of advanced disease and poor prognosis. METHODS: We utilized a retrospective cross-sectional database of CRC patients treated at two inner-city public hospitals from January 1, 2016, to June 1, 2021. The database includes demographic and routine clinical variables collected at the time of CRC diagnosis. Nutrition markers of interest included hemoglobin, albumin, and body mass index (BMI). Demographic and clinical variables of interest encompassed age, sex, race, smoking status, indication for colonoscopy (screening versus diagnostic), and continuity of care. Continuity of care was defined as having a primary care evaluation at least once within 5 years of a CRC diagnosis. Data were evaluated using one-way multivariate analysis of variance (MANOVA). RESULTS: Analysis was conducted on 259 patients (Gender: M = 119, F = 138; Race: White = 15.1%, African American/Black = 43.2%, Hispanic = 30.9%, Asian = 6.9%, Other = 4%). Data were divided into five cancer stages based upon CRC disease progression, with stages 0–2 representing cancer limited to the colonic wall, and stages 3–4 where the tumor extends beyond the colonic wall. MANOVA rejected hypothesis of equality of means between different stages (P = 0.0097). Individual one-way ANOVA for testing equality of means for different clinical stages with each outcome variable (adjusted for multiplicity of testing) produced significant results for albumin (P = 0.0099) and hemoglobin (P = 0.01), but not BMI (P = 0.06). When the mean of patients in stages 0–2 was compared to those in stages 3–4 (advanced versus non-advanced disease), significant differences were observed for albumin (P = 0.024) and hemoglobin (P = 0.017). CONCLUSIONS: Our results suggest that nutritional alterations are not observed in CRC patients until the tumor has advanced beyond the colonic wall (stages 3–4); our data were derived from a largely minority population. Future investigations should seek to replicate this work with more precise definitions of malnutrition, including more specific inflammatory markers besides albumin. FUNDING SOURCES: None.