We undertook descriptive epidemiology of spot urine sodium-to-potassium ratio (Na/K) in a population sample to clarify the close relationship between Na/K and blood pressure level independently of potential confounding factors.
Study participants consisted of 9144 apparently healthy citizens (aged 54 ± 13 years). All clinical parameters were obtained at baseline.
Na/K was significantly higher in hypertensive individuals irrespective of antihypertensive medication status (normotension, 3.12 ± 1.82; untreated hypertension 3.50 ± 1.96; treated hypertension, 3.72 ± 2.53). As urinary Na (β = 0.092, P < 0.001) and K (β = −0.050, P < 0.001) levels were inversely associated with BP, Na/K (β = 0.118, P < 0.001) was more closely associated with BP than Na or K alone, as well as daily salt intake estimated from urinary Na (β = 0.088, P < 0.001). Several factors were significantly associated with Na/K, namely age, sex, obesity, blood pressure, renal function, salt restriction status, serum phosphate and urinary creatinine level, and fasting period and season at urine sample collection. However, the association between Na/K and BP was independent of these factors (adjusted β = 0.112, P < 0.001). No direct association was observed between Na/K and large arterial remodeling assessed by pulse wave analysis (P = 0.496) or retinal arteriolar morphological change (P = 0.431). Further, a genome-wide association study failed to identify any particular genotype influencing urinary Na and K levels.
Although we clarified several factors that might affect spot urine Na/K, these relationships were not substantial enough to confound the association between urinary Na/K and BP. A simple measure of Na/K might be more representative of salt loading obtained from spot urine samples than Na excretion alone.