Postherpetic Neuralgia After Herpes Zoster and Incident Dementia Risk

Overview

This retrospective observational cohort study uses de-identified electronic health record data from the TriNetX Global Collaborative Network to evaluate whether postherpetic neuralgia after herpes zoster is associated with an increased risk of incident dementia. Adults aged 40 to 120 years with incident herpes zoster between 1 October 2015 and 31 December 2024 are identified using diagnostic codes. Postherpetic neuralgia is defined by International Classification of Diseases, 10th Revision, Clinical Modification (ICD-10-CM) codes B02.22 or B02.29 recorded between 90 and 365 days after the index herpes zoster date, and comparators have no such codes within 365 days after index. The primary analysis uses 1:1 propensity score matching and a 365-day landmark design, including only individuals alive and free of dementia at the landmark. Time-to-event analyses estimate hazard ratios for incident dementia and related outcomes.

Data source and study design This is a retrospective cohort study using de-identified electronic health record data accessed through the TriNetX platform. The primary analysis is conducted in the TriNetX Global Collaborative Network. Replication is performed in the TriNetX Asia Pacific (APAC) network as a geographic validation. Population and index date Eligible individuals are aged 40 to 120 years and have an incident herpes zoster diagnosis (ICD-10-CM B02) recorded between 1 October 2015 and 31 December 2024, with a 3-year washout period free of herpes zoster before the index date. Baseline covariates are assessed up to 1 day before the index date. Exposure definition Postherpetic neuralgia is defined by ICD-10-CM codes B02.22 or B02.29 recorded between 90 and 365 days after the index date. The comparator cohort includes individuals with incident herpes zoster and no record of B02.22 or B02.29 within 365 days after the index date. A prespecified stricter exposure definition requires at least two postherpetic neuralgia codes separated by at least 30 days. Landmark design and follow-up The primary analysis uses a 365-day landmark after the index date. Only individuals alive and free of the specified outcome at the landmark enter the risk set. Follow-up starts at the landmark and continues until the first record of the outcome, death, loss to follow-up, or 31 December 2024. Matching and analysis Cohorts are matched 1:1 using nearest neighbour propensity score matching with a caliper of 0.1. Balance is assessed using standardised mean differences. Time-to-event analyses use Kaplan-Meier methods and Cox proportional hazards models, reporting hazard ratios with 95% confidence intervals. Outcomes The primary outcome is incident all-cause dementia defined by ICD-10-CM codes F01, F02, F03, or G30. Prespecified secondary outcomes include Alzheimer disease, vascular dementia, other or unspecified dementia, all-cause mortality, ischaemic stroke, and hip fracture. The Benjamini-Hochberg false discovery rate is applied to the six secondary outcomes only. A negative control outcome of acute appendicitis (ICD-10-CM K35) is included to assess residual bias. Sensitivity and validation analyses Prespecified analyses include an alternative 730-day landmark, the stricter postherpetic neuralgia definition, additional adjustment for recorded zoster vaccination as a baseline covariate, and replication in the TriNetX APAC network using prespecified matching settings.