Risk of Parkinson Disease Among Adults With vs Without Posttraumatic Stress Disorder

Key Points Question Is a history of posttraumatic stress disorder (PTSD) associated with an increased risk of Parkinson disease (PD)? Findings In this cohort study of 8336 patients with PTSD matched with 8336 patients without PTSD, patients with PTSD had an increased risk of PD compared with those without PTSD. Risks were higher among men older than 72 years. Meaning These findings suggest that older men with a history of PTSD should be monitored for PD.


Introduction
Parkinson disease (PD) is the second most common neurodegenerative disease, after Alzheimer disease, with a lifetime prevalence of approximately 1.5%. 1,2PD is manifested by the loss of dopamine-producing neurons in the substantia nigra, resulting in various motor and nonmotor symptoms. 3PD risk is associated with multiple environmental, behavioral, and biological 4 risk factors.
Posttraumatic stress disorder (PTSD) is a mental disorder that occurs as a result of an exposure to a traumatic event. 5Dopamine is implicated in the regulation of fear conditioning and anxiety. 6In individuals with PTSD, there is a genetic component associated with dopamine metabolism that governs whether an individual develops PTSD, as well as what symptoms they may display. 6The prevalence of traumatic exposure worldwide is 70.4% (range, 28.6%-84.6%),and Israel ranks high, with a prevalence of 74.8%. 7Consequently, it has been estimated that 9.4% of Israel's adult population has PTSD, 8 more than twice the global prevalence of 3.9%. 95][16] The association between stress (including PTSD) and PD has been hypothesized since the 1980s, 17,18 but longitudinal cohort studies using systematic data collection are scarce.A cohort study 19 in Taiwan found a 3.46-fold risk for PD in late life compare with patients without PTSD.Recently, a case-control study 20 has shown that PTSD is associated with higher odds of PD (odds ratio, 2.7) in veterans in the US.However, that study population was almost 99% male and assessed prevalent PD; thus, the measured PD risk might be biased and overestimated.In the current study, we evaluated the association between PTSD and incident PD, using data from the second largest health plan in Israel over the course of a 20-year followup period.

Study Design and Population
This retrospective cohort study was performed using data from the Maccabi Health Care Services (MHS) automatically computerized clinical database.MHS is a nationwide health plan (payer-provider) representing approximately one-quarter of the population in Israel (approximately 2.6 million people) with less than 1% per year moving out.The National Health Insurance Law passed in Israel in 1995 mandates that all legal residents be required to enroll in 1 of 4 not-for-profit health plans and guarantees free choice among them. 21The study was conducted in accordance with the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use Guideline for Good Clinical Practice, Good Epidemiology Practices 22 and Declaration of Helsinki, Ethical Principles for Medical Research Involving Human Subjects. 23A waiver of consent was approved by the Maccabi ethics committee because this retrospective database study was performed using data that were automatically collected in the MHS databases as part of routine data collection in a large health maintenance organization.All data were anonymous.
This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

MHS members born before 1970 with a PTSD diagnosis (International Classification of Diseases,
Ninth Edition [ICD-9] code 309.81) (1) given by selected specialists (ie, psychiatrists, psychologists, or neurologists), (2) extracted from a hospital discharge report, or (3) documented as a chronic diagnosis (defined as such by the primary physician), between January 1, 2000, and December 31, 2015, were identified.We required at least 1 year of continuous membership in MHS before the first PTSD diagnosis to ensure that cases were incident.Patients with PD diagnosed before the first PTSD diagnosis were excluded from the analysis.All eligible patients with PTSD were randomly matched to MHS members without PTSD with at least 1 year of membership on a 1:1 basis by exact birth year and sex.The index date was defined as the date of the first PTSD diagnosis.Patients without PTSD

Variables and Measurements
Patients (including both cardiovascular disease and stroke data), diabetes, 27 hypertension, 28 chronic kidney disease, 29 and cancer. 30Additional comorbidities (ie, depression, migraine, epilepsy, and traumatic brain injury) were defined as present when at least 3 diagnoses were documented, with at least first mention before the index date.End of follow-up was defined as the earlier of (

