Association between chiropractic spinal manipulation and adverse vascular events of the head and neck: systematic review and meta-analysis

Abstract Background Observational studies have raised concerns about a potential association between chiropractic spinal manipulation (CSM) and craniocervical vascular events (CCVE), including cervical artery dissection and stroke. We tested the null hypothesis of no significant association between CSM and CCVE, measured by odds ratio (OR) at 1-day and 7-day windows post-CSM, secondarily exploring 14 and 30 days. Methods We searched multiple databases spanning inception to March 3, 2024, including epidemiologic studies examining the CSM-CCVE association. Article selection, primary/secondary outcome data extraction, quality ratings, and certainty of evidence were completed by multiple co-authors. Meta-analysis included a random-effects model, funnel plots, and Egger's regression tests. Statistical significance was set at P = 0.025, to account for co-primary outcomes. Results Of 346 studies screened, 10 were included, most commonly using case–control designs ( n = 9). Included studies comprised 93,474 subjects at 1 day and 1,366,036 at 7 days. Five studies used a composite CCVE definition; none examined stroke in isolation. There was no significant association between CSM and CCVE at our coprimary endpoints of 1 day [95% confidence intervals] (OR = 1.12 [0.53–2.36]; P = 0.7654) and 7 days (OR = 1.76; [1.04–3.00]; P = 0.0363). Secondary analyses yielded OR = 2.00 [0.94–4.22]; P = 0.0707 at 14 days and OR = 1.95 [1.33–2.87]; P = 0.0006 at 30 days. Prediction intervals were wide across all time windows (1-day: 0.14–8.81; 7-day: 0.26–11.97; 14-day: 0.14–28.10; 30-day: 0.52–7.32). Certainty of evidence decreased sequentially due to increasing heterogeneity and indirectness over longer time between CSM and CCVE (moderate [1 day], low [7 and 14 days], very low [30 days]). Conclusion The present findings demonstrate no definitive evidence of a meaningful association between CSM and CCVE. Point estimates at follow-up windows of 7, 14, and 30 days demonstrated small positive associations, yet were characterized by high heterogeneity, and low to very low certainty of evidence. Wide prediction intervals, due to between-study variability, suggest future studies could yield associations ranging from apparent protective effects to potential harm. Given that patients may present to chiropractors or other clinicians with early CCVE symptoms (e.g., neck pain/headache), better control over protopathic bias and verification of cervical CSM administration is needed in future studies. Trial registration Systematic review registration: Open Science Framework ( https://osf.io/cdtf6 ).