BACKGROUND: True resistant hypertension (RH) is defined as uncontrolled office and ambulatory blood pressure (BP) in spite of an optimal regimen with at least three antihypertensive drugs. The aim of this study is to identify, in the office, clinical, laboratory, electrocardiographic, and echocardiographic variables associated with the occurrence of true RH. METHODS: These variables were recorded in a cross-sectional study involving 497 resistant hypertensive patients diagnosed by ambulatory BP monitoring as true RH (63.0%) or white coat RH (37.0%). Statistical analysis included bivariate and multivariate logistic regression. RESULTS: In bivariate analysis, true RH patients were younger, more frequently men, and had significantly higher office BP than white coat RH patients. They also had higher prevalence of physical inactivity, heart failure, and retinopathy, higher fasting glycemia, 24-h proteinuria and albuminuria, and lower serum potassium. In addition, these patients had higher electrocardiographic Sokolow and Cornell voltages and echocardiographic left ventricular mass index and hypertrophy. In multivariate logistic regression the variables best associated with true RH were male sex (P = .026), office systolic BP > or =180 mm Hg (P = .016), fasting glycemia > or =7.0 mmol/L (P = .042), serum potassium <4.5 mmol/L (P = .037), abnormal microalbuminuria (P < .001), adjusted Cornell voltage > or =2.6 mV (P = .002), and echocardiographic left ventricular hypertrophy (P = .009). In an alternative simpler model, proteinuria substituted microalbuminuria and echocardiographic data was excluded. Both predictive models have areas under receiver operating characteristic curve of 0.70. CONCLUSIONS: True RH can be recognized in the office in selected RH patients. We propose a simple scoring system with these variables that can be used in clinical practice.
BACKGROUND: True resistant hypertension (RH) is defined as uncontrolled office and ambulatory blood pressure (BP) in spite of an optimal regimen with at least three antihypertensive drugs. The aim of this study is to identify, in the office, clinical, laboratory, electrocardiographic, and echocardiographic variables associated with the occurrence of true RH. METHODS: These variables were recorded in a cross-sectional study involving 497 resistant hypertensivepatients diagnosed by ambulatory BP monitoring as true RH (63.0%) or white coat RH (37.0%). Statistical analysis included bivariate and multivariate logistic regression. RESULTS: In bivariate analysis, true RH patients were younger, more frequently men, and had significantly higher office BP than white coat RH patients. They also had higher prevalence of physical inactivity, heart failure, and retinopathy, higher fasting glycemia, 24-h proteinuria and albuminuria, and lower serum potassium. In addition, these patients had higher electrocardiographic Sokolow and Cornell voltages and echocardiographic left ventricular mass index and hypertrophy. In multivariate logistic regression the variables best associated with true RH were male sex (P = .026), office systolic BP > or =180 mm Hg (P = .016), fasting glycemia > or =7.0 mmol/L (P = .042), serum potassium <4.5 mmol/L (P = .037), abnormal microalbuminuria (P < .001), adjusted Cornell voltage > or =2.6 mV (P = .002), and echocardiographic left ventricular hypertrophy (P = .009). In an alternative simpler model, proteinuria substituted microalbuminuria and echocardiographic data was excluded. Both predictive models have areas under receiver operating characteristic curve of 0.70. CONCLUSIONS: True RH can be recognized in the office in selected RH patients. We propose a simple scoring system with these variables that can be used in clinical practice.