J Am Geriatr Soc. 2016;64(3):602-607.
AUTHORS: Erin J. Aiello Bowles, MPH et al
Objectives: To evaluate the associations between anesthesia and dementia or Alzheimer’s disease (AD) risk using prospectively collected data.
Design: Cohort study.
Participants: Community-dwelling members of the Adult Changes in Thought cohort aged 65 and older and free of dementia at baseline (N = 3,988).
Measurements: Participants self-reported all prior surgical procedures with general or neuraxial (spinal or epidural) anesthesia at baseline and reported new procedures every 2 years. People undergoing high-risk surgery with general anesthesia, other surgery with general anesthesia, and other surgery with neuraxial anesthesia exposures were compared with those with no surgery and no anesthesia. Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for dementia and AD associated with time-varying lifetime and recent (past 5 years) anesthesia exposures.
Results: At baseline, 254 (6%) people reported never having anesthesia; 248 (6%) had had one or more high-risk surgeries with general anesthesia, 3,363 (84%) had had one or more other surgeries with general anesthesia, and 123 (3%) had had one or more surgeries with neuraxial anesthesia. High-risk surgery with general anesthesia was not associated with greater risk of dementia (HR = 0.86, 95% CI = 0.58–1.28) or AD (HR = 0.95, 95% CI = 0.61–1.49) than no history of anesthesia. People with any history of other surgery with general anesthesia had a lower risk of dementia (HR = 0.63, 95% CI = 0.46–0.85) and AD (HR = 0.65, 95% CI = 0.46–0.93) than people with no history of anesthesia. There was no association between recent anesthesia exposure and dementia or AD.
Conclusion: Anesthesia exposure was not associated with of dementia or AD in older adults.
Approximately 9.5 million people aged 65 and older undergo surgery with anesthesia every year in the United States. One troubling complication is postoperative cognitive decline (POCD), which may occur in up to 80% of individuals after cardiac surgery and 26% after noncardiac surgery. There is a perception that POCD may increase risk of dementia and Alzheimer’s disease (AD). Animal and molecular studies support this perception by demonstrating associations between general anesthesia exposure and AD pathogenesis.[4–9] Surgery itself may also contribute to POCD and subsequent dementia or AD risk. High-risk surgery such as cardiac surgery may increase risk of postoperative delirium, although the exact mechanism is unclear.[10,11]
Human studies of the association between anesthesia and dementia have not consistently confirmed findings from animal models. A meta-analysis of 15 studies found no association between general anesthesia exposure and risk of AD (pooled odds ratio (OR) = 1.05, 95% confidence interval (CI) = 0.93, 1.19), but prior studies have used case–control designs, which are prone to limitations, including recall bias and the use of non-population-based controls.[2,12,13] Furthermore, no studies have been able to evaluate whether the stress of surgery itself was associated with dementia or AD, in addition to anesthesia exposure. A more-recent, high-quality case–control study attempted to control biases noted in other studies and found no association between anesthesia and dementia (OR = 0.89, 95% CI = 0.73–1.10). A second population-based case–control study with 5,345 individuals with dementia and 21,380 matched controls found a greater risk of dementia in people exposed to general anesthesia through endotracheal tube intubation (OR = 1.34, 95% CI = 1.25–1.44) or intravenous or intramuscular injection (OR = 1.28, 95% CI = 1.14–1.43).
Using data from the Adult Changes in Thought (ACT) cohort, the current study evaluated associations between anesthesia exposure and incident dementia and AD using prospectively collected data. In addition, it was attempted to evaluate the associations with high-risk surgery separately from those with anesthesia.
This study cohort was drawn from members of Group Health (GH), an integrated delivery system in Washington State that provides healthcare and health insurance to approximately 600,000 members. GH’s human subjects review committee approved study procedures; all study participants provided written informed consent.
The ACT study has been described previously. Briefly, in 1994, the ACT study began enrolling randomly selected, community-dwelling adults who were members of GH and lived in or near Seattle, Washington. Participants had to be aged 65 and older and dementia free at enrollment. In-person visits occur every 2 years until the participant dies, develops dementia, or withdraws from ACT. Follow-up was ongoing as of June 2015.
