A recent trial presented at the ESC meeting addresses the question of reducing blood pressure meds in the elderly. It’s an important study—for numerous reasons.
There are so many trials trying to succeed in reducing polypharmacy and by this making real health outcomes better. Nearly all of them failed, but most of them did not show any harm by the patientients in whom you succeeded in reducing meds. This one is special as the patients (even so highly selected) are really very old, living in nursing homes and they are to a high degree very frail or frail.
So I disagree with this sentence of John:
"Wisdom also holds that we don’t want older patients to have stroke from severe hypertension".
In my understanding there are four types of stroke:
1.) one you only see by testing (CT/MRI) as a confounder: nobody cares in a 90 year old
2.) one which results in deficits, which can be resolved in the near future by logopedia or physio: also not so interesting
3.) one which kills you right away: in my office I know really nobody in this age not longig for such a death
4.) an disabeling one: but consider here are 40 percent already frail or very frail, some of the might have severe dementia or are bed ridden anyways.
So even in the worst case, it might not be the worst. Is a painfull femur frecture or bone metastasies due to pancreatic cancer better? If you are 90 and you live in a nursing home, you have far less than 5 years on average. Everybody wants a quick death without pain. Stroke might be one of the best choices.
Glad to read this. I work as a home health nurse doing medication setup with mostly elderly patients with many chronic health problems. I've learned to vigilantly check blood pressure readings on all of my patients taking antihypertensives and promptly report out of range trends to their prescribers.
My father, an interventional cardiologist from the times of phonocardiograms, mahogany case EKG machines with suction electrodes for the pre cordial leads, when they shaped the catheters themselves and use film that had to be developed and then played back and forth on a viewer, used to repeat a sentence to myself and my brother as we went through med school “ The patient died in electrolyte balance “. Recognizing when not to intervene is important too. and, no, we can’t help all.
So the study shows that a med reduction program didn't prolong life. Common sense tells me that it also showed that these elderly patients didn't need the meds they were taking. I would bet that they probably didn't need many of the other 6 they were taking on average.
This chimes with my personal experience. A frail elderly relative, inclined to fall over, was started on BP-lowering meds and immediately started falling over more. Not a good outcome.
Critique of John Mandrola’s Opinions on Antihypertensive Use in the Elderly
1. The fragility of the “preventive” premise
Mandrola states that antihypertensive drugs are essential in middle-aged patients so that they can reach 80 or 90 years of age, but that they lose their purpose after that point. Although appealing, this reasoning oversimplifies reality.
- Vascular risk does not disappear at 90. On the contrary, the incidence of stroke and heart failure remains high, and even a small preventive effect can mean preserving autonomy for months or years, something crucial for patients and families.
- Secondary prevention (patients with prior MI, stroke, or HF) cannot be confused with primary prevention. Many elderly patients continue to benefit from rigorous blood pressure control, not for longevity but for quality of life and reduced disability.
---
2. The overvaluation of “common sense medicine”
Mandrola repeatedly appeals to “common sense” as a guide. In geriatrics, this is dangerous:
- What seems like common sense (“less medication is always better”) can be misleading. Many octogenarians and nonagenarians, even frail ones, tolerate antihypertensives well—especially in low doses—and benefit from reduced risk of disabling stroke.
- Geriatric physiology is complex: orthostatism, variations in cardiac output, renal decline. Generalizing based on “common sense” may lead to undertreatment.
---
3. Limitations of RETREAT-FRAIL that he downplays
The study is important, but Mandrola interprets it in a biased way:
- Extremely selective population: 10,000 screened to enroll 1,000. Results are hardly generalizable, especially to community-dwelling older adults (not only nursing home residents).
- Very high-mortality population: 60% dead within 3 years. Under such conditions, it was indeed unlikely to see mortality differences. The absence of difference does not mean absence of benefit in subgroups (less frail, less polymedicated).
