The most recent study on the long-term cognitive effects of anesthesia and surgery suggests that elderly people going under the knife have little reason to worry. According to a report in the November issue of the journal Anesthesiology, the rate of cognitive decline was the same in people who had surgery, were hospitalized for non-surgical illness, or in controls who avoided the hospital bed altogether. The work relied on long-term observational data from the Alzheimer’s Disease Research Center (ADRC) at Washington University in St. Louis, Missouri. However, the authors emphasize that their small study, with a heterogeneous group of subjects, must be repeated in a larger size before the field can form firm conclusions.

Short-term delirium or cognitive dysfunction following surgery is a well-known phenomenon, but beyond that, many patients, and anesthesiologists, are concerned that inhaled anesthetics might have long-term effects such as accelerating Alzheimer’s (see ARF related news story). Elderly people facing surgery, and their families, should find the data “reassuring,” said principal investigator Michael Avidan, an anesthesiologist at Washington University School of Medicine. Alex Evers, another anesthesiologist, co-designed the study. The authors discuss their work in a video on the university website.

Collecting solid data on the relationship between surgery and cognition can be tricky. “It is really hard to get appropriate controls for people who are about to undergo surgery,” Avidan noted. In addition, scientists generally have no data on a person’s rate of cognitive change before surgery. A patient might be in the process of gradual decline, but family members might not notice until after the surgery.

Avidan, Evers, and colleagues attempted to address these issues by joining forces with John Morris, director of WashU’s ADRC. Center scientists have been collecting cognitive data on volunteers aged 50 and older since 1985. Participants, some with and some without symptoms of Alzheimer’s, typically visit the center annually for a workup including a set of psychometric tests. During their relationship with the center, some people had surgery. People who remained healthy provided one control group. Those who were hospitalized for illness without surgery provided another control group—perhaps a better match for the surgical group because they had health problems. And because many people had two or more ADRC checkups before they needed surgery or hospitalizations, the study authors were able to track their rate of change in cognitive scores both before and after the event. The researchers also included in the analysis people with mild or moderate dementia, a population potentially at high risk for worsening cognition after surgery. “This important study has set new standards for the approach to peri-operative influence on cognition in the elderly,” wrote Roderic Eckenhoff, an anesthesiologist at the University of Pennsylvania in Philadelphia, in an e-mail to ARF (see full comment below).

The researchers examined data from 575 participants, calculating the slope of their cognitive scores over time before and after surgery or illness. For the control group, they calculated slopes before and after an imaginary “event,” occurring at approximately the same times as the events in the hospitalized populations. The slopes steepened after surgery or illness, indicating an increasing rate of cognitive decline. However, the same was true in the control group. The likely explanation is that, in general, elderly people in the study experienced a slightly increasing speed of cognitive decline. However, their decline after surgery or illness was probably no worse than it would have been without reason for hospitalization.

The authors acknowledged several limitations of their work. Participants had a variety of surgeries and illnesses, so it is possible that certain conditions that do affect cognition were present but went unnoticed in the heterogeneous data set. The researchers did not have data on the types of anesthetics used, so could not analyze the effect of different drugs. In addition, it remains possible that certain people are susceptible to cognitive decline after surgery. Although the researchers included those with mild dementia, they did not study people with more severe dementia, who might experience different effects after anesthesia. Genetics might also influence a person’s response to anesthesia.

A large, multicenter study is the next step, wrote Zhongcong Xie, of Massachusetts General Hospital in Boston, in an e-mail to ARF (see full comment below). Avidan and collaborators have begun to set up such a project, tapping Alzheimer’s research centers at several universities, but Eckenhoff estimates it will take at least five years to complete the work.—Amber Dance


  1. Many studies have suggested that cognitive dysfunction may occur after anesthesia and surgery. However, opposite reports exist as well. In the recent retrospective cohort study, the authors have employed an approach of obtaining multiple assessments before the surgery or major illness and included participants with early Alzheimer disease. They have reported that there is no significant difference among surgery, major illness, or control groups in terms of the decline of cognitive function. However, they did find that the participants with dementia may decline more markedly than the participants without dementia.

    These findings are important. But like all other retrospective studies, this study has several limitations, including the difficulty to find appropriate controls, as described in the manuscript. Therefore, the findings from this study and other post-operative cognitive dysfunction studies strongly suggest that there is a need to perform an adequately powered, multicenter human study to further define post-operative cognitive dysfunction.

