Risk of Parkinson's disease after anaesthesia and surgery

Editor—We previously examined the association between the exposure of surgery and anaesthesia for appendicitis and the subsequent development of Alzheimer's disease and related dementias (ADRD).^1,2 We chose to focus on appendicectomy because appendicitis appears to occur fairly randomly (acting as a natural experiment), and its sequelae are generally not long term, unlike other procedures possibly needed as a result of underlying illnesses that may affect long-term outcomes. Literature has suggested that there are common pathways of neurodegeneration in Parkinson's disease (PD) and ADRD.3, 4, 5 We used the same cohort of patients who received an appendicectomy for appendicitis with similarly matched patients who did not have an appendicectomy, to investigate an association between exposure to surgery and anaesthesia and PD. This study was approved by the Children's Hospital of Philadelphia Institutional Review Board.

Covariate balance after matching is shown in Supplementary Appendix 1, Section 1. The details of our matching techniques were published previously.¹ Briefly, for each of 55,000 older patients undergoing an appendicectomy for acute appendicitis, we matched five controls on 211 covariates reflecting health before the appendicectomy and followed these patients for up to 15 yr to determine if exposure to appendicectomy was associated with an increased rate of PD. Previous studies have suggested a 0.1–0.3% annual incidence rate of PD among older individuals in the USA (Supplementary Appendix 1, Section 7). This study had >80% power to detect a 1.1 hazard ratio (HR) for PD in treated (appendicectomy) patients vs matched controls with a two-sided alpha error of 5% (Supplementary Appendix, Section 5).

For our main outcome, we used diagnostic codes for primary PD (ICD9: 3320 and ICD10: G20). We defined a patient to have PD if PD codes were present in the claims separated by at least 1 yr or if one PD code from a neurologist was observed. The date of onset of PD was the date of the first code (Supplementary Appendix 1, Section 2 for details of the PD definition validation and related approach).² We looked at two outcomes and their combination. The first outcome was the time to PD, measured in terms of the hazard rate of PD. Some people are spared PD because they die before developing PD. The second outcome was death without PD. The composite outcome was the time to PD or death. For each outcome, we compared treated and control groups using Cox's stratified proportional hazards model.

In treated vs controls patients, the HR was 0.97 (0.95–0.98), P=0.0002 for PD or death. For death, the HR was 0.97 (0.95–0.99), P=0.0022. The HR for PD was 0.92 (0.86–0.98), P=0.0064 (Table 1). After appendicectomy, the 7.5-yr rates for alive without PD was 75.9% (vs 75.3% in controls, P=0.0088); for being alive the rates were 77.3% (vs 76.9% in controls, P=0.0465); and for PD the rates were 1.9% (vs 2.2% in controls, P=0.0045) (Table 1).

Table 1.

Hazard ratios between those treated with appendicectomy (N=54,996) and controls (N=274,980), and outcomes at 5, 7.5, and 10 yr.

Cox model results∗
	Hazard ratio†	(95% CI)	P-value
Parkinson’s disease or death	0.967	(0.949–0.984)	0.0002
Death	0.972	(0.954–0.990)	0.0022
Parkinson’s disease	0.916	(0.861–0.976)	0.0064
Rates of outcomes‡
At 5 yr
	Appendix treated N=47,876	Matched controls N=237,431	P-value
Alive w/o Parkinson’s disease (%)	85.51	85.30	0.2102
Alive (%)	86.53	86.40	0.4085
Parkinson’s disease (%)	1.21	1.33	0.0428
At 7.5 yr
	Appendix treated N=36,361	Matched controls N=179,841	P-value
Alive w/o Parkinson’s disease (%)	75.92	75.31	0.0088
Alive (%)	77.31	76.87	0.0465
Parkinson’s disease (%)	1.92	2.16	0.0045
At 10 yr
	Appendix treated N=25,098	Matched controls N=124,285	P-value
Alive w/o Parkinson’s disease (%)	65.09	63.79	0.0001
Alive (%)	66.72	65.48	0.0001
Parkinson’s disease (%)	2.73	2.92	0.1206

CI, confidence interval; HMO, Health Maintenance Organization; w/o, without.

For the yearly 5-, 7.5-, and 10-yr rates, the outcomes are defined as rates for all patients who did not go into HMO or lose Part A or B or reaching the end of 2017 data before 5, 7.5, and 10 yr.

^∗In the Cox model output (first three columns) the hazards were estimated before going to HMO or losing Part A or B, or before reaching the end of 2017 data.

^†The hazard ratio represents the Cox model stratified by the matched sets, with the reference group non-appendicectomy or untreated group. P-value reflects the significance of the hazard ratio coefficient from the Cox model.

^‡These rates reflect the status at 5, 7.5, and 10 yr (they are the weighted estimates of rates, with each matched set receiving weights equal to the number of eligible appendicectomy patients in that set). N at top of each column provides the numbers still eligible for analysis after 5, 7.5, and 10 yr. P-values reflect results from performing the Mantel-Haenszel test stratified by matched set.

We examined six subgroups: three based on patient characteristics (sex, age, and race), and three based on patient presentation (history of colon disease, appendicitis presentation, and the period of the appendicectomy procedure). The hazard for PD was not higher for appendicitis patients than controls in any of these six subgroup analyses (Supplementary Appendix 1, Section 3). A formal sensitivity analysis⁶ (Supplementary Appendix 1, Section 4) showed that slightly lower rates of PD in the appendicectomy group should not be interpreted as a protective effect of anaesthesia.

Some investigators have theorised that the appendix is a potential site for initiation of PD pathology because it contains an abundance of α-synuclein relative to other areas of the gastrointestinal system.⁷ Our study did not confirm this hypothesis. To date, empirical studies have been inconsistent, and two large meta-analyses failed to find a beneficial effect on lowering PD rates after appendicectomy.^8,9 There are a few known factors such as age, sex, race, and diseases of the colon that are possibly associated with appendicitis. We matched closely on such factors thereby removing potential bias if these same factors were associated with PD.

The results of our study are also relevant for procedures other than appendicectomy that resemble appendicectomy concerning surgery and anaesthesia. The type of anaesthesia used for appendicectomy in the USA is almost exclusively general anaesthesia (Supplementary Appendix 1, Section 6). Based on our previous work, anaesthesia time for appendicectomy reflected in the anaesthesia bill was ∼90 min.¹⁰ We have also found that a herniorrhaphy takes ∼100 min and a laparoscopic cholecystectomy ∼112 min.¹⁰ Therefore, in terms of anaesthesia exposure time, our findings for appendicectomy should also apply to those procedures. However, the pathophysiology of the diseases that are indications for herniorrhaphy and cholecystectomy are different than appendicectomy, which could impact differentially on the outcomes of such comparisons.

Despite this study utilising a natural experiment based on the randomness of appendicitis, we cannot be certain that appendicitis is perfectly random, so the possibility remains that there is an imbalance in an unmeasured covariate that was not controlled by matching. To the extent that the benefits and harms from appendicectomy are representative of other forms of surgery, the results of this study should reassure older patients without a past diagnosis of PD, cognitive impairment, or other neurodegenerative diseases when needing surgery that requires general anaesthesia.

In conclusion, in this study of patients who underwent general anaesthesia for appendicectomy, we did not identify an increased risk of subsequent Parkinson's disease. However, further research on this question is still needed.

Declaration of interest

The authors declare that they have no conflicts of interest.

Funding

US National Institute on Aging (Bethesda, MD, USA, RF1AG055390).

Appendix A. Supplementary data

The following is the Supplementary data to this article: