Urinary Potassium and Kidney Function Decline in the Population—Observational Study

Abstract

Background—Some data suggest favorable effects of a high potassium intake on kidney function. The present population-based study investigated cross-sectional and longitudinal relations of urinary potassium with kidney function. Methods—Study cohort included 2027 Gubbio Study examinees (56.9% women) with age ≥ 18 years at exam-1 and with complete data on selected variables at exam-1 (1983–1985), exam-2 (1989–1992), and exam-3 (2001–2007). Urinary potassium as urinary potassium/creatinine ratio was measured in daytime spot samples at exam-1 and in overnight timed collections at exam-2. Estimated glomerular filtration rate (eGFR) was measured at all exams. Covariates in analyses included demographics, anthropometry, blood pressure, drug treatments, diabetes, smoking, alcohol intake, and urinary markers of dietary sodium and protein. Results—In multivariable regression, urinary potassium/creatinine ratio cross-sectionally related to eGFR neither at exam-1 (standardized coefficient and 95%CI = 0.020 and −0.059/0.019) nor at exam-2 (0.024 and −0.013/0.056). Exam-1 urinary potassium/creatinine ratio related to eGFR change from exam-1 to exam-2 (0.051 and 0.018/0.084). Exam-2 urinary potassium/creatinine ratio related to eGFR change from exam-2 to exam-3 (0.048 and 0.005/0.091). Mean of urinary potassium/creatinine ratio at exam-1 and exam-2 related to eGFR change from exam-1 to exam-3 (0.056 and 0.027/0.087) and to incidence of eGFR < 60 mL/min per 1.73 m2 from exam-1 to exam-3 (odds ratio and 95%CI = 0.78 and 0.61/0.98). Conclusion—In the population, urinary potassium did not relate cross-sectionally to eGFR but related to eGFR decline over time. Data support the existence of favorable effects of potassium intake on ageing-associated decline in kidney function.

Diet and Other Modifiable Factors in Long-Term Decline of Kidney Function: Observational and Population-Based Cohort Study

“…Although the influence of modifiable factors is often regarded as limited, several studies have reported that dietary habits and other modifiable factors may modulate the decline of kidney function over time [5,6]. In agreement with those studies [5,6], a lower physical activity in leisure time, a lower alcohol intake, a higher protein intake, a higher sodium intake, and a lower potassium intake in the Gubbio study population associated with a greater kidney function, decline over time independently of each other [7][8][9][10]. The Gubbio study is a prospective, longitudinal project investigating cardiovascular disease, kidney disease, and other public health objectives in a sample of the general population residing in a city of north-central Italy [11,12].…”

“…The baseline and the follow-up questionnaires included specific questions on the habitual physical activity in leisure time expressed as min/d and on the habitual intake of four types of alcohol containing beverages: wine, beer, aperitifs/cocktails, and spirits [11,12]; the total intake of alcohol as g/d was calculated as the rounded sum of the alcohol intake due to the four types of beverages using the Italian standards of volume and alcoholic gradation of each beverage [8]. Urinary levels of urea, sodium, and potassium in the baseline overnight collections were measured as indices of the dietary intake of protein [9,14,15], sodium [7,14,16], and potassium [10,14,17], respectively. Urinary levels of urea, sodium, and potassium were expressed as ratio to urine creatinine for exclusion of errors in completeness, and timing of urine collection and were analyzed separately by gender and age strata to reduce the confounding of gender and age on urinary creatinine [16][17][18].…”

“…The cumulative score was a priori conceived to obtain a single numerical variable ranging from 0 to 10 that reflected the cumulative and graded prevalence of the five modifiable factors in each examinee. To this aim, each one of the five modifiable factors was scored from 0 to 2 based on the linear relationships previously found in the Gubbio cohort between quantiles of the given factor and the eGFR change over time [7][8][9][10]. In particular, given that higher urinary urea and higher urinary sodium related to a greater eGFR decline [7,9], urinary urea and urinary sodium were scored 0 when the value was in the lowest tertile, 1 when in the intermediate tertile, and 2 when in the highest tertile (low risk, intermediate risk, and high risk, respectively); vice versa, given that lower physical activity, lower alcohol intake, and lower urinary potassium related to a greater eGFR decline [7,8,10], physical activity, alcohol intake, and urinary potassium were scored 0 when the value was in the highest tertile, 1 when in the intermediate tertile, and 2 when in the lowest tertile.…”

Relation of Alcohol Intake to Kidney Function and Mortality Observational, Population-Based, Cohort Study

“…The study adheres to the Declaration of Helsinki and includes informed consent and approval by the institutional committee (CEAS-Umbria #2850/16). The study also investigates the relationship of dietary factors with renal endpoints [ 26 , 27 , 28 , 29 , 30 ]. The study design, involvement of the invited population, response rates, and characteristics of the Gubbio study cohort are reported elsewhere [ 23 , 24 , 25 ].…”

“…Erythrocyte mean corpuscular volume and gamma-glutamyl-transpeptidase were used as objective markers for the validation of the reported alcohol intake [ 32 ]. Variables selected for the analyses were as follows: gender, age, and date of the exam; sodium/creatinine ratio and potassium/creatinine ratio in the Exam-1 daytime urine, and in the Exam-2, overnight urine used as indices of dietary intake of sodium and potassium, respectively [ 28 , 29 , 33 ]; urea nitrogen/creatinine ratio in the Exam-2 overnight urine used as an index of dietary intake of protein [ 34 ]; 24 h urinary creatinine used as an index of creatinine generation and muscle mass [ 35 ]; body mass index used as an index of overweight; blood pressure and use of antihypertensive drugs; serum total cholesterol; smoking; serum glucose and use of antidiabetic drugs; date of death. Urinary albumin/creatinine ratio was measured and included in the analyses only for those examinees aged 45–64 years at Exam-2 [ 36 ].…”