The analyses for this article include data collected through September 30, 2012. Analyses were limited to people with at least one follow-up visit to ascertain possible incident dementia. The analytical sample included 3,988 people.
Self-reported surgery and anesthesia data were collected through interview at baseline and follow-up study visits. Participants were asked, “Have you had any surgery that involved a general or spinal anesthetic?” If yes, participants reported detailed information including the type of surgery, age at surgery, and type of anesthetic (general, spinal, unknown). Study interviewers did not distinguish between neuraxial anesthesia types such as epidural and spinal anesthesia, so all spinal anesthesia is referred to as neuraxial. Participants reported all prior procedures at baseline; new procedures since the prior study visit were reported at each follow-up visit. To reduce potential reporting errors, an anesthesiologist (RPP) reviewed all participant-reported procedures and anesthesia combinations. Any procedure–anesthesia combinations that were highly improbable, such as neuraxial anesthesia for coronary artery bypass grafting, were recoded assuming the reported surgery was accurate and the anesthesia was more prone to error.
Each participant’s lifetime and recent (within the past 5 years) anesthesia exposure was classified. High-risk procedures were defined as those having greater than average risk of mortality or nonfatal myocardial infarction, as determined by an anesthesiologist (RPP). High-risk procedures included skull or brain procedures, cardiovascular procedures, and liver transplants, all of which are performed under general anesthesia. Using a hierarchy to define anesthesia exposure groups, participants with any history of high-risk surgery were classified as high-risk surgery with general anesthesia. Participants with no high-risk surgery but any history of other procedures with general anesthesia were classified as other surgery with general anesthesia. Participants with only neuraxial anesthesia procedures were classified as other surgery with neuraxial anesthesia. Participants with no surgical procedures requiring anesthesia were classified as no anesthesia.
Hierarchical classification methods for lifetime and recent anesthesia exposures. Each of the three examples below illustrates the hierarchy used to classify participants’ lifetime and recent exposures to anesthesia as time-varying exposures. Each line represents the time from baseline to end of follow-up. For example, the first bar shows a person who reported no anesthesia exposure before baseline. In Year 3, he reported a general anesthesia procedure and had his exposure classified as such. In Year 7, he reported a high-risk surgery and had his exposure classified as such. His exposure remained high-risk surgery until the end of follow-up, even though he reported a general anesthesia procedure in Year 9, because high-risk surgery outweighed general surgery in the exposure hierarchy. Similarly, in Example 2, the general anesthesia procedure reported before baseline outweighed the neuraxial anesthesia procedure reported in Year 3, and the high-risk procedure in Year 7 the general anesthesia procedure outweighed. In the third example, examining recent anesthesia exposure, exposures last only for 5 years unless another procedure in the hierarchy outweighs them. So the first general anesthesia procedure exposure defined exposure status for 4 years until a high-risk surgery procedure outweighed it. The high-risk procedure defined exposure status for 5 years even though another general anesthesia procedure occurred in Year 9. Once 5 years had elapsed after the high-risk exposure, the person would be classified as exposed to the prior general anesthesia procedure for 2 more years or until the end of follow-up, whichever came first.
The number of procedures requiring general anesthesia was considered to evaluate any dose-response effect between general anesthesia and risk of incident dementia or AD. High-risk and other surgeries with general anesthesia were combined for these analyses.
Procedures used to identify incident dementia and AD have been described previously. The Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, criteria were used to define dementia and the National Institute of Neurological and Communicative Disorders and Stroke—Alzheimer’s Disease and Related Disorders Association criteria to define possible or probable AD. The onset dates for dementia and AD were defined as the date midway between the visit that triggered the evaluation leading to a positive diagnosis and the preceding study visit.
