A New Twist in the J‐Shape Curve?

In this issue of the Journal, the paper titled “A U-Shaped Association Between Home Systolic Blood Pressure and 4-Year Mortality in Community-Dwelling Older Men, ” by Okumiya and colleagues, 1 represents the latest addition to the ongoing controversy about what constitutes ideal blood pressure (BP) in older adults. As noted by the authors, there have been many studies, including hypertension treatment trials, that find “a U-shaped or J-shaped association between BP and mortality. ”1 As with most areas of research, this area contains many conflicting reports. Reasons for these differences include groups studied, variables measured (both primary and secondary, including confounders), and study design. Clearly, chart review studies of the relationship of blood pressure and mortality finding a U-shaped relationship may be explained by a biological model in which the disease gives rise to the either low or high levels of BP. However, a longitudinal model in which both low and high levels of BP seem to predict subsequent mortality reduces the plausibility of the reverse causation explanation. The paper by Okumiya et al. was the result of a longitudinal study in which a group of rural Japanese residents aged 65 and older were followed for 4 years. Specifically, the finding that older adults with systolic blood pressure (SBP) less than 125 mm Hg had a nearly twofold higher risk of all-cause mortality compared with subjects with SBP of 125 to 134 mm Hg supports a J-shape curve for systolic blood pressure. Also, the survival curve figure for men indicates that the low SBP group has the greatest mortality in the later 2 years of follow-up, whereas the highest SBP group has the most mortality in the first 2 years of follow-up. If low SBP were indicative of current illness, the exact opposite might be expected, given that we know high SBP is related to CVD mortality. These findings represent a new and potentially important contribution. Repeated measures on the factor under study, i. e. , SBP, would improve the design used in the article by Okumiya et al. If relationships exist that show a change in the factor associated with a change in the outcome, then time-invariant alternative explanations such as pre-existing disease at baseline are even less plausible as alternative explanations. Clearly, if low SBP is hypothesized to be predictive of subsequent mortality in men, then repeated measuring of SBP would improve the ability to test this hypothesis. In relatively healthy populations, the relationship between blood pressure and adverse events seems to be linear: lower blood pressure is better. However, in older populations, and in those with prevalent cardiovascular disease, there is increasing evidence that reducing diastolic blood pressure (DBP) below a level of 80 to 85 mm Hg is associated with increased risk of cardiovascular endpoints and all-cause mortality. This point is most recently elaborated in the work of Farnett et al. , who evaluated 13 studies and found a consistent J-shaped relationship for cardiac events and diastolic blood pressure. The beneficial therapeutic threshold point was 85 mm Hg. 2 Identifying a J-shaped curve in epidemiologic research is difficult for several reasons. First, low blood pressure may represent an artifact of greater comorbidity. Thus, comorbidity, rather than blood pressure per se, may explain observed increases in events associated with low blood pressure. This point of view is supported by Glynn, 3 who found no threshold effect of blood pressure among Establishment Populations for Epidemiologic Studies of the Elderly participants once early deaths were eliminated and potential confounders were controlled for. A second difficulty is related to sample size: there are few subjects, and even fewer events, in the tails of the blood pressure distribution, which makes nonlinear inferences about risk imprecise. This was demonstrated in the Hypertension Optimal Treatment (HOT) study; although the HOT investigators concluded that intensive treatment of blood pressure is “safe, ” careful examination of results shows a U-shaped curve in DBP for major cardiovascular events and cardiovascular mortality. Subjects with an achieved mean DBP of 70 mm Hg had about the same risk of cardiovascular mortality as subjects with an achieved mean DBP of 105 mm Hg. 4 Despite these findings, this study is now cited widely as supporting the “no threshold” hypothesis. 5 The SHEP study is another example cited widely by “no threshold” proponents. 