- Already low baseline BP (113 mmHg). The potential for benefit was minimal. It cannot be extrapolated that older hypertensive patients with SBP of 150 mmHg should loosen control.
---
4. Disregard for clinical nuances
Mandrola lumps all adults “over 80–90” together, ignoring heterogeneity:
- A 90-year-old, lucid, independent, with preserved functional reserve, may want active treatment to reduce stroke risk.
- A terminally ill older person, with extreme frailty and polypharmacy, may indeed benefit from deprescribing.
The critique here is that he universalizes management, whereas geriatrics is, by definition, about individualization.
---
5. The bias against “preventive medicine in the elderly”
He argues that there are no positive trials in nonagenarians. This is true in terms of overall mortality, but it omits that:
- Trials such as **HYVET** (patients ≥ 80 years) clearly showed benefits of treating hypertension in outcomes like stroke.
- The goal in geriatrics is not only to “live longer,” but to **live better and die better**. Avoiding a disabling stroke may be more valuable than extending life by a few months.
---
6. The final message: good, but poorly substantiated
I agree with Mandrola when he emphasizes **reducing the burden of being a patient**: appointments, tests, polypharmacy. That is central in geriatric practice.
But his reasoning sounds simplistic: he takes a negative trial in an ultra-frail population and turns it into a near-universal rule for deprescribing. The practical result may be dangerous—encouraging less experienced clinicians to stop antihypertensives without proper clinical and functional assessment.
---
Critical conclusion
John Mandrola is right to draw attention to polypharmacy and to the need to reconsider preventive drugs in very frail elderly patients. However, he overreaches by using a negative study as justification for generalized conduct.
Geriatrics requires **individualization, nuance, and contextual analysis**: chronological age is not enough, and the absence of a mortality gain does not negate benefits in preventing disability. Taken literally, his discourse risks trivializing hypertension in older adults—something we know is devastating in terms of stroke and loss of autonomy.
I’ve always tried to be reductionist for myself. If diet and exercise can treat my issue, I’ll do it no other way. I appreciate the careful demedicating though, and applaud those who know enough about their care (or surround family knowledge) not to waste time on unneeded tests.
I agree with the takeaway from this study, but for different reasons.
The authors postulated testing for 25% RRR in mortality to show superiority….in a trial involving a reduction in med Rx. That seemed like an extremely unnecessary high bar.
The proposed strategy would have reduced polypharmacy in a frail elderly population, as you pointed out. That alone would have been a sufficient win in my book. I’m disappointed the trial wasn’t structured as non-inferiority for efficacy, and superiority for safety such as symptomatic hypotension and injurious falls.
In the end, this trial “failed to show a difference” in mortality….but there certainly did not seem to be a clinical outcome penalty for reducing BP meds. That is good enough for me, for the reasons you mentioned, and will be incorporated into my practice.
No mention of the many natural resources to try that come with very little risk and no blow-back. Take your blood pressure readings over a week, everyday maybe 3-4 times a day. After sitting, or moving around, or walking, in the morning, in the evening. You might have 25-30 readings and they could be over quite a large range. So which one is your blood pressure? ALL of them...not the one time reading the nurse or doctor takes in an office. Using that lone figure as your never-to-be-questioned BP is outrageous.
My wife is on a BPM and it gets scary when the pressure falls below 90 on the upper (systolic) reading. That is not safe in my view. Oh, but wait, you have other drugs to balance things out. What a freaking joke medicine has become. I am elderly and on NO meds. There is no reason to be.
Crixcyon: elderly & on NO meds is truly exceptional & something to be thankful for
Agreed the one time reading in the practitioners office is often inaccurate & shouldn't be used for determining anyone's BP drug needs.
No one on maintenance meds needs to take their BP 3-4 times a day or deal with 25-30 readings. Establish a routine (same time, same place, same circumstances) for daily BP readings. Record 4 readings at one sitting. Throw out the high & low. Average the other two. Do this for a week then average the daily averages. Now you know your BP. Do this each time you have a med change. It's a bit complicated but not truly hard.