Information on demographic characteristics including age, sex, education level, and self-reported race was collected at baseline. Information on self-reported comorbidities (hypertension, diabetes mellitus, stroke, coronary heart disease, Parkinson’s disease), smoking, regular exercise (≥15 minutes, ≥3 times/week), self-rated health, and level of difficulty with activities of daily living (walking around the house, getting out of a bed or chair, feeding oneself, dressing oneself, bathing or showering oneself, getting to or using the toilet) was collected at baseline and each follow-up visit. Depression was assessed at each visit using a modified version of the Center for Epidemiological Studies Depression Scale, with a score of 10 or greater indicating depression.[19,20] Body mass index was calculated from measured height and weight at each study visit. Charlson comorbidity scores were calculated using diagnosis, procedure, and surgical codes in automated data, excluding codes for dementia. All of these characteristics were treated as time varying in the analyses.
Participants’ demographic and health characteristics are described according to lifetime exposure status at baseline. Three analyses were conducted: lifetime exposure, recent exposure (limited to the past 5 years), and lifetime cumulative exposure to general anesthesia (categorized as 0 through ≥7 procedures). Follow-up time (person-years) and number of dementia and AD outcomes, stratified according to exposure status, were calculated. Cox proportional hazards regression with participant age as the time axis was used to evaluate the associations between each time-varying exposure group and the two outcomes. For each outcome, adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated comparing each anesthesia exposure group with the reference group of no anesthesia. Individuals without a dementia diagnosis were censored at their last ACT study visit. For the AD analyses, subjects were censored at the time of dementia diagnosis if they developed dementia of a type other than AD.
All models adjusted for ACT study cohort, baseline age, sex, education, hypertension, diabetes mellitus, smoking, stroke, coronary heart disease, regular exercise, self-rated health, body mass index, depression, Parkinson’s disease, Charlson comorbidity score, and number of activities of daily living that a person had difficulty with. For the recent exposure analyses, a variable was also included in the models indicating whether a person had any history of surgery before the 5-year exposure window. Whether proportional hazards assumptions were met was assessed by evaluating plots and tests of Schoenfeld residuals; all models had P > .05, suggesting that proportional hazards assumptions were met.
Several sensitivity analyses were conducted. First, the analysis was limited to anesthesia and surgeries reported before age 65 to reduce healthy-user bias, which could occur if people with early or impending cognitive impairment were less likely to undergo surgery than those without impairment. Age 65 was chosen because this was the earliest age someone could enroll in the ACT study, and at time of study entry, they were evaluated as cognitively healthy. Second, after noting that women were more likely to report a history of anesthesia, all labor and delivery–related procedures were excluded to evaluate whether these procedures overly influenced results. Third, 22 subjects who had had high-risk surgical procedures on the brain were excluded because these procedures might be directly related to the risk of developing cognitive dysfunction. All analyses were conducted in Stata/MP 13.1 (Stata Corp., College Station, TX).
Of 3,988 participants, at baseline 248 (6%) had at least one high-risk surgery with general anesthesia, 3,363 (84%) had no history of high-risk surgery but at least one other surgery with general anesthesia, 123 (3%) had only other surgery with neuraxial anesthesia, and 254 (6%) had no history of surgery or anesthesia (Table 1). A smaller proportion of people who had other surgery with general anesthesia were male than in the other exposure groups. People who had high-risk surgery with general anesthesia had a higher prevalence of hypertension, diabetes mellitus, and coronary heart disease than the other exposure groups.
