The J-curve in hypertension

Curr Cardiol Rep. 2003 Nov;5(6):441-52. doi: 10.1007/s11886-003-0105-1.


The J-curve debate has continued for 25 years. Dependency upon observational and retrospective studies has confused the issue; only the full publication of data from the prospective, randomized Hypertension Optimal Treatment (HOT) study has thrown genuine light on the problem. Many examples of the J-curve relationship between blood pressure and cardiovascular/noncardiovascular events are due to reverse causality, where underlying disease (eg, poor left ventricular function, poor general health, poorly compliant/stiff arteries) is the cause of both the low blood pressure and the increased risk of both cardiovascular and noncardiovascular events. The J-curve in patients with stiff arteries (wide pulse pressure) may be exacerbated by treatment. From the full publication of the HOT study database it is now reasonable to conclude that for nonischemic hypertensive subjects the therapeutic lowering of diastolic blood pressure (DBP) to the low 80s mm Hg is beneficial, but it is safe (though unproductive) to go lower. However, in the presence of coronary artery disease (limited coronary flow reserve) there is a J-curve relationship between treated DBP and myocardial infarction, but not for stroke. In such high-risk (for myocardial infarction) cases it would be prudent to avoid lowering DBP to below the low 80s mm Hg.

Publication types

  • Review

MeSH terms

  • Antihypertensive Agents / therapeutic use
  • Cardiovascular Diseases / etiology*
  • Cardiovascular Diseases / prevention & control
  • Diastole / physiology*
  • Humans
  • Hypertension / complications
  • Hypertension / drug therapy
  • Hypertension / physiopathology*


  • Antihypertensive Agents