3, 6 However, we have recently reanalyzed SHEP data7 and found, among subjects randomized to active treatment, that each 5 mm Hg lower achieved value in DBP increases the risk for stroke by 14%, coronary heart disease (CHD) by 8% and all cardiovascular disease (CVD) by 11%. We also found that a lowering of DBP over time was associated with an increased risk for stroke, CHD, and CVD. See also p 1415 How should geriatricians interpret these findings? There is some evidence to suggest that low blood pressure may not represent a desirable state of health in older adults. We have been trained to diagnose and treat high blood pressure, but perhaps those with low blood pressure deserve more careful monitoring. In health, renal, endocrine, and autonomic systems protect against low perfusion states in the settings of hypotension. Perhaps long-standing very low blood pressure among older persons stimulates physiological responses that contribute to increased risk of endpoints. Blood pressure is essential for survival, yet very high pressure leads to long-term morbidity and mortality. Similar arguments can be made for cholesterol and glucose. With the advent of stringent treatment guidelines, and powerful pharmacologic agents, we now seek to normalize blood pressure (as well as cholesterol and glucose). Usually these goals are stated with a “<” sign, with no explicit minimum. In practice, the treatment goals are translated to minimization rather than optimization. In some patients, however, we will overshoot the mark. Although the results of Okumiya do not reflect directly on the treatment of hypertension, findings from SHEP suggest that the risk of cardiovascular disease increases once diastolic BP is less than 70 mm Hg. Are there (other than economic) hazards associated with overzealous treatment of hypercholesterolemia or hyperglycemia? Further research is needed. STATEMENT OF OWNERSHIP MANAGEMENT ANO CIRCULATION (Required by 39 U. S. C. 3685). Publication Title: JOURNAL OF THE AMERICAN GERIATRICS SOCIETY;. Publication no.: 0002–8614. Filing Date: 10-01-99; 4. Issue Frequency: monthly;. No. of issues published annually: 12;. Annual subscription price: 161. 00. . Complete mailing address of known office of publication: 12107 Insurance Way, Hagerstown, MD 21740. . Complete mailing address of the headquarters or general business offices of the publisher: 227 East Washington Square, Philadelphia, PA 19106. . Full names and complete mailing address of publisher, editor, and managing editor: Publisher Lippincott Williams Editor: Dr. William Applegate. M. D. , MPH, Editor, Wake Forest University, School of Medicine, Winston-Salem, NC 27157–1052. Managing Editor: Elizabeth Webb, Editorial Coordinator, The University of TN, Memphis School of Medicine, Department of Preventive Medicine, 66 N. Pauline, Suite #633, Memphis, TN 38105. . Owner: American Geriatrics Society, 770 Lexington Ave. Suite 300. New York. NY 10021. . Known bondholders, rrrortgagees, and other security holders owning or holding 1 percent or more of total amount of bonds, mortgages, or other securities: None. . Purpose, function, and nonprofit statue: Has not changed during preceding 12 months. . Publication Name: JOURNAL OF THE AMERICAN GERIATRTCS SOCIETY. Issue Date for circulation data: September, 1999. Vol. 47, #9. . Extent and nature of circulation: Average number of copies each Issue during preceding 12 months: (a) Total no. copies (Net Press Run), 10, 245. (b) Paid and/or requested circulation; (1) Paid/requested outside-county mall subscriptions stated on Form 3541: 8, 717; (2) Paid in-county subscriptions: N/A; (3) Sales through dealers and carriers, street vendors and counter sales and other non-USPS paid distribution: N/A; (4) Other classes mailed through the USPS: N/A. (c) Total paid and/or requested circulation (sum of 15b (1) (2) (3) and M): 8, 717: (d) Free distribution by mail: (1) Outside-county as stated on Form 3541: 252: (2) In-county as stated on Form 3541. N/A; (3) Other classes mailed through the USPS: N/A. (e) Free distribution outside the mail (carriers or other means): N/A. (f) Total free distribution (sum of 15d and 15e). 252. (g) Total distribution (Sum of 15c and 15f) 8, 969. (h) Copies not distributed: 1, 276. (i) Total (sum of 15g and 15h), 10, 245. Percent paid and/or requested circulation (15c/15g × 100) 97%. Actual no. copies of single Issue published nearest to filing date: (a) Total No. copies (Net Press Run), 10, 125. (b) Paid and/or requested circulation: (1) Paid/requested outside-county mail subscriptions stated on Form 3541: 9, 110; (2) Paid In-county subscriptions: N/A; (3) Sales through dealers and carriers, street vendors and counter sales and other non-USPS paid distribution: N/A: ; (4) Other classes mailed through the USPS: N/A. (c) Total paid and/or requested circulation (sum of 15b (1) (2) (3) and (4): 9, 110 (d) Free distribution by mail: (1) Outside-county as stated on Form 3541: 296; (2) In-county as stated on Form 3541: N/A; (3) Other classes mailed through the USPS: N/A. (e) Free distribution outside the mail (carriers or other means): N/A. (f) Total free distribution (sum of 15d and 15e). 296. (g) Total distribution (Sum of 15c and 15f) 9, 406. (h) Copies not distributed: 719. (i) Total (sum of 15g and 15h), 10, 125. Percent paid and/or requested circulation (15c/15g × 100) 97%. . This Statement of Ownership will be printed In the December issue of this publication. . I certify that the statements made by me above are correct and complete. Carole Pippin Publisher