“…The relationship between the endpoints, the presence of baseline SL-, CV-, and CP-LVH and other baseline variables (age, gender SBP, DBP, heart rate, BMI, serum cholesterol, hemoglobin, creatinine and uric acid concentration, history of CVD, diabetes, antihypertensive drug treatment at baseline, smoking and alcohol use and randomization to intervention/placebo group) was determined using a Cox proportional hazard regression model. The variables selected have been associated with the presence of ECG LVH in previous studies[15,22,23]. Information was missing for serum uric acid in 17 participants, cholesterol in six, hemoglobin level in four, heart rate measurements in three and BMI in one participant.…”