Abstract
BACKGROUND: While numerous studies have investigated the association between alkaline phosphatase (ALP) levels and the prevalence of cancer, the epidemiological evidence regarding the relationship between ALP and specific cancer types, as well as cancer prognosis, remains inconclusive. This study aimed to fill this gap. METHODS: This study employed retrospective cohort design utilizing data from the National Health and Nutrition Examination Survey (NHANES) database, covering the period from 2003 to 2016. Mortality data up to December 31, 2017, were retrieved from the National Death Index (NDI). Weighted multivariable logistic regression and restricted cubic spline (RCS) plots were employed to explore nonlinear associations between ALP levels and cancer. Additionally, weighted Cox proportional hazards models and Kaplan-Meier (KM) curves were used to analyze the relationship between ALP levels and all-cause mortality, as well as specific-cause mortality. Subgroup analyses were conducted to assess the consistency of the results. RESULTS: The study included 21,698 participants, with an average age of 47.21 ± 0.27 years, of whom 51.39% were female and 2064 had cancer. Participants were categorized into three models (crude model to model 2) based on different adjustment variables. Fully adjusted weighted logistic regression analysis revealed positive association between ALP levels and higher cancer prevalence (OR: 1.15, 95% CI: 1.00-1.19, P = 0.002). Consistent trend was observed across ALP quartiles (Q4 vs. Q1 OR: 1.42, 95% CI: 1.19-1.70, P for trend < 0.001). The smooth curve fitting model confirmed linear increasing trend between cancer prevalence and ALP levels (P for nonlinearity = 0.390). Further analysis by cancer type indicated that ALP levels were associated with breast cancer (OR: 1.09, 95% CI: 1.05-1.14), digestive system cancers (OR: 1.06, 95% CI: 1.01-1.11), urinary system cancers (OR: 1.05, 95% CI: 1.02-1.08), bone tumors (OR: 1.01, 95% CI: 1.01-1.02), and other cancer types (OR: 1.08, 95% CI: 1.05-1.10). Additionally, the results of multivariable Cox regression analysis showed that the hazard ratio (HR) for all-cause mortality in the Q4 group was 1.81 (95% CI: 1.55-2.12), and the HR for cancer-specific mortality was 1.40 (95% CI: 1.02-1.91). However, no significant trend was observed for cardiovascular mortality (P > 0.05). The KM survival curves for all-cause mortality, cancer-specific mortality, and heart-cause mortality indicated that the mortality rate for individuals in Group Q4 was significantly higher (P log-rank < 0.05). Stratified analysis results showed no significant interaction effects for variables, except for age and LDH (P for interaction > 0.05). CONCLUSION: Our study indicates that elevated ALP levels are associated with higher prevalence of cancer. Additionally, ALP levels possess significant clinical value in predicting the risks of all-cause and cancer-specific mortality among American adults. Further prospective studies are needed to confirm these results.