On average, people were followed for 7 years. During follow-up, 946 (24%) people were diagnosed with dementia, including 752 (19%) with AD; in addition, 26% died, 9% disenrolled, and 42% reached the current end of study follow-up before a diagnosis of dementia. The adjusted hazard for dementia or AD was not greater in people with high-risk surgery with general anesthesia than in the reference group (no anesthesia) (HR = 0.86, 95% CI = 0.58–1.28 for dementia; HR = 0.95, 95% CI = 0.61–1.49 for AD) (Table 2). The adjusted HRs for other surgery with general anesthesia and their confidence intervals were less than 1, suggesting a lower risk of dementia. None of the HRs suggested strong evidence for associations between recent anesthesia exposure and risk of dementia or AD. When comparing the results for high-risk surgery with general anesthesia directly with the results of other surgery with general anesthesia (to explore for an association with high-risk surgery in addition to that of general anesthesia), the risks of dementia (HR = 1.37, 95% CI = 1.04–1.80) and AD (HR = 1.46, 95% CI = 1.07–1.99) were statistically significantly greater in people who underwent high-risk surgery with general anesthesia than in those with general anesthesia alone. When the cumulative number of surgeries with general anesthesia (data not shown) results was evaluated, no dose response was noted.
In sensitivity analyses limited to surgeries before age 65 or excluding labor-related procedures, the associations between anesthesia exposure and dementia or AD, outcomes were attenuated toward the null. Excluding high-risk surgeries on the brain did not change the results
The primary results do not support greater risk of dementia or AD after lifetime or recent anesthesia exposure in older adults. To the knowledge of the authors, this is one of the largest population-based studies with data on anesthesia exposure to evaluate this association. These results may help quell the perception that anesthesia exposure increases risk of dementia and AD.
The results were similar to those from previous case–control studies, which have found no association between anesthesia and AD. Authors of a recent review and metaanalysis used Newcastle-Ottawa criteria to identify “high quality” case–control studies.Two of the four high-quality studies identified had nonsignificant odds ratios less than 1.0.[23,24] Recall bias in prior studies may help explain results from two prior studies that showed that anesthesia exposure increases dementia and AD risk.[9,15]
Exposure to general anesthesia within the last 5 years was not associated with subsequent risk of dementia or AD. This result should be reassuring to older adults, their families, and providers when considering surgery late in life. The authors are unaware of any other studies that have examined recent anesthesia exposure. One study limited analyses to general anesthesia exposure after age 45 and found a nonsignificant association with dementia (OR = 0.89, 95% CI = 0.73–1.10) similar to the current results for recent exposure.
This study attempted to explore whether the stress of surgery was associated with dementia or AD, in addition to general anesthesia exposure, by directly comparing high-risk surgeries with all other procedures involving general anesthesia. The risks of dementia and AD associated with high-risk surgery were statistically significantly greater than the risks associated with other surgery with general anesthesia. This result is consistent with the hypothesis that high-risk surgery adds additional stress that often leads to postoperative delirium and subsequently dementia or AD.[10,11] Additional studies are needed to confirm this observation. It is also possible that there are underlying medical conditions that contribute to the need for high-risk surgery and are also risk factors for dementia, confounding this association.
A major strength of this study is that all surgical procedures with anesthesia were recorded before the onset of dementia or AD. The outcomes of dementia and AD were also assessed prospectively and defined using standard criteria. Finally, the analysis took advantage of a large, population-based community cohort of older adults; with nearly 4,000 study participants, this is one of the largest studies on anesthesia use and dementia outcomes to date.
Limitations include that exposure data were collected through self-report and were not confirmed by medical record review. In addition, exposures collected at baseline may be subject to recall limitations. Information was not available on the specific medications used or the duration of anesthesia. Although it was attempted to disentangle surgery and anesthesia exposures, such efforts are inherently limited because of the strong correlation between the two: typically, no one receives undergo anesthesia without undergoing surgery, and no one receives surgery without anesthesia. Finally, as with all observational studies, residual confounding may have limited the results
These results suggest that anesthesia is not associated with risk of dementia or AD in older adults. Future studies should continue to explore whether there is truly greater risk of dementia and AD after high-risk surgery.
- Buie VC, Owings MF, DeFrances CJ et al. National hospital discharge survey: 2006 summary. Vital Health Stat 2010;13:1–79.
- Avidan MS, Evers AS, Planel E et al. Review of clinical evidence for persistent cognitive decline or incident dementia attributable to surgery or general anesthesia. J Alzheimers Dis 2011;24:201–216.