"elderly & on NO meds is truly exceptional & something to be thankful for"
Remember that long before blood pressure medication, there were more than a few who lived to 90 and beyond: John Adams (90), Oliver Wendell Holmes Jr. (93), George Bernard Shaw (94), Pablo Casals (96), Frank Lloyd Wright (91), Sophocles (90).
Feeding pills to frail elderly is always exciting. One never knows when the next bout of aspiration will occur. As the good doctor says, food is more important than pills at that age.
With an average systolic of 113 wouldn't we say they are over medicated even if they were 40. I would find more evidence in this trial if the average systolic were 140. So the question is do you treat an 85 year old in good health if their systolic is 140? Sprint says you lower mortality if you do. I would guess that you do not treat , but we don't have evidence for that conclusion.
This column reminds me of a recent story about a possible link between overmedication and fall-induced mortality in the elderly: Paula Span, "Why Are More Older People Dying After Falls?" The New York Times, 7 September 2025, https://www.nytimes.com/2025/09/07/health/falls-deaths-elderly-drugs.html.
I think you have to distinguish between 90 year olds who are living in a nursing home and those who are out living independently in the community. The former on average will have 2 more years and the latter who knows, but in any case will not have the same degree of frailty. Doctors have to keep cautious in every patient taking anti-hypertensives which are the underlying cause of orthostatic hypotension more than 90% of the time. We don't ask patients to have follow up visits for no reason.
"The former on average will have 2 more years AND THE LATTER WHO KNOWS"
A 90-year-old can expect, on average, about 3.9 more years (men) and 4.6 more years (women) (Source: Social Security) . . . so the average life expectancy for a 90 year-old "living independently in the community" is only a little longer--3.9 to 4.6 years--than for one in a nursing home (2 years).
agree
I strongly agree with you. I think there are too many people that are made patients by us while they shouldn't.
I also like saying:
There are only two types of death: with and without doctor's help.
There are so many trials trying to succeed in reducing polypharmacy and by this making real health outcomes better. Nearly all of them failed, but most of them did not show any harm by the patientients in whom you succeeded in reducing meds. This one is special as the patients (even so highly selected) are really very old, living in nursing homes and they are to a high degree very frail or frail.
So I disagree with this sentence of John:
"Wisdom also holds that we don’t want older patients to have stroke from severe hypertension".
In my understanding there are four types of stroke:
1.) one you only see by testing (CT/MRI) as a confounder: nobody cares in a 90 year old
2.) one which results in deficits, which can be resolved in the near future by logopedia or physio: also not so interesting
3.) one which kills you right away: in my office I know really nobody in this age not longig for such a death
4.) an disabeling one: but consider here are 40 percent already frail or very frail, some of the might have severe dementia or are bed ridden anyways.
So even in the worst case, it might not be the worst. Is a painfull femur frecture or bone metastasies due to pancreatic cancer better? If you are 90 and you live in a nursing home, you have far less than 5 years on average. Everybody wants a quick death without pain. Stroke might be one of the best choices.
Glad to read this. I work as a home health nurse doing medication setup with mostly elderly patients with many chronic health problems. I've learned to vigilantly check blood pressure readings on all of my patients taking antihypertensives and promptly report out of range trends to their prescribers.
The referenced article and Dr. J. Mandrola's comments encourage us not to lose sight of the possibility of therapeutic inertia in medical care.
My father, an interventional cardiologist from the times of phonocardiograms, mahogany case EKG machines with suction electrodes for the pre cordial leads, when they shaped the catheters themselves and use film that had to be developed and then played back and forth on a viewer, used to repeat a sentence to myself and my brother as we went through med school “ The patient died in electrolyte balance “. Recognizing when not to intervene is important too. and, no, we can’t help all.
