LAUSR

One of my patients, a woman in her late 80s, has required 4 drugs at high doses for many years to keep her systolic blood pressure under 160 mm Hg. Over a 3-month period earlier this year, she went from 4 drugs to 1, with her systolic blood pressure now under 140 mm Hg. She and her family are ecstatic; I, less so. It has been known for many years that decreasing blood pressure in the elderly is a bad sign, unless related to more aggressive treatment.1 Decreases in blood pressure are a component of the so-called terminal decline, a constellation of signs and symptoms including decreases in activity, weight, cognition, and psychological outlook that often occur in the year or 2 before death, particularly death in very old age.2 Of these, blood pressure is the most noticeable, because it is numerical, frequently measured, and valued by patients and physicians. Weight loss is also a bad sign.3 We have all encountered octogenarian patients proud of finally shedding that excess 20 pounds after 50 years of failed attempts. Once again, not a good sign. Delgado et al4 now report that blood pressure actually starts a downward trajectory more than 10 years before death, when examined at a population level. They conducted 2 sets of analyses using a database of electronic medical records from 674 primary care practices. First, they plotted the blood pressure of 46 634 patients in the 20 years before death. Blood pressures started decreasing approximately 14 to 18 years before death, with steeper declines in those dying in their 80s and 90s, and with the steepest declines occurring in the 2 years prior to death. In the second study, they matched by age and sex each patient who died with a patient who survived at least 9 years longer, and compared the changes in blood pressure for the 2 groups in the period 10 to 2 years before death in the dying cohort. There were declines in blood pressure in both cohorts, but they were significantly steeper in those who died. One straightforward interpretation of these findings is that many people who die in old age have been on a dying trajectory for a decade or more. Something is happening to them that is not happening in people of the same age who are not going to die over the next decade. Are there other components of that trajectory, in addition to blood pressure decreases? The entire predeath trajectory was best illustrated in a brilliant and data-dense series of analyses of the Cardiovascular Health Study by Diehr and colleagues.5 They examined changes in body mass index, cognition, self-rated health, gait speed, distance walked, and activities of daily living over time, starting at different ages, for example, changes between age 75 and 80 years. They then stratified those changes based on when the participant ultimately died, for example, at 80 or 85 years or older than 87 years. This allowed them to distinguish the effect of being 1 year older from being 1 year closer to death. For all variables, there were clear decreases in the 7 years before death, independent of age. For most of the variables, the “effect” of being a year closer to death was larger than the “effect” of being a year older. The declines in functions in the decade before death suggest that epidemiological studies that examine the association of any of those functions with mortality are susceptible to reverse causation. The observational studies that show that high blood pressure has survival value in very old age may be because more individuals with lower pressures are in terminal decline. Similar biases would tend to show positive associations with high body mass index, or more physical activity, and survival. Epidemiologic studies describe average effects across populations or subpopulations. Physicians deal with individuals. How does the former activity inform the latter? An individual does not, on average, survive to 80 years. He or she dies at a specific age, with or without a preceding decline. The changes described by Delgado et al4 and by Diehr et al5 represent the averages of many different individuals. Perhaps one of them was hit by a car at the end of a 10-mile run on his 80th birthday. Another died after a 5-year struggle with metastatic breast cancer. Many died of “old age” or “natural causes,” diagnoses that do not appear on mortality statistics but are the most parsimonious ways to describe cause of death in many old people.6 It is the individuals dying of old age, or after many years with a serious chronic disease, that I picture when I read the graphs created by Delgado et al4 and the earlier ones by Diehr et al.5 Just because declines in function in the years before death are predictable does not mean that they are inevitable. It is reasonable to expect that gait speed and physical functioning and affect could all be improved with interventions in the decade before death. On the other hand, the much steeper decline in blood pressure in the last 2 years of life may have a different interpretation. This decline is the subject of recent reports from another research group using the same primary care database.7 The blood pressure decreases are not subtle, and they are accompanied by similar unsubtle decreases in physical activity, cognition, weight, and affect.2 The challenge for physicians treating the very elderly is when to “let go,” when to stop urging more social engagement, more exercise, more food, and to realize that our patients are near death. No physician wants to clutter the last 10 to 20 months of a person’s life with irrelevant concerns and activities. Surely an individual in the last year of life has more important things to think about. The clinical picture of the terminal decline may not meet hospice criteria. Perhaps death is still a year or 2 away. But the response should be similar, with a focus on supportive care. We do not much teach or study or talk about the various trajectories before death. This was not the case a century ago, when effective treatments were few and prognostication a valued activity. My interest in this topic was stimulated many years ago by another patient—a woman in her 80s with inoperable stomach cancer—who asked me what might happen to her. I Related article Decreasing Blood Pressure in Older Patients Invited Commentary