- Fong HK, Sands LP, Leung JM. The role of postoperative analgesia in delirium and cognitive decline in elderly patients: A systematic review. Anesth Analg 2006;102:1255–1266.
- Craddock TJ, St George M, Freedman H et al. Computational predictions of volatile anesthetic interactions with the microtubule cytoskeleton: Implications for side effects of general anesthesia. PLoS One 2012;7:e37251.
- Papon MA, Whittington RA, El-Khoury NB et al. Alzheimer’s disease and anesthesia. Front Neurosci 2011;4:272.
- Planel E, Krishnamurthy P, Miyasaka T et al. Anesthesia-induced hyperphosphorylation detaches 3-repeat tau from microtubules without affecting their stability in vivo. J Neurosci 2008;28:12798–12807.
- Planel E, Richter KE, Nolan CE et al. Anesthesia leads to tau hyperphosphorylation through inhibition of phosphatase activity by hypothermia. J Neurosci 2007;27:3090–3097.
- Whittington RA, Bretteville A, Dickler MF et al. Anesthesia and tau pathology. Prog Neuropsychopharmacol Biol Psychiatry 2013;47:147–155.
- Hussain M, Berger M, Eckenhoff RG et al. General anesthetic and the risk of dementia in elderly patients: Current insights. Clin Interv Aging 2014;9:1619–1628.
- Cerejeira J, Batista P, Nogueira V et al. The stress response to surgery and postoperative delirium: Evidence of hypothalamic-pituitary-adrenal axis hyperresponsiveness and decreased suppression of the GH/IGF-1 Axis. J Geriatr Psychiatry Neurol 2013;26:185–194.
- Deiner S, Silverstein JH. Postoperative delirium and cognitive dysfunction. Br J Anaesth 2009;103(Suppl 1):41–46.
- Seitz DP, Shah PS, Herrmann N et al. Exposure to general anesthesia and risk of Alzheimer’s disease: A systematic review and meta-analysis. BMC Geriatr 2011;11:83.
- Seitz DP, Reimer CL, Siddiqui N. A review of epidemiological evidence for general anesthesia as a risk factor for Alzheimer’s disease. Prog Neuropsychopharmacol Biol Psychiatry 2013;47:122–127.
- Sprung J, Jankowski CJ, Roberts RO et al. Anesthesia and incident dementia: A population-based, nested, case-control study. Mayo Clinic Proc 2013;88:552–561.
- Chen CW, Lin CC, Chen KB et al. Increased risk of dementia in people with previous exposure to general anesthesia: A nationwide population-based case-control study. Alzheimers Dement 2014;10:196–204.
- Kukull WA, Higdon R, Bowen JD et al. Dementia and Alzheimer disease incidence: A prospective cohort study. Arch Neurol 2002;59:1737–1746.
- American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th Ed. Washington, DC: American Psychiatric Association, 1994.
- McKhann G, Drachman D, Folstein M et al. Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology 1984;34:939–944.
- Radloff LS. The CES-D scale: A self report depression scale for research in the general population. Appl Psychol Meas 1977;1:385–401.
- Andresen EM, Malmgren JA, Carter WB et al. Screening for depression in well older adults: Evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). Am J Prev Med 1994;10:77–84.
- Charlson ME, Pompei P, Ales KL et al. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373–383.
- Wells GAS, Shea D, O’Connell D et al. The Newcastle-Ottawa Scale (NOS) for Assessing the Quality of Nonrandomised Studies in Meta-Analyses [on-line]. Available athttp://www.ohri.ca/programs/clinical_epidemiology/oxford.asp Accessed Febuary 5, 2015.
- Tyas SL, Manfreda J, Strain LA et al. Risk factors for Alzheimer’s disease: A population-based, longitudinal study in Manitoba, Canada. Int J Epidemiol 2001;30:590–597.
- Yip AG, Brayne C, Matthews FE et al. Risk factors for incident dementia in England and Wales: The Medical Research Council Cognitive Function and Ageing Study. A population-based nested case-control study. Age Ageing 2006;35:154–160.