So the study shows that a med reduction program didn't prolong life. Common sense tells me that it also showed that these elderly patients didn't need the meds they were taking. I would bet that they probably didn't need many of the other 6 they were taking on average.
This chimes with my personal experience. A frail elderly relative, inclined to fall over, was started on BP-lowering meds and immediately started falling over more. Not a good outcome.
Critique of John Mandrola’s Opinions on Antihypertensive Use in the Elderly
1. The fragility of the “preventive” premise
Mandrola states that antihypertensive drugs are essential in middle-aged patients so that they can reach 80 or 90 years of age, but that they lose their purpose after that point. Although appealing, this reasoning oversimplifies reality.
- Vascular risk does not disappear at 90. On the contrary, the incidence of stroke and heart failure remains high, and even a small preventive effect can mean preserving autonomy for months or years, something crucial for patients and families.
- Secondary prevention (patients with prior MI, stroke, or HF) cannot be confused with primary prevention. Many elderly patients continue to benefit from rigorous blood pressure control, not for longevity but for quality of life and reduced disability.
---
2. The overvaluation of “common sense medicine”
Mandrola repeatedly appeals to “common sense” as a guide. In geriatrics, this is dangerous:
- What seems like common sense (“less medication is always better”) can be misleading. Many octogenarians and nonagenarians, even frail ones, tolerate antihypertensives well—especially in low doses—and benefit from reduced risk of disabling stroke.
- Geriatric physiology is complex: orthostatism, variations in cardiac output, renal decline. Generalizing based on “common sense” may lead to undertreatment.
---
3. Limitations of RETREAT-FRAIL that he downplays
The study is important, but Mandrola interprets it in a biased way:
- Extremely selective population: 10,000 screened to enroll 1,000. Results are hardly generalizable, especially to community-dwelling older adults (not only nursing home residents).
- Very high-mortality population: 60% dead within 3 years. Under such conditions, it was indeed unlikely to see mortality differences. The absence of difference does not mean absence of benefit in subgroups (less frail, less polymedicated).
- Already low baseline BP (113 mmHg). The potential for benefit was minimal. It cannot be extrapolated that older hypertensive patients with SBP of 150 mmHg should loosen control.
---
4. Disregard for clinical nuances
Mandrola lumps all adults “over 80–90” together, ignoring heterogeneity:
- A 90-year-old, lucid, independent, with preserved functional reserve, may want active treatment to reduce stroke risk.
- A terminally ill older person, with extreme frailty and polypharmacy, may indeed benefit from deprescribing.
The critique here is that he universalizes management, whereas geriatrics is, by definition, about individualization.
---
5. The bias against “preventive medicine in the elderly”
He argues that there are no positive trials in nonagenarians. This is true in terms of overall mortality, but it omits that:
- Trials such as **HYVET** (patients ≥ 80 years) clearly showed benefits of treating hypertension in outcomes like stroke.
- The goal in geriatrics is not only to “live longer,” but to **live better and die better**. Avoiding a disabling stroke may be more valuable than extending life by a few months.
---
6. The final message: good, but poorly substantiated
I agree with Mandrola when he emphasizes **reducing the burden of being a patient**: appointments, tests, polypharmacy. That is central in geriatric practice.
But his reasoning sounds simplistic: he takes a negative trial in an ultra-frail population and turns it into a near-universal rule for deprescribing. The practical result may be dangerous—encouraging less experienced clinicians to stop antihypertensives without proper clinical and functional assessment.
---
Critical conclusion
John Mandrola is right to draw attention to polypharmacy and to the need to reconsider preventive drugs in very frail elderly patients. However, he overreaches by using a negative study as justification for generalized conduct.
Geriatrics requires **individualization, nuance, and contextual analysis**: chronological age is not enough, and the absence of a mortality gain does not negate benefits in preventing disability. Taken literally, his discourse risks trivializing hypertension in older adults—something we know is devastating in terms of stroke and loss of autonomy.
