J‐curve Relation Between Blood Pressure and Decline in Cognitive Function in Older People Living in Community, Japan

To the Editor: There is controversy about whether high blood pressure is associated with impaired cognitive function or dementia.1, 2 Recently Skoog and colleagues3 revealed an association between high blood pressure at age 70 and the development of dementia 10 to 15 years later. We show the J-curve relation between blood pressure and decline in cognitive function in older Japanese people in a 3-year follow-up study. We assessed blood pressure (BP) and Mini-Mental State Examination (MMSE) scores in older people living in Kahoku,4 a rural Japanese town, in 1992 and 1995. We studied 155 older subjects (58 men and 97 women, mean age in 1992: 78 years; range: 70 to 91 years). In 1992, the subjects were divided into four groups according to their systolic blood pressure (SBP): SBP < 125 (Group I, n = 38); SBP 125 to <150 (Group II, n = 63,); SBP 150 to <175 (Group III, n = 40); SBP > 175 (Group IV, n = 14). There were no significant differences in age and MMSE scores among the four groups in 1992. The rates of antihypertensive medications prescribed in each group were 13.5%, 47.5%, 40.0%, and 71.4%, respectively, and those did not change significantly 3 years later. Figure 1 shows MMSE scores in 1992 and 1995 and the decline in the scores during the 3 years by each blood pressure group. Groups I, II, and IV showed significant deterioration in MMSE scores during the 3-year period (Group I from 27.2 ± 3.7 to 26.3 ± 4.2; Group II 28.2 ± 2.2 to 27.6 ± 2.4; Group IV 27.9 ± 2.2 to 26.1 ± 3.8; each P < .05 by paired t test), whereas Group III showed no significant change in MMSE scores (from 27.5 ± 3.2 to 27.9 ± 2.4). The difference in MMSE scores in each group between 1992 and 1995 showed significant decline in the scores in Groups I and IV (P < .05, Fisher PLSD, ANOVA: P < .05) compared with that in Group III, revealing the J-curve relation between SBP and the decline in MMSE scores. The relationship between the decline in MMSE scores and diastolic BP also showed a tendency of J-curve relation but failed to be significant (P < .09). . MMSE scores in 1992 and 1995 and decline in the scores during 3 years by each blood pressure group. °P < .05, MMSE scores in 1995 vs in 1992 (paired t test). *P < .05, decline in MMSE compared with that in Group III (Fisher PLSD) (ANOVA: P < .05). Although the duration of follow-up was only 3 years, we revealed an association of high BP with deterioration in cognitive function, adding the J-curve relation between blood pressure and decline in cognitive function in a longitudinal study. Our results also support the findings of Meyer and colleagues,5 which showed improvement of cognition after control of SBP from 135 to 150 mm Hg but decline in cognition after lowering SBP below the level of 135 mm Hg in hypertensive patients with multi-infarct dementia. In conclusion, it is worth remembering that not only high blood pressure, but also lower blood pressure, may be associated with later decline in the cognitive function of older people.