Background and aims: Clusters of metabolic abnormalities resembling phenotypes of metabolic syndrome predicted outcome in the LIFE study, independently of single risk markers, including obesity, diabetes and baseline ECG left ventricular hypertrophy (LVH). We examined whether clusters of two or more metabolic abnormalities (MetAb, including obesity, high plasma glucose without diabetes, low HDL-cholesterol) in addition to hypertension were associated to levels of ECG LVH reduction comparable to that obtained in hypertensive subjects without or with only one additional metabolic abnormality (no-MetAb). Methods and results: We studied 5558 non-diabetic participants without MetAb (2920 women) and 1235 with MetAb (751 women) from the LIFE-study cohort. MetAb was defined by reported LIFE criteria, using partition values from the ATPIII recommendations. Time-trends of Cornell voltage-duration product (CP) over 5 years was assessed using a quadratic polynomial contrast, adjusting for age, sex, prevalent cardiovascular disease and treatment arm (losartan or atenolol). At baseline, despite similar blood pressures, CP was greater in the presence than in the absence of MetAb (p < 0.0001). During follow-up, despite similar reduction of blood pressure, CP decreased less in patients with than in those without MetAb, even after adjustment for the respective baseline values (both p < 0.002). Losartan was more effective than atenolol in reducing CP independently of MetAb. Conclusions: Clusters of metabolic abnormalities resembling phenotypes of metabolic syndrome are related to greater initial ECG LVH in hypertensive patients with value of blood pressure similar to individuals without metabolic abnormalities, and are associated with less reduction of ECG LVH during antihypertensive therapy, potentially contributing to the reported adverse prognosis of metabolic syndrome.