I’ve always tried to be reductionist for myself. If diet and exercise can treat my issue, I’ll do it no other way. I appreciate the careful demedicating though, and applaud those who know enough about their care (or surround family knowledge) not to waste time on unneeded tests.
I agree with the takeaway from this study, but for different reasons.
The authors postulated testing for 25% RRR in mortality to show superiority….in a trial involving a reduction in med Rx. That seemed like an extremely unnecessary high bar.
The proposed strategy would have reduced polypharmacy in a frail elderly population, as you pointed out. That alone would have been a sufficient win in my book. I’m disappointed the trial wasn’t structured as non-inferiority for efficacy, and superiority for safety such as symptomatic hypotension and injurious falls.
In the end, this trial “failed to show a difference” in mortality….but there certainly did not seem to be a clinical outcome penalty for reducing BP meds. That is good enough for me, for the reasons you mentioned, and will be incorporated into my practice.
No mention of the many natural resources to try that come with very little risk and no blow-back. Take your blood pressure readings over a week, everyday maybe 3-4 times a day. After sitting, or moving around, or walking, in the morning, in the evening. You might have 25-30 readings and they could be over quite a large range. So which one is your blood pressure? ALL of them...not the one time reading the nurse or doctor takes in an office. Using that lone figure as your never-to-be-questioned BP is outrageous.
My wife is on a BPM and it gets scary when the pressure falls below 90 on the upper (systolic) reading. That is not safe in my view. Oh, but wait, you have other drugs to balance things out. What a freaking joke medicine has become. I am elderly and on NO meds. There is no reason to be.
Crixcyon: elderly & on NO meds is truly exceptional & something to be thankful for
Agreed the one time reading in the practitioners office is often inaccurate & shouldn't be used for determining anyone's BP drug needs.
No one on maintenance meds needs to take their BP 3-4 times a day or deal with 25-30 readings. Establish a routine (same time, same place, same circumstances) for daily BP readings. Record 4 readings at one sitting. Throw out the high & low. Average the other two. Do this for a week then average the daily averages. Now you know your BP. Do this each time you have a med change. It's a bit complicated but not truly hard.
"elderly & on NO meds is truly exceptional & something to be thankful for"
Remember that long before blood pressure medication, there were more than a few who lived to 90 and beyond: John Adams (90), Oliver Wendell Holmes Jr. (93), George Bernard Shaw (94), Pablo Casals (96), Frank Lloyd Wright (91), Sophocles (90).
Feeding pills to frail elderly is always exciting. One never knows when the next bout of aspiration will occur. As the good doctor says, food is more important than pills at that age.
With an average systolic of 113 wouldn't we say they are over medicated even if they were 40. I would find more evidence in this trial if the average systolic were 140. So the question is do you treat an 85 year old in good health if their systolic is 140? Sprint says you lower mortality if you do. I would guess that you do not treat , but we don't have evidence for that conclusion.
This column reminds me of a recent story about a possible link between overmedication and fall-induced mortality in the elderly: Paula Span, "Why Are More Older People Dying After Falls?" The New York Times, 7 September 2025, https://www.nytimes.com/2025/09/07/health/falls-deaths-elderly-drugs.html.
I think you have to distinguish between 90 year olds who are living in a nursing home and those who are out living independently in the community. The former on average will have 2 more years and the latter who knows, but in any case will not have the same degree of frailty. Doctors have to keep cautious in every patient taking anti-hypertensives which are the underlying cause of orthostatic hypotension more than 90% of the time. We don't ask patients to have follow up visits for no reason.
"The former on average will have 2 more years AND THE LATTER WHO KNOWS"
A 90-year-old can expect, on average, about 3.9 more years (men) and 4.6 more years (women) (Source: Social Security) . . . so the average life expectancy for a 90 year-old "living independently in the community" is only a little longer--3.9 to 4.6 years--than for one in a nursing home (2 years).
Thanks for the info but I still think the general condition of the 2 groups is different relative to meds.