Home blood pressure measurement in the frail elderly

The elderly is the fastest growing segment of the population in affluent (and also some other) societies. In the United States, for example, people of at least 65 years are projected to account for a quarter of the population (https://www.census.gov/population/projections/data/national/2014/summarytables.html), and a fifth of them are projected to be at least 85 years old, by 2060. Hypertension is common among the elderly and is a major contributor to morbidity and mortality also in this population. The residual lifetime hypertension risk of middle-aged people approached 90% [1]. Indeed at the ages of 85 and 90 years, in the population representative Jerusalem Longitudinal study, 91 and 88%, respectively, were hypertensive [2,3]. Therefore, as of now, and even more so in the future, dealing with hypertension in the elderly will involve many if not all health professionals. Elderly population is very heterogeneous, some individuals are robust as middle-aged people are, and some develop increasing functional and cognitive limitations, especially after the age of 80 years [4]. Many physicians are reluctant to treat hypertension in the elderly, who are frequently treated with polypharmacy for concurrent medical conditions. By 2010, there were worries that antihypertensive treatment of the elderly may increase mortality, a notion that was supported by meta-analyses of studies on hypertension in the elderly, which found, despite reduction of stroke and cardiovascular disease, a tendency for increased all-cause mortality [5]. However, the Hypertension in the Very Elderly Trial (HYVET) study, a double-blind placebo-controlled study of antihypertensive treatment in people with hypertension aged 80 years and over, had to be terminated early because of a significant reduction of all-cause mortality as well as other important outcomes [6]. A subsequent Expert consensus document on hypertension in the elderly was somewhat ambiguous about treating the very elderly for hypertension [7]. More recently the randomized Systolic Blood Pressure Intervention Trial (SPRINT) compared an intensive (<120 mmHg) versus standard (<140 mmHg) treatment goal and included a prospectively recruited elderly subgroup of 75 years and more (average 80 years old). In this elderly cohort, a significant 27% reduction in all-cause mortality, as well as other important outcomes was found [8]. Yet, another Expert opinion group had suggested caution in treatment of frail elderly with the worries that intensive treatment may worsen both cognitive and physical health [9]. Why is there such a discussion about treatment of hypertension in the elderly, when randomized controlled trials, presumably the highest level of scientific evidence, point to such impressive outcomes benefits? Paradoxically, many, if not all, observational studies do not show that lower blood pressure (BP) is associated with better outcome. As a matter of fact many actually suggest that uncontrolled or untreated hypertensive patients fare better [2,3,10–16]. How can such contradictory findings be reconciled? One way to reconcile such apparent contradictory results was suggested by Odden et al.[17]. In a sample of the NHANES survey, these authors found that among the elderly, high BP predicted adverse outcome in fast walkers, but it did not in slower walkers, and in those unable to perform the rather trivial walking task, higher BP was associated with better outcome as in many of the observational studies. In another population not only walking speed, but also handgrip strength modified the predictive value of high BP [18]. The same authors found, in another population, that walking was not predictive, yet, dependence in performing activities of daily living (ADL) was [19]. This body of evidence (and of course additional studies) promoted the notion that frailty may modify the predictive potential of high BP in the elderly. Researchers involved in the HYVET and SPRINT (>75 years old), however, have applied an index of frailty to their data and concluded that in their populations varying levels of frailty (or physical function, if one wants) did not significantly modify the beneficial effects of antihypertensive treatment [20,21]. On this background, Murakami et al.[22] examined in the present issue of the Journal another validated measure of physical function, the Medical Outcomes Study (MOS) questionnaire, in a study of a rural Japanese population, the Ohasama study. The MOS was investigated as a modifying factor of the relationship of either a single office or home measured BP (about 26 measurements over a period of few days) with risk of stroke and mortality. The main study result was represented by the finding that neither stroke incidence nor all-cause mortality was associated with office BP both in individuals with normal and in those with impaired function according to MOS. On the other hand, elevated home SBP and DBP were significantly associated with stroke incidence after appropriate adjustments, including office BP. However, both for stroke and all-cause mortality, there was no interaction with physical function. This brings up the issue of BP response to measurement (loosely called white-coat effect), which may be exaggerated in the elderly. As an example, it averaged about 32/10 mmHg in the HYVET study [23]. Nevertheless, physical function and frailty assessed in different ways may not have been consistent modifiers of the relationship between BP and different outcomes. There may be several explanations: in the Ohasama study participants were relatively young (69 years old on average), and neither HYVET nor SPRINT (>75 years old) used the MOS. They used an index of frailty which accumulated various comorbidities and functional deficits, but not consistently the same ones in all participants. In the SPRINT, the frailty index was surprisingly higher in younger than in older participants. Also, a detailed analysis of the SPRINT findings according to a frailty index demonstrated that patients aged 70–85 years had a similar levels of frailty compared with national samples from the National Long Term Care Study and from the Survey of Health, Ageing, and Retirement in Europe data. On the other hand, at the age of 85 years, a divergence was observed, suggesting that the very old SPRINT participants were less frail and more robust [21]. However, in our 90-year olds study dependence in ADL, handgrip strength and a comorbidity index did not modify the overall lack of association of BP with mortality [3]. The HYVET participants had an unusually high DBP (around 90 mmHg), quite distinctively higher than typical population based, or even clinic based, elderly patients. Indeed in the Jerusalem Longitudinal Study, at the age of 85 years, only 9% would have been HYVET eligible [2], and about 48% of 90-year olds would have been SPRINT eligible [3]. The comorbidities among the 80-year olds were several folds higher than those of the HYVET: 270% higher prevalence of previous stroke, 1212% higher prevalence of previous myocardial infarction and 438% higher prevalence of previous heart failure [2]. Another attempt to uncover the discrepancy between the randomized and observational studies is a suggestion of reverse causality: those with normal or low BP, may have it because of apparent or silent morbidity, which in itself increases their mortality risk, and therefore, those with high BP appear to have better survival. Such an explanation is suggested by a massive UK study of electronic health records [24]. The authors found that in the last couple of years preceding death, BP consistently decreased, supporting the possibility of such reverse causality. However, in the Jerusalem Longitudinal Study, the first third of follow-up deaths were excluded from analysis (to account for such ‘reverse causality’), but this did not essentially change the results [2,3]. Another suggestion was that it may be the trend of BP with time that predicts mortality. A decreasing trend in SBP between 85 and 90 years was associated with increased mortality compared with an average SBP trend, independent of the SBP at age 90 years [25]. So, there is no clear explanation for the discrepancy between observational and randomized studies. Of course, survival bias may be, in part at least, an explanation: if you survive to an older age with a potentially fatal risk factor such as hypertension, maybe you have some sort of resilience to that risk factor. Another prominent difference between randomized controlled studies and some of the observational studies is that participants in the former have to come to research centers, whereas in some of the latter examination is performed at home. This type of selection may conceal greater differences in frailty than the different indexes of frailty and function. ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.