Abstract
BACKGROUND: While numerous studies have reported associations between low dietary potassium intake and adverse clinical outcomes, methods to estimate potassium intake, mainly self-reported dietary measures and urinary potassium excretion, entail certain limitations. Self-reported measures are subject to underreporting and overreporting. Urinary potassium excretion is affected by multiple factors including renal function. Revealing the degree of bias inherent in these measures would help accurately assess potassium intake and its association with disease risk. We aim to summarize evidence on the strength of the associations between potassium intake estimated from 24-h urinary potassium excretion and potassium intake estimated from self-reported dietary measures or objective quantification methods in populations with different kidney function levels and age groups. We also aim to identify factors that affect the association strength. METHODS: We will search for potentially eligible studies that examined associations between self-reported potassium intake, 24-h urinary potassium excretion, and objectively quantified potassium intake, using MEDLINE (PubMed), Embase, Web of Science, and Scopus. Studies on children, adolescents, adults, and the elderly are eligible. Studies of patients on dialysis will be excluded. Collective study results, including a meta-analysis, will be synthesized if an adequate number of studies examining similar dietary potassium intake estimation methods are found. Analyses will be performed separately according to age groups and renal function. For the meta-analysis, fixed-effects or random-effect models will be employed depending on the degree of study heterogeneity to combine across studies the correlation coefficient, ratio, or standardized mean difference for potassium intake, comparing dietary potassium intake based on self-reported or objectively quantified methods and intake based on 24-h urinary potassium excretion. The degree of heterogeneity among included studies will be examined by calculating I(2) statistics. To investigate sources of study heterogeneity, random-effects meta-regression analyses will be performed. DISCUSSION: Revealing the strength of the association between dietary and urinary measures in populations with different levels of kidney function and age groups will enhance researchers' and clinicians' ability to interpret studies that utilize these measures and help establish a more solid evidence base for the role of potassium intake in changing chronic disease risk. Identifying factors that modify the associations between these measures may aid in developing predictive models to estimate actual potassium intake. SYSTEMATIC REVIEW REGISTRATION: PROSPERO CRD42022357847.