Statistical Analysis
Data analysis was performed from February to June 2022.Descriptive statistics are presented using frequencies and proportions for categorical variables and mean values with SD or medians with IQRs as appropriate.Standard mean differences were calculated to assess the absolute effect size for each baseline comparison.A standardized mean difference less than 0.1 was taken to indicate a negligible difference in the mean or prevalence of a covariate between exposure groups. 31Time to incident PD in patients with and without PTSD was calculated using Kaplan-Meier curves and compared using the log-rank test.The proportionality assumption was assessed using Schoenfeld residuals test and verified both by a nonsignificant P value (P = .greater than among unexposed individuals), we examined the association between PTSD and negative control outcomes, including neurological diseases (eg, glioblastoma multiform amyotrophic lateral sclerosis, Creutzfeldt-Jakob disease, and Huntington disease) that may share the same potential ascertainment bias but are not caused by PTSD.The presence of these outcomes was assessed after the index date using the same criteria as for PD.We excluded patients with an event before the index date.To maintain balanced cofounders, each excluded case was excluded with their matched control.Fifth, we also restricted the analyses to those with no brain injury history to assess the possibility of confounding due to brain injury.In addition, we performed a sensitivity analysis replacing the categorical SES with the original continuous variable.

Patients' Characteristics
We Graphs show data for entire cohort (A), men only (B), patients older than 72 years at the index date (C), Holocaust survivors (D), patients with depression at the index date (E), and patients with hypertension at the index date (F).
observed in the main analysis (overall adjusted HR, 1.40; 95% CI, 1.03-1.91)(Table 2 and eTable 1 in the Supplement).A negative control outcomes analysis found no difference in the risk among patients with and without PTSD (event rate in both groups, 0.2%; HR, 1.00; 95% CI, 0.50-1.90;P > .99)(eTables 2 and 3 in the Supplement).Restriction of the main analysis to those with no indication of brain injury history showed estimated risk similar to that found by the main analysis (overall adjusted HR, 1.54; 95% CI, 1.15-2.05)(eFigure in the Supplement).The final sensitivity analysis using SES as a continuous variable produced results almost identical to those in the main analysis (overall adjusted HR, 1.50; 95% CI, 1.13-2.00).Stratified analyses by age and sex of sensitivity analyses are presented in Table 2 and in eTable 1 in the Supplement.

Discussion
The current retrospective cohort study spanning more than 20 years of follow-up found that incident PTSD was associated with an increased risk for PD in male patients, particularly among those aged 72 years and older.These findings are consistent with the results of a recent study by White et al 20 among male veterans in the US showing a 2.7-fold excess risk for PD.The association between PTSD and PD among elderly male patients in our analysis was robust and not affected by adjustments for depression and traumatic lifetime events, such as the Holocaust and experiencing terror attacks.In addition, the results of the main analysis were solid across several sensitivity analyses, including restriction for patients with no history of brain injury.Although in our analysis the association between PTSD and PD in female patients was not significant, a 3.46-fold increased risk for PD in patients with history of PTSD compared with those without PTSD was previously reported in a predominantly female cohort. 19er the last few decades, emotional and physical stress have been linked to PD risk. 18rthermore, there might be a link between the duration of PTSD and PTSD-related neurodegeneration. 32Moreover, sleep disturbance, a common symptom of PTSD manifested as nightmares and difficulty in falling asleep and maintaining sleep, may disrupt the balance of oxidants and antioxidants toward a more oxidative environment. 32Nonetheless, the causal association between stress or anxiety and PD has not been fully established.A recent study 33 has shown that adjustment disorder, an emotional or behavioral reaction to a stressful event, was associated with

Limitations and Strengths
The current analysis has several limitations.First, data regarding the type of trauma and PTSD  39 was accepted in Israel, making patients with PTSD (among others) entitled to a package of social services.This reform might have encouraged previously symptomatic undiagnosed patients to be officially diagnosed at more late stage of the disorder.Third, despite a known female predominance of PTSD, 40 this analysis has found similar proportions of PTSD in men and women.In Israel, there is a higher proportion of men rather than women in combat units in the Israeli defense forces.In men, combat is the second leading cause of PTSD, with approximately 40% of the men experiencing combat developing PTSD. 5 In the current analysis, we included Israeli citizens born before 1970, a population who was exposed to several regional wars either in the army (age 18 years) or in the reserve services (age Ն21 years).
Additional limitations of this study are those inherent in the use of database analysis, including the use of ICD-9-Clinical Modification codes to identify patients with PTSD.It is possible that some individuals are misclassified as having as PTSD or trauma brain injury that share a similar clinical presentation.This is indicated by the high proportion of brain injury among patients with PTSD.In addition, unlike International Classification of Diseases, 11th Revision, ICD-9 provides no distinction between PTSD and complex PTSD, which is characterized by additional clusters that reflect disturbances in self-organization, such as affective dysregulation, negative self-concept, and disturbances in relationships. 41Complex PTSD was associated with sustained, repeated, or multiple forms of traumatic exposure.The specific relationship between complex PTSD and PD could not have been investigated in the current study and warrants future research.
This study has several strengths.First, to best of our knowledge, this is one of the first largescale longitudinal cohort studies examining the risk of PD in patients with PTSD.Second, owing to the long duration of data availability, we were able to include a large number of patients with PTSD and allow for at least 5 years of postexposure follow-up.Third, we were able to minimize the concern of ascertainment bias by showing no difference in other neurological diagnosis between those with and without PTSD.

Conclusions
Our study showed that PTSD diagnosed in men over the age of 72 years is associated with an increased risk of developing PD later in life.Thus, PTSD may be considered to be a factor associated

JAMA Network Open | Neurology
19; global test, P = .74)and by visual assessment of the residual plot (data not shown).Cox proportional hazards regression was used to estimate hazard ratios (HRs) and 95% CIs of PD among patients with PTSD compared with those without PTSD.Multivariable Cox proportional hazards was performed to adjust for suspected confounders measured at baseline (variable with standard mean difference >0.1 at baseline).Analysis stratification was performed by age at index date (according to the median age at index date of the PD cases, 72 years) and sex.A few sensitivity analyses were performed.First, we excluded patients (and their matched controls) with an index date before 2004, to further ensure incident cases of PTSD.Second, we excluded patients who received a diagnosis of PD in the first year after the PTSD index date (and their matched controls) to ensure that PTSD precede the development of PD.Third, we excluded patients with PD (and their matched controls) who had at least 1 purchase of antipsychotic medication in the year before their first PD diagnosis because of the motor effects of antipsychotic medications.Fourth, to address the possibility of ascertainment bias (which might be introduced when intense surveillance or screening for outcomes among individuals with a history of PTSD is JAMA Network Open | Neurology Risk of Parkinson Disease Among Adults With vs Without Posttraumatic Stress Disorder JAMA Network Open.2022;5(8):e2225445.doi:10.1001/jamanetworkopen.2022.25445( Risk of Parkinson Disease Among Adults With vs Without Posttraumatic Stress Disorder JAMA Network Open.2022;5(8):e2225445. doi:10.1001/jamanetworkopen.2022.25445(Reprinted) August 4, 2022 2/11 Downloaded From: https://jamanetwork.com/ on 09/21/2023 were given the index date of the their respective matched patient with PTSD.All patients without PTSD were PD free at the index date.

Table 2 .
36s and 95% CIs for PD in Patients With PTSD Compared With Patients Without PTSD Adjusted for age at index, sex, socioeconomic status, smoking status, Holocaust survivor and survivor of terror attack, hypertension, depression, migraine, and traumatic brain injury.Another study34has shown that anxious personality increases the risk of PD in men by 2-fold and by 1.76-fold in people aged 50 to 69 years at the time they took the Minnesota Multiphasic personality inventor test.Moreover, depression and anxiety are well recognized prodromal symptoms of PD.35PTSD may present the extreme manifestation of less-acute disorders, such as anxiety, depression, or stressful existence.The fact that PTSD in men older than 72 years was associated with increased risk of developing PD later in life in our study may indicate that PTSD in older patients may actually be a prodromal symptom of PD, especially in the light of the fact that PARK2, a PD gene involved in dopamine regulation, is associated with PTSD in men, along with other novel genes and noncoding RNAs.36 a JAMA Network Open.2022;5(8):e2225445.doi:10.1001/jamanetworkopen.2022.25445(Reprinted) August 4, 2022 7/11 Downloaded From: https://jamanetwork.com/ on 09/21/2023 increased risk of PD.
38verity were not available for extraction; thus, we were unable to further investigate the association of PTSD with PD.Second, sensitivity analysis excluding PD cases in the first year after PTSD diagnosis showed an attenuation of risk estimates.It is well known that PD has a long prodromal phase37; thus, exposure to risk factors would have accrued at least 10 to 15 years before PD diagnosis (ie, the appearance of distinct motor symptoms).However, it is also known that PTSD diagnosis is not necessarily documented at first appearance of symptoms because of (1) strict diagnosis criteria when subsyndromal symptoms accumulate with time until being sufficient for diagnosis; (2) early PTSD episodes being unrecognized as such, undisclosed, or forgotten; or (3) remanifestation of PTSD in older age as a result of stressors such as loneliness and physical illness.38Inaddition, in 2000 the Community Rehabilitation of Persons with Mental Health Disability law