Association between hemoglobin levels and cancer risk: An analysis using NHANES 2017–2020 data.

Abstract

e22588 Background: Hemoglobin (Hb) levels have been implicated in various malignancies. Anemia serves as an early indicator for leukemia, conversely, elevated Hb levels may be linked to a specific cancer. This study aims to evaluate the association between Hb levels and specific cancers while adjusting for demographic factors. Methods: A retrospective analysis was conducted using data from the National Health and Nutrition Examination Survey (NHANES) of the U.S. population from 2021 to 2023. Hb levels were categorized as low, normal, or high. A chi-square test was performed to evaluate the relationship between Hb categories and cancer diagnosis. Logistic regression analysis was used to calculate the odds ratio (OR) for specific cancer types based on Hb levels, adjusted for age, gender, and ethnicity, using the SAS proc survey methodology. Results: A total of 15,560 participants' data was analyzed. Chi-square analysis revealed a significant association between Hb categories and cancer diagnosis (χ² = 1233.93, p < 0.0001), where individuals with low or normal Hb levels are more likely to be diagnosed with cancer. Logistic regression indicated that higher Hb levels were associated with an increased likelihood of cancer diagnosis (OR = 1.08, p = 0.028), while age decreased the odds (OR = 0.93, p < 0.0001). The model showed a concordance of 78.9% suggesting a strong predictive accuracy. Non-Hispanic Asians exhibited higher odds of cancer diagnosis (OR = 1.48, p = 0.012) and significantly higher odds specifically for thyroid cancer (OR = 6.453, p = 0.047). Multinomial logistic regression showed Hb levels were inversely related to the odds of specific cancer types such as esophageal (OR = 0.53, p = 0.014) and mouth/tongue/lip (OR = 0.47, p = 0.028); increasing age was a significant risk factor for bladder (OR = 1.13, p = 0.02) and prostate cancer (OR = 1.05, p = 0.012) whereas it decreased the odds for ovarian (OR = 0.93, p = 0.012) and cervical cancer (OR = 0.95, p = 0.003). Lastly, female gender strongly influenced breast cancer risk (OR = 53.67, p = 0.0003). Conclusions: Our study shows the complex relationship between Hb levels and cancer diagnosis, influenced by demographic factors. Non-Hispanic Asian ethnicity association with a cancer diagnosis and thyroid cancer specifically may highlight the importance of incorporating Hb levels and demographic factors in cancer risk assessments. Further studies are required to support our results.