  2. This is an important study. It sets new standards for the approach to peri-operative influences on cognition in the elderly by using established AD cohorts within the ADRC groups, and by examining and comparing both pre-surgery and post-surgery cognitive trajectories. Although this initial study could not detect a difference in these trajectories among control, illness, and surgery
    groups, the authors quite properly identified weaknesses that
    prevent it from being definitive at this point.

    These weaknesses include:

    1. Modest size: some groups had as few as 27 patients.

    2. Heterogeneity with respect to surgical procedures.

    3. A lack of detail with respect to surgery and anesthesia.

    4. Capture of procedures that occurred primarily at Wash U.
    hospitals; i.e., controls may have had surgery/illness elsewhere during
    the study period.

    5. No information on surgery or illness prior to ADRC entry.

    6. Elderly nature of all patients: vulnerable window not yet

    7. Controls are not "wild-type" in that many are likely to be
    family members and therefore at higher risk of having AD neuropathology.

    To address these limitations, the authors have begun to recruit a
    much larger cohort involving multiple ADRCs, including those at UPenn, Harvard, WashU, Columbia, and others. This effort will take much longer but should be more definitive. It still has the problem of a narrow age window. Like head trauma is known to do, it could be that anesthesia and surgery change the trajectory of the neuropathology much earlier, long before any cognitive changes are
    evident, a possibility that this study design cannot capture. To at least capture more acute changes in neuropathology, I am advocating for well-designed peri-operative biomarker and/or imaging studies. We and other groups interested in this issue will be discussing both the basic and clinical science at a small meeting in Toronto in June 2010.

    View all comments by Roderic G. Eckenhoff
  3. Your report was quite informative. While there still may be few conclusive studies to link AD to anesthesia, my personal suspicions of a related link still hold. My Mom at the age of 80, active, clear-minded, and driving, had denture repair at University at Buffalo School of Medicine by a student who for some reason choose to put my mother and another patient to sleep to fit dentures! She was so confused afterwards she couldn't remember how to get home. She said she's never been the same since, with mild confusion, and now at age 85, on meds for early-stage AD. One may default the cause to being age related, but in view of my personal knowledge of the her behavior and history, I beg to differ with the theory that there is no connection.

  4. The age-old question of general anesthesia such as clindamycin causing shock to the autonomic nervous system by blunting it is documented. The adjustment to unconsciousness and then consciousness can produce cognitive decline.


    . Shock Produced by General Anesthesia With Relation to Disturbances of the Blood and Gastrointestinal Tract. JAMA. 1903; XL(18):1206-12.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Inhaling Alzheimer’s? Hazy Picture Links Anesthesia, AD

External Citations

  1. video

Further Reading


  1. . Acceleration and persistence of neurofibrillary pathology in a mouse model of tauopathy following anesthesia. FASEB J. 2009 Aug;23(8):2595-604. PubMed.
  2. . Anesthesia-induced hyperphosphorylation detaches 3-repeat tau from microtubules without affecting their stability in vivo. J Neurosci. 2008 Nov 26;28(48):12798-807. PubMed.
  3. . The inhalation anesthetic desflurane induces caspase activation and increases amyloid beta-protein levels under hypoxic conditions. J Biol Chem. 2008 May 2;283(18):11866-75. PubMed.
  4. . Brain and behavior changes in 12-month-old Tg2576 and nontransgenic mice exposed to anesthetics. Neurobiol Aging. 2008 Jul;29(7):1002-10. PubMed.
  5. . Post-operative delirium is associated with poor cognitive outcome and dementia. Dement Geriatr Cogn Disord. 2006;21(4):221-7. PubMed.
  6. . Delirium and cognitive dysfunction in the intensive care unit. Curr Psychiatry Rep. 2007 Feb;9(1):26-34. PubMed.
  7. . Ether stress-induced Alzheimer-like tau phosphorylation in the normal mouse brain. FEBS Lett. 2007 Mar 6;581(5):891-7. PubMed.

Primary Papers

  1. . Long-term cognitive decline in older subjects was not attributable to noncardiac surgery or major illness. Anesthesiology. 2009 Nov;111(5):964-70. PubMed.