ADT is associated with risk for Alzheimer’s and dementia in largest study to date


According to the newest and largest known analysis of data currently available, there is a small but statistically significant association between treatment with androgen deprivation therapy (ADT) and increase in risk for Alzheimer’s disease and/or dementia over time.

The full text of the new paper by Jayadevappa et al., which is based on an analysis of data from > 295,000 US men diagnosed with prostate cancer between 1996 and 2003 as recorded in the Surveillance, Epidemiology, and End Results (SEER)–Medicare linked database. We should therefore point out that this analysis is really only relevant for men who were 65 years of age or older at the time of their diagnosis.

Of the total of > 295,000 men identifiable in the SEER-Medicare database:

  • 154,049 met the study criteria.
  • 62,330 received ADT within 2 years of their diagnosis (and had an average [mean] age of 76.0 years).
  • 91,759 did not receive ADT (and had a men age of 74.3 years).

Mean follow-up for these 154,049 men was 8.3 years.

Here are the core findings of the study:

  • Risk for a diagnosis of Alzheimer’s disease was
    • 13.1 percent among all patients exposed to treatment with ADT
    • 9.4 percent among men not exposed to treatment with ADT
    • Hazard ratio (HR) = 1.14
  • Risk for a diagnosis of dementia was
    • 21.6 percent among all patient exposed to ADT
    • 15.8 percent among all patients not exposed to ADT
    • Hazard ratio = 1.20

There was also a small increase in risk for patients who had five of more doses of ADT compared to those who had only one to four doses:

  • For patients who had 1 to 4 doses of ADT
    • HR for Alzheimer’s disease was 1.19
    • HR for dementia was 1.19
  • For patients who had 5 to 8 doses of ADT
    • HR for Alzheimer’s disease was 1.28
    • HR for dementia was 1.24
  • For patients who > 8 doses of ADT
    • HR for Alzheimer’s disease was 1.24
    • HR for dementia was 1.21

However, it appears that the risk is mainly based on the first four doses of ADT.

The authors do note a number of significant limitations to the accuracy and interpretation of their data. However, their conclusion states that:

… our population-based study spanning 10 years or more following the diagnosis of prostate cancer shows that exposure to ADT was associated with increased hazard of both Alzheimer disease and dementia among elderly fee-for-service Medicare beneficiaries with prostate cancer. … Our results suggest that clinicians need to carefully weigh the long-term risks and benefits of exposure to ADT in patients with a prolonged life expectancy and stratify patients based on dementia risk prior to ADT initiation.

The “New” Prostate Cancer InfoLink would agree with that set of conclusions and add that prostate cancer patients with a family history of Alzheimer’s disease and/or dementia later in life might want to take particular note of this set of findings.

We also recommend this paper to prostate cancer support group leaders and other prostate cancer educators as one for discussion at appropriate meetings.

10 Responses

  1. Right around the same time this study was published, another report came out linking dementia to commonly prescribed anti-cholinergic drugs.

    These drugs include familiar names like warfarin, Zantac, Lasix and more. We have commented on both studies on our blog; note the caveat often cited by our Sitemaster — such large epidemiological studies are correlative, not causative.

  2. One must be very careful about drawing actionable conclusions from database series like this. I would rush to explain that association is not causation, and that database studies like this are fraught with unmeasured variables. Iif you think about it, you’ll see that the real cause is difficult to tease out. Older men are more likely to have dementia and they are more likely to have prostate cancer requiring ADT; smoking, diabetes, cardiovascular disease, lack of exercise, depression, drugs used to treat other cancers, body mass index (BMI), poor eating habits, etc., all increase as we age and they are all risk factors for dementia.

    In the latest study I’ve seen about this (the authors of the current study inexplicably did not reference it), by NIH of the Medicare database (which is much larger), they noticed a small association for both Alzheimer’s and dementia; however, after adjusting for risk factors, the association disappeared entirely. In fact, there was a very small diminution of Alzheimer’s risk. Furthermore, there was no dose effect — men on ADT for longer duration had no more dementia risk than men on ADT for shorter duration. (The up-and-down dose effect in the current study should make us suspicious.) Also, it is incumbent upon those who analyze this association to come up with a plausible cause for it.

