FP088COMBINATION OF URINE PH AND FEMALE GENDER FOR SCREENING PRIMARY ALDOSTERONISM IN NEWLY DIAGNOSED HYPERTENSIVE PATIENTS

Abstract

Introduction and Aims: Hypokalemia is not a sensitive marker of primary aldosteronism (PA). The method for screening PA by routine clinical examinations has not yet been established. We examined whether combination of clinical features of PA that had been reported is useful for screening PA in newly diagnosed hypertensive patients. Methods: DESIGN and SETTING and PARTICIPANTS: We retrospectively reviewed medical records of consecutive adult patients who visited a single general hospital in Japan from April 2012 to November 2014 for newly diagnosed hypertension. Almost all hypertensive patients tested plasma aldosterone concentration to plasma renin activity ratio (ARR) at the first visit. Exclusion criteria were age of 75 or older, no record of ARR measurement, previous diagnosis of secondary hypertension, treatment with steroids, NSAIDs, Chinese herbal medicines or uric acid (UA)-lowering drugs and ARR data previously measured in other hospitals. MAIN OUTCOME MEASURES: Main outcome was diagnosis of PA and it was defined in accordance with the guideline of the Japan Endocrine Society: ARR over 20 and one or more positive results in captopril challenge test, furosemide-upright test, saline infusion test and/or ACTH stimulation test. We defined potential predictors of PA and their cut-off values based on literatures and clinical viewpoints: hypokalemia as K+ below 3.5 mEq/L, metabolic alkalosis as Na + minus Cl- 40 or above, high urine pH (u-pH) as 7 or above, low serum UA as below 4.0 mg/dL, moderate to severe hypertension, female gender and age under 40 years old. Results: Among 130 patients, 24 patients were diagnosed as PA and 106 patients were classified into essential hypertension (eHT). PA group had significantly higher u-pH (6.8 +/- 0.8 vs 6.3 +/-0.7, p < 0.01) and higher proportion of female (46 % vs 21 %, p = 0.01) compared with eHT group. Multivariable logistic regression analysis by using all potential predictors indicated that high u-pH was independently associated with PA (OR 3.7, 95%CI 1.3 to 10.0) and the area under the ROC curve (AUC) of 0.73 (95%CI 0.61 to 0.85). To evaluate the influence of each independent variables on AUC, we assessed AUC when each variable was removed from the logistic regression. Removing high u-pH, female gender and hypokalemia from the regression decreased AUC by 0.059, 0.035 and 0.0011, respectively. Removing other variables led to less changes on AUC. When both high u-pH and female gender were removed, AUC was decreased by 0.11 (95%CI 0.0025 to 0.22, p=0.045). Using a scoring model in which one point was provided for each of high u-pH and female gender, AUC of this model was 0.69 (95% CI 0.56 to 0.81), whereas AUC of hypokalemia was 0.55 (95%CI 0.46 to 0.65). The prevalences of PA in patients with 0, 1 and 2 points in the scoring model were 11, 17 and 53 %, respectively. Conclusions: Combination of high u-pH and female is a simple and useful predictor for screening PA in newly diagnosed hypertensive patients.