    What we want is a higher level of evidence, ideally one that shows causation. In the only randomized clinical trial (Level 1 evidence) I’ve seen related to this, they found that hypogonadal men treated with TRT had the same risk of dementia as those who did not receive TRT. So we can be certain that replacing testosterone does not prevent dementia.

    Several observational studies have suggested a positive association (see, for example, here and here.

    Several observational studies have suggested no association (see, for example, here and here.

    There will never be definitive evidence of causation or its lack. That would require a randomized clinical trial where one group got ADT and one group didn’t. It would be unethical to withhold ADT from men with incurable prostate cancer. So each man has to decide whether the possibility of an increased risk is worth dying sooner.

  3. I am in the midst of reading The Virility Paradox by Charles Ryan, MD, a medical oncologist who used to be at UCSF but is now Director, Division of Hematology, Oncology and Transplantation, in the Department of Medicine at the University of Minnesota. I know Chuck well and very much respect his research and clinical prowess that includes being the PI for key studies leading to the approval of abiraterone acetate.

    I recently finished Chapter 6 on Testosterone & Alzheimer’s … in which Chuck argues there may well be a higher risk for Alzheimer’s from ADT. He spends many words addressing the impact of testosterone on brain physiology. But as he points out finally, ADT could result in dementia, while no ADT could result in death from prostate cancer! He concludes we have to determine what factors place men at highest risk for dementia complications.

  4. It is a not uncommon experience for patients receiving ADT to develop dementia-like symptoms which go away when ADT is stopped. So I have to wonder: how many of the men in this study who received a diagnosis of dementia were really suffering from this reversible dementia-like syndrome?

    Unfortunately the paper does not tell us how many of these men were receiving ADT at the time of their dementia/Alzheimer diagnosis, which could have some bearing on this question.

  5. Another Possible Confounding Influence in Association of ADT and Alzheimer’s/Dementia

    Thanks Sitemaster for posting this, and thanks to Rick, Allen, and Tom for astute comments.

    For those of us getting tired of playing whack-a-mole in pointing out alternate likely causes of this association, it is actually helpful that we are seeing two really big databses — SEER and Medicare — used to investigate the association.

    For support group leaders and advocates, the takeaway should be that the risk of ADT causing these conditions is very small — a handful of percentage points, and that it is quite possible that no true causal connection exists or that it is extremely small. In fact, if you look at the “confidence interval” limits — the range within which the true, known to God, percentage of increased (or decreased) risk lies, it is possible that ADT actually is protective, to a small extent. For those of us with training in statistics and experimental design, the rather large confidence limit ranges in the context of extremely large databases suggests that other influences are affecting the results. (I had 15 semester hours in statistics and experimental design in the experimental track of a rigorous undergraduate major in psychology, plus 24 additional hours of mathematics, plus related graduate work.) Age, which was not specifically examined as a primary driver in these studies, is the prime suspect, but not the only one for influences other than ADT that are affecting the results. The studies did adjust for age, but the method is not stated, and that could be critical. It is clear that when age and other variables were considered, what looked like a large influence plummeted to one that looked small and marginally significant, at best.

    The Act of Observation as a Confounder

    The physicist Heisenberg is known for his “uncertainty principle.” In simple terms, it means the act of observing can influence the behavior of what is being observed. Applied here, I strongly suspect that looking at prostate cancer patients with cognitive impairment versus those without is also looking at patients that are observed a lot (those on ADT) versus those who are not (those not on ADT), and this is related to more diagnoses of cognitive impairment. At least one paper I have read indicates there are more diagnoses of cognitive impairment in some medical settings that probably have more frequent interaction with patients (such as Medicare) than in others (such as PPOs). ADT patients are likely to be observed more frequently, that is, they have more medical interactions, because ADT therapy tends to be ongoing or at least intermittent (versus a more one-and-done, with some monitoring for non-ADT), because ADT patients tend to be older, because ADT patients have side effects that merit more medical attention, and because older patients tend to have more medical attention.

    This could be put to a test with these large databases. Other medical conditions that require more medical interventions could be a proxy for ADT visits for patients who are not on ADT. In other words, non-ADT patients with diabetes, for example, could be age-matched with ADT patients and the rates of cognitive impairment examined. Or, if resources permitted, all medical interactions could be tallied in an age-adjusting research design.

    Another method would be to tally the number of medical interventions for both ADT and non-ADT patients to see if that “simple” tally suggests that medical attention may be driving the residual small increases seen after a method of adjusting for age was employed.

  6. Thanks for all these comments. I’m not worried. I had 3 years of ADT in 2009 — 12 treatments — and at 76 feel privileged to have survived high-risk disease this long. I would like though, to read some statistics and a few of the most important papers about ADT’s effects on cognition. Any suggestions? I’m not afraid of maths.

  7. George:

    Have you already visited all the links in this thread that include live URLs to several papers? Chapter 6 of Chuck Ryan’s book on testosterone and Alzheimer’s includes 13 footnote citations.

  8. Gerorge, re papers on ADT and cognition:

    I just did a search on http://www.pubmed.gov for — prostate cancer AND (“androgen deprivation therapy” OR “hormonal therapy”) AND (cognition OR dementia OR Alzheimer’s) — and got a list of 86 papers after filtering for abstracts.

    I have read some of these abstracts or papers in the past and have noted a rather common flaw: failure to adequately control for age, which of course is strongly associated with cognitive issues. As we know, there is a strong “publish or perish” culture in the research world, with negative studies often not being published, and I believe this has pushed studies in this area toward weak study designs that yield misleading results. In some studies possible (to me extremely likely) confounding by age is not even considered as a limitation. I also mentioned another likely confounder above — the act of observation, i.e., more medical interactions for older folks which makes a diagnosis of a cognitive issue more likely. I have yet to see a paper that makes a convincing case for a true increase in Alzheimer’s/dementia with ADT as a causal factor rather than just an association.

    At least we in the prostate cancer community are reducing the risk of cognitive diagnoses by not being female, for women check two of the blocks for increased incidence of Alzheimer’s and dementia disease, which is well documented for women versus men by research: older age due to living longer, and more medical attention because that’s what women do, in comparison to us. Bravo for us guys!

  9. Rick:

    Thanks for the note.

    No, I have not but will do so.

  10. Dementia Mimicked For Some Patients Due to Anticholinergic Drugs — Another Confounding Influence for Studies Examining Association of ADT with Dementia/Alzheimer’s

    We older prostate cancer patients are not only more likely to be on ADT than other prostate cancer patients, but we are also more likely to be taking drugs for other health conditions just because we are older, and some of those drugs have effects that mimic dementia. This is another confounding factor for some studies that look into the association of ADT drugs and Alzheimer’s/dementia.

    The Washington Post had an article about anticholinergic drugs that have caused apparent dementia for some patients. The article contained these key sentences:

    “An estimated 1 in 4 older adults take anticholinergic drugs — a wide-ranging class of medications used to treat allergies, insomnia, leaky bladders, diarrhea, dizziness, motion sickness, asthma, Parkinson’s disease, chronic obstructive pulmonary disease and various psychiatric disorders.

    “Older adults are highly susceptible to negative responses to these medications. Since 2012, anticholinergics have been featured prominently on the American Geriatrics Society Beers Criteria list of medications that are potentially inappropriate for seniors. ‘The drugs that I’m most worried about in my clinic, when I need to think about what might be contributing to older patients’ memory loss or cognitive changes, are the anticholinergics,’ said Rosemary Laird, a geriatrician and medical director of the Maturing Minds Clinic at AdventHealth in Winter Park, Fla.”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: