Advertisement

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Driving under viral impairment: Linking acute SARS-CoV-2 infections to elevated car crash risks

  • Baran Erdik

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    baran@erdik.co.nz

    Affiliations Department of Healthcare Administration, American Vision University, Anaheim, California, United States of America, Hygia Health, Miami, Florida, United States of America

    ⨯

Abstract

This study explores the linkage between acute SARS-CoV-2 and car crashes across U.S. states, correlating with COVID-19 mitigation strategies, vaccination rates, and Long COVID prevalence. This investigation analyzed aggregate COVID-19 and car crash data spanning 2020–2023, with data collection occurring between March and May 2024. Analysis was done via a Poisson regression model, adjusted for population. Key variables included vaccination status, month-specific effects relating to initial pandemic shutdowns, and Long COVID rates. Results demonstrated a significant association between acute COVID-19 infections and an increase in car crashes, independent of Long COVID status to the tune of an OR of 1.25 [1.23-1.26]. This association was observed despite varying mitigation efforts and vaccination rates across states. The study found no protective effect of vaccination against car crashes, challenging prior assumptions about the benefits of vaccination. Notably, the risk associated with COVID-19 was found to be analogous to driving impairments seen with alcohol consumption at legal limits. Findings suggest significant implications for public health policies, especially in assessing the readiness of individuals recovering from COVID-19 to engage in high-risk activities such as pilots or nuclear plant employees. Further research is necessary to establish causation and explore the exact effects of COVID-19 within the CNS affecting cognition and behavior.

Citation: Erdik B (2025) Driving under viral impairment: Linking acute SARS-CoV-2 infections to elevated car crash risks. PLOS Glob Public Health 5(4): e0004420. https://doi.org/10.1371/journal.pgph.0004420

Editor: Jose Ignacio Nazif-Munoz, Université de Sherbrooke: Universite de Sherbrooke, CANADA

Received: May 8, 2024; Accepted: February 20, 2025; Published: April 8, 2025

Copyright: © 2025 Baran Erdik. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting information files.

Funding: The author received no specific funding for this work.

Competing interests: The author has declared that no competing interests exist.

Introduction

Coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) first reported in December 2019. Since then, it has resulted in over one million deaths in the US alone as well as an estimated incidence of almost 3.5 infections per person as of early 2024 [1]. Beyond acute hospitalizations and associated mortality and morbidity, one of the main concerns surrounding COVID-19 is the post-COVID-19 conditions termed long COVID or post-acute sequelae of SARS-CoV-2 infection (PASC) which affect at least a quarter of patients or more surviving even mild cases of COVID. A recent longitudinal follow-up study demonstrated that almost 63% of those who have had COVID-19 met the criteria for long-term COVID [2,3].

Studies in this context have demonstrated that Long Covid is associated with numerous symptoms affecting all bodily systems and can be clustered within categories. One of the main symptom clusters is cognitive complaints, particularly memory and concentration deficits with data indicating more than 30% of those that have had acute COVID-19 reporting such symptoms [4]. Further, it has been shown that reinfections confer increased risk of Long Covid, and the data demonstrating protective effect of vaccination on Long Covid is conflicting [5,6]. In light of such data, as well as removal of all mitigation efforts in the US, the question of Long Covid evolves into a matter of when, not if [7,8].

Several hypotheses have been proposed regarding Long Covid, mainly viral persistence, immune dysregulation, and viral reactivation [9]. From a neuropathological perspective, it has been shown that immune dysregulation within the immune system is evident with persistent immune activation in the context of antibody secreting B cells [10], microglial activation associated with poor cognitive function [11,12], as well as effects of SARS-CoV-2 on the microvasculature, which can explain pan-cerebral issues along with perturbations of synaptic function in general [13,14]. Further, it’s also been shown that macrophages can enter the brain tissue post COVID-19, and stay in perpetuity leading to neural inflammation [15,16]. Not surprisingly, studies have also demonstrated loss of brain volume, increased risk of Alzheimer’s and numerous autonomic nervous system issues as postural tachycardia syndrome (POTS), some with sufficient severity that it may lead to Ondine’s curse [17–19]. Additionally, neural viral reservoirs may perpetuate immune dysregulation, culminating in T-cell exhaustion [20,21]. Regardless of the pathogenesis, the phenomenon commonly referred to as “brain fog,” characterized primarily by cognitive impairments, is becoming increasingly prevalent. Data indicate that even mild acute infections can have lasting effects on brain function, potentially leading to significant challenges in tasks that demand high levels of attention and cognitive processing [22].

One such cognitively demanding task is driving, which relies heavily on psychomotor speed, executive function, and visual processing [23,24]. It is well established that impairments in these domains, as seen in conditions such as ADHD and dementia, correlate with an increased risk of motor vehicle crashes. Despite this, most states do not require specific medical evaluations or tests to obtain a driver’s license, with a few exceptions for conditions like epilepsy or for older adults [25–27]. Given the effects of Long COVID and the sheer scale of numbers affected, the full impact of COVID-19 on driving remains largely unknown. In this sense, studies have analyzed increases in car crashes as well as crash mortality in the US contemporarily with COVID-19 pandemic [28] While these analyses did not directly attribute these increases to COVID-19 beyond government-imposed “stay-at-home” orders -shutdowns-, they may have correctly identified a trend without recognizing the role of SARS-CoV-2 as a contributing factor. The virus has been shown to increase aggression, impair visuoconstructional abilities, and induce cognitive dysfunction—factors that collectively elevate the risk of automobile crashes [29,30].

Initial studies during the pandemic consistently documented a reduction in car crash rates during the early stages of the pandemic [31]. These decreases were mainly observed in less severe collisions occurring during rush hour or in minor-moderate crashes that were not fatal but would require hospitalization [32–36]. Conversely, fatal crashes increased, largely attributed to speeding and reduced traffic volume [37,38].

Beyond speeding, factors such as reckless driving, inattentiveness, and aggression have also been implicated in the rise of severe crashes, although no direct links to SARS-CoV-2’s effects on the central nervous system (CNS) have been proposed [28]. Further, studies in numerous states have repeated these findings, associating increases risk-taking behaviors such as failure to use a seatbelt with increases in fatality and severity of crashes. In the same sense, studies have also demonstrated that SARS-CoV-2 vaccine hesitancy was associated car crashes and severity, indicating that inherent risk hesitancy and associated risk taking maybe exacerbated by SARS-CoV-2 [29,30,34,39].

Further complicating the picture, while the total number of crashes decreased, fatal crashes increased due to single-vehicle crashes, often attributed to risk-taking behavior [40]. Demonstrating the possible neurophysiological role of SARS-CoV-2 as the main driver of increases in severe and fatal crashes, studies have shown increased risk-taking behavior in recovered COVID-19 patients, possibly due to ongoing damage to the limbic system, mainly the amygdala [41–44]. While some researchers have argued that such risk taking maybe secondary to boredom and direct effects of the pandemic mitigation efforts, this explanation fails to account for global patterns observed across diverse cultural contexts and varying levels of pandemic restrictions [45]. The changes in car crash composition and risk-taking behavior were demonstrated to be a global phenomenon, strengthening our hypothesis that the CNS related changes owing to SARS-CoV-2 are to blame [46,47].

Moreover, although some states suspended on-the-road testing for teenage drivers during the early pandemic, the outcomes did not significantly deviate from the norm, as most states do not require behind-the-wheel training. While young and newly licensed drivers typically have higher crash rates, these incidents are usually non-fatal [48–50]. Lower socioeconomic status (SES) remains a significant factor in road safety and risk-taking behavior within this demographic, reinforcing the link between SES and increased risk. Internationally, similar changes in driving behavior have been observed, such as in Greece and Saudi Arabia, where speeding and fatal crash rates increased despite robust pandemic mitigation efforts, further challenging the notion that these trends result solely from mitigation measures like stay-at-home orders [51,52]. The demographic and psychological profiles of individuals who defied pandemic mitigation measures may offer additional insights into these patterns. Such individuals—often characterized by lower SES, educational attainment, employment in the service industry, ownership of older or less safe vehicles, younger age, male gender, and a history of substance abuse—were more likely to be on the road during stay-at-home orders, increasing their exposure to SARS-CoV-2 and, consequently, their likelihood of subsequent reinfections [53–56]. Supporting this perspective, data suggest that drivers self-reported engaging in fewer risky behaviors during the pandemic [57]. However, international data, such as from Thailand, where a second stay-at-home order following a surge in COVID-19 cases resulted in an increase in crashes and fatalities, contradicts the notion that mitigation efforts alone can account for these changes [58] Additionally, the rise in motorcycle crashes and non-compliance with safety measures globally suggests a broader shift in risk-taking behavior, potentially linked to the neurophysiological effects of SARS-CoV-2 [59].

Materials & methods

Study design

In response to the growing SARS-CoV-2 and Long COVID-19 epidemic, we conducted an in-depth analysis of the correlation between acute COVID-19 cases and the incidence of car crashes across seven states. This analysis was stratified by SARS-CoV-2 vaccination rates, Long COVID-19 prevalence as reported by the Centers for Disease Control and Prevention (CDC), and early pandemic policies, specifically in states with minimal or no COVID-19 restrictions [60].

Data collection

Data collection spanned March to May 2024, covering data from 2020 to 2023 to ensure finalization and completeness. To ensure robust findings, we amalgamated car crash data from each state’s Department of Transportation (DOT) or equivalent, cross-referencing with the National Highway Traffic Safety Administration (NHTSA) data repository to ensure accuracy and consistency. Specifically, crash data were obtained from the following sources: the TxDOT Crash Query Tool (C.R.I.S. Query) for Texas, the MassDOT IMPACT tool for Massachusetts, the Washington State Collision Analysis Tool for Washington State, the Iowa DOT ICAT for Iowa, the Connecticut Crash Data Repository for Connecticut, the Florida Highway Safety and Motor Vehicles (FLHSMV) Crash Dashboard for Florida, and the GDOT Crash Data Portal for Georgia [61–68]. This multi-source approach ensured that discrepancies or incomplete data from one repository could be cross-verified and resolved.

COVID-19 vaccination data were sourced from the CDC’s National Center for Immunization and Respiratory Diseases (NCIRD) [69]. Given the inherent variability in test reporting and the widespread use of at-home antigen testing—results of which are largely unreported—we employed PCR positivity rates obtained from the U.S. Department of Health & Human Services/ Centers for Disease Control and Prevention (CDC) as a surrogate marker for COVID-19 incidence [70]. PCR positivity rates were validated using time-series data from the Pandemic Mitigation Collaborative (PMC) and wastewater SARS-CoV-2 RNA levels from the CDC [1,70,71]. Refined wastewater metrics, accounting for population size, wastewater flow rates, and environmental conditions, showed strong positive correlations with PCR positivity rates (R² > 0.92, p < 0.001), with validation performed via regression analysis. These validations confirm PCR positivity as a reliable proxy for COVID-19 case numbers in line with previous literature [72] not only validating PCR positivity rates as a reliable measure of COVID-19 incidence but also underscoring the utility of such in epidemiological analyses, particularly when traditional testing data may be incomplete or biased.

Statistical reasoning

The Poisson regression model was used to assess the relationship between acute COVID-19 incidence and car crashes. Fixed effects were applied to control for state-specific characteristics, as the analysis focused on state-level data. State-level fixed effects accounted for differences in policy implementation, population density, and road usage patterns. Overdispersion was evaluated using the Pearson dispersion statistic. The absence of significant overdispersion justified the use of the Poisson model over a Negative Binomial alternative.

The log of the population was included as an offset in the Poisson regression equation to account for differences in population size among states. The outcome variable, the count of car crashes, was modeled as a log-transformed variable. Covariates included vaccination rates, Long COVID-19 prevalence, and PCR positivity rates. Interaction terms between mitigation measures and vaccination rates were considered but ultimately excluded due to lack of statistical significance and lack of available enforcement/compliance data. Descriptive statistics for all datasets are detailed in Table 1.

thumbnail
Download:
Table 1. Descriptive Statistics.

https://doi.org/10.1371/journal.pgph.0004420.t001

Correlation analysis and heat map

A correlation analysis was conducted to assess whether changes in one variable were associated with changes in another. Highly correlated variables (above 0.75) were excluded from the multivariable model to avoid collinearity. Pearson’s correlation coefficient was used to quantify the strength and direction of associations. To visualize correlations, a heat map was constructed, categorizing correlations as high (above 0.75), medium (0.5–0.75), or low (below 0.5). The heat map revealed no significant correlations among predictor variables, validating the independence of variables used in the Poisson regression model (Fig 1). The heat map also informed variable selection by excluding predictors with potential collinearity, thereby improving the robustness of the regression model.

thumbnail
Download:
Fig 1. Correlation Heat Map.

https://doi.org/10.1371/journal.pgph.0004420.g001

Software

All analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC). No custom scripts or extensions were applied to the software during the analysis.

Supplementary data

The complete dataset is available in S1 and S2 Data. S1 Data (Wastewater Data) includes three tabs: percentage subsets of wastewater metrics, concentration data, and a master list of facilities. S2 Data provides the remaining datasets. Changes in facility names, primarily within the Verily-sourced wastewater data, are noted.

Statistical analysis

The 2020–2022 data were fitted to a Poisson regression model. The log of population (per the US Census) was an offset variable for differences among seven states (Connecticut, Florida, Georgia, Iowa, Massachusetts, Texas, and Washington). (Table 2) Binary variables for March, April, and May were included separately because those were the months of the shutdowns in most US states and were included in the model that found no correlation between car crashes and concurrent COVID-19 rate (% positive PCR). The form of the equation with log-transformation was: log(μi) = β0 + β1 (COVID-19 rate) + β2 (vaccination rate) + β3 (March) + β4 (April) + β5 (May) + log(populationi), Where: μi represents the expected number of car crashes in state i; log(populationi) is included as an offset to account for population differences.

thumbnail
Download:
Table 2. State Characteristics.

https://doi.org/10.1371/journal.pgph.0004420.t002

A Poisson regression model was chosen for its effectiveness in modeling count data, particularly for rare events like car crashes, where the data is characterized by rare occurrences and skewed distributions, as supported by prior literature [25,37]. Fixed effects were used in the model to account for specific state-level differences, given the focus on evaluating how various state policies, mitigation efforts, and vaccination rates influenced crash rates to control for observed, non-random variability across states while ensuring the model’s estimates were not biased by unobserved heterogeneity. This approach was appropriate given the study’s focus on state-specific impacts of SARS-CoV-2 incidence and policy rather than broader variation across states, which would have warranted random effects. The log of the population was included as an offset variable to adjust for differing population sizes across states. Overdispersion, a common issue in count data where variance exceeds the mean, was assessed using the dispersion statistic (ratio of the deviance to the degrees of freedom), being critical in ensuring the model’s robustness and accuracy, as overdispersion can lead to underestimation of standard errors, resulting in misleading inferences. In our model, overdispersion was specifically assessed by comparing the variance and mean of the outcome variable, with the dispersion statistic (ratio of the deviance to the degrees of freedom) suggesting minimal overdispersion (value close to 1), indicating that the Poisson model appropriately fits the data without underestimating standard errors. Consequently, a negative binomial model, typically used for significant overdispersion, was deemed unnecessary due to the observed minor overdispersion. Ultimately, Poisson regression is well-suited to situations requiring adjustments for varying population sizes across different States, thus allowing for a more nuanced analysis of the incidence rate of car crashes as a function of predictor variables. By incorporating variables such as COVID-19 infection rates alongside other relevant covariates, Poisson regression has in this context facilitated a robust framework for analyzing the relationship between SARS-CoV-2 and road safety outcomes while accounting for state-level population differences and other contextual factors.

Results

The increases in crash frequency temporally coincided with the initial dip in SARS-CoV-2 rates thanks to stay-at-home orders, and subsequently ever increased following removal of most mitigations and advent of ever-virulent variants, e.g., Delta and Omicron. Not surprisingly, our data demonstrate lower COVID-19 rates in states with more stringent mitigation efforts, namely masking mandates and higher vaccination rates. (Tables 2 and 3) As discussed above, to evaluate the relationship between variables, the correlation analysis conducted visually summarizes the correlations in a heat map (Fig 1), categorizing them as high (>0.75), medium (0.5–0.75), or low (<0.5). The heat map (Fig 1) revealed no significant correlations among predictor variables, validating their independence and inclusion in the Poisson regression model. Ultimately, our analysis demonstrated that acute COVID as measured by PCR positivity was associated with an odds ratio (OR) of 1.25 (1.23-1.26 95%CI) for subsequent car crashes indicating a 25% increase in crash risk associated with acute COVID-19. (Table 3). The results were consistent across univariate and multivariate analyses, demonstrating statistically significant results (p<0.05) in all states except Iowa and Washington. (Table 3)

thumbnail
Download:
Table 3. Univariable and Multivariable Analysis on Car Crashes by COVID-19 Rate through Poisson Regression Model.

https://doi.org/10.1371/journal.pgph.0004420.t003

Discussion

Previous studies utilizing similar Poisson regression has demonstrated a connection between the number of accumulated COVID-19 cases and an increase in road fatalities, but not simultaneously and in a temporally associated manner [73]. Our analysis extends these findings by establishing a temporal link, showing that car crashes are associated with increased COVID-19 rates acutely. Furthermore, we found that this increase in crashes is independent of Long COVID rates, focusing on acute infections as the primary driver. Interestingly, vaccination does not appear to confer a protective effect against crash risk, contradicting prior studies associating vaccine hesitancy with elevated crash risk [39]. States with extended mask mandates, like Connecticut, experienced prolonged periods of reduced traffic crash rates, whereas our inability to establish statistical significance in Iowa and Washington likely stems from data reporting issues and differences in pandemic-related behavioral changes, such as work-from-home arrangements in Washington and a lack of mitigation efforts in Iowa.

Further, we chose not to include gas prices as a variable, as previous data has demonstrated that gas demand is inelastic, meaning driving choices do not immediately reflect pricing, and this variable was deemed irrelevant to our focus on acute COVID-19. Additionally, over 10% of US vehicles currently on the road are powered by electricity reducing the relevance of traditional fuel price metrics [74,75].

Our observed odds ratio 1.25 [1.23-1.26] is comparable to that associated with a blood alcohol concentration (BAC) of 0.08%, the legal threshold for driving under the influence (DUI) in many states. Comparatively, states such as Utah, which set lower BAC limits of 0.05%, show similar crash risk odds ratios, underlining the substantial impact of acute COVID-19 on driving safety [76]. This finding is also consistent with odds ratios linked to various problematic driving behaviors, such as habitual speeding or running red lights—activities that are illegal and heavily enforced due to their significant risk profiles drawing the ire of public [77]. Although this study examined aggregate data cannot definitively determine the causality of COVID’s effects on the central nervous system (CNS) & car crash frequency, the hypothesis is defensible both mechanistically and statistically.

Our analysis faced several limitations, including the rising prevalence of uninsured drivers, which may contribute to crash underreporting, and the inability to directly confirm acute COVID-19 infections in individuals involved in these crashes, given the population-level nature of the data. Despite these limitations, our findings provide substantial support for the hypothesis that acute COVID-19 contributes to increased car crash risk.

Some prior studies have reported decreases in car crashes frequencies concurrent with rising COVID-19 case rates. However, these studies were conducted in countries such as Japan and Greece, where stringent mitigation measures significantly restricted population mobility, leading to immediate reductions in traffic volume. Importantly, not all studies demonstrated corresponding decreases in fatalities beyond reductions in crash frequency [78,79]. While these studies addressed changes in crash occurrences alongside fatalities and injuries, our analysis highlights the distinct role of acute COVID-19 in contributing to the risk of car crashes. This distinction is critical, as crash severity involves unique risk factors that differ from those influencing crash frequencies alone. For example, prior research shows that population density and traffic density as inversely correlated risk factors for crashes and fatalities [80,81].

Previous studies on exploration of crash heterogeneity further support the importance of distinguishing between outcomes such as crash frequency and severity. The causal mechanisms underlying crashes of varying severity differ significantly, and generalizing across these categories can lead to erroneous conclusions, highlighting the need for targeted analysis [82]. Ultimately, this perspective strengthens our hypothesis: in settings without restrictive mobility measures, the persistence of crashes suggests that acute COVID-19 may exacerbate risk through mechanisms such as impaired cognition or judgment.

To further elucidate the implications of COVID-19 sequelae on car crashes, additional research—such as case-control or prospective cohort studies—is urgently needed. This is particularly critical given the impacts of COVID-19 on the CNS, especially its effects on cognition, which may lead to judgment errors in high-stakes environments, such as flight decks or nuclear energy plants, with potentially catastrophic consequences.

Conclusions

With the COVID-19 Public Health Emergency (PHE) expiring on May 11, 2023, society has primarily suspended protection against COVID-19. In light of lack of mitigations surrounding COVID-19 spread and the failure of the vaccination-only public health approach that has focused mainly on acute hospitalizations and deaths, further research is urgently needed into Long COVID as well as treatments to help manage the sequelae of an Acute Covid infection.

Legislatures and public health experts should not only view COVID-19 in the sense of acute mortality and morbidity. As brief neuropsychological tests are predictive of Long COVID and validated for driving risk, agencies responsible for driving licenses should implement a short questionnaire at license renewal inquiring about Long COVID/COVID and refer applicants to neuropsychological testing as needed. Perhaps even asking if drivers have had ongoing taste and smell disturbances post-COVID might link applicants with ongoing neurological sequelae of COVID [83–85].

Finally, clinicians, particularly those dealing with Long Covid patients in the cognitive setting, such as neurologists, must remember their obligation to report patients who potentially constitute medically impaired drivers. Such patients may include those with COVID-19 or those who suffer from the after-effects. Indeed, clinicians, particularly primary care practitioners, should contemplate warning patients with COVID-19 that they should minimize driving and report back if they feel any cognitive sequelae.

Supporting information

S1 Data. Dataset underlying the main analysis with tabs separating data for differing states.

https://doi.org/10.1371/journal.pgph.0004420.s001

(XLSX)

S2 Data. Wastewater data including three tabs: percentage subsets of wastewater metrics, concentration data, and a master list of facilities, changes in facility names, primarily within the Verily-sourced wastewater data, are noted.

https://doi.org/10.1371/journal.pgph.0004420.s002

(XLSX)

Acknowledgments

I would like to thank Kaitlyn Bruneau, LCSW with her assistance in data collection as well as Sara Anne Willette in her assistance with raw wastewater data collection and analysis.

References

  1. 1. Hoerger M. U.S. SARS-CoV-2 wastewater levels, COVID-19 case estimates, and 4-week forecast: Report for March 11, 2024. Pandemic Mitigation Collaborative [Internet]. 2024; (March 11 2024). Available from: https://www.pmc19.com/data/PMC_Report_Mar11_2024.pdf
  2. 2. Sedgley R, Winer-Jones J, Bonafede M. Long COVID Incidence in a Large US Ambulatory Electronic Health Record System. Am J Epidemiol. 2023;192(8):1350–7. pmid:37073410
  3. 3. Wentz E, Ni Z, Yenokyan K, Vergara C, Pahwa J, Kammerling T, et al. Cohort profile: the Johns Hopkins COVID Long Study (JHCLS), a United States Nationwide Prospective Cohort Study. medRxiv. 2024:2024.04.25.24306387. pmid:38712261
  4. 4. Seeßle J, Waterboer T, Hippchen T, Simon J, Kirchner M, Lim A, et al. Persistent Symptoms in Adult Patients 1 Year After Coronavirus Disease 2019 (COVID-19): A Prospective Cohort Study. Clin Infect Dis. 2022;74(7):1191–8. pmid:34223884
  5. 5. Bowe B, Xie Y, Al-Aly Z. Acute and postacute sequelae associated with SARS-CoV-2 reinfection. Nat Med. 2022;28(11):2398–405. pmid:36357676
  6. 6. Byambasuren O, Stehlik P, Clark J, Alcorn K, Glasziou P. Effect of covid-19 vaccination on long covid: systematic review. BMJ Med. 2023;2(1):e000385. pmid:36936268
  7. 7. Doll MK, Waghmare A, Heit A, Levenson Shakoor B, Kimball LE, Ozbek N, et al. Acute and Postacute COVID-19 Outcomes Among Immunologically Naive Adults During Delta vs Omicron Waves. JAMA Netw Open. 2023;6(2):e231181. pmid:36853602
  8. 8. Marjenberg Z, Leng S, Tascini C, Garg M, Misso K, El Guerche Seblain C, et al. Risk of long COVID main symptoms after SARS-CoV-2 infection: a systematic review and meta-analysis. Sci Rep. 2023;13(1):15332. pmid:37714919
  9. 9. Davis HE, McCorkell L, Vogel JM, Topol EJ. Long COVID: major findings, mechanisms and recommendations. Nat Rev Microbiol. 2023;21(3):133–46. pmid:36639608
  10. 10. Mina Y, Enose-Akahata Y, Hammoud DA, Videckis AJ, Narpala SR, O’Connell SE, et al. Deep Phenotyping of Neurologic Postacute Sequelae of SARS-CoV-2 Infection. Neurol Neuroimmunol Neuroinflamm. 2023;10(4):e200097. pmid:37147136
  11. 11. Braga J, Lepra M, Kish SJ, Rusjan PM, Nasser Z, Verhoeff N, et al. Neuroinflammation After COVID-19 With Persistent Depressive and Cognitive Symptoms. JAMA Psychiatry. 2023;80(8):787–95. pmid:37256580
  12. 12. Fernández-Castañeda A, Lu P, Geraghty AC, Song E, Lee M-H, Wood J, et al. Mild respiratory COVID can cause multi-lineage neural cell and myelin dysregulation. Cell. 2022;185(14):2452–2468.e16. pmid:35768006
  13. 13. Guedj E, Campion JY, Dudouet P, Kaphan E, Bregeon F, Tissot-Dupont H, et al. 18F-FDG brain PET hypometabolism in patients with long COVID. Eur J Nucl Med Mol Imaging. 2021;48(9):2823–33. pmid:33501506
  14. 14. Partiot E, Hirschler A, Colomb S, Lutz W, Claeys T, Delalande F, et al. Brain exposure to SARS-CoV-2 virions perturbs synaptic homeostasis. Nat Microbiol. 2024;9(5):1189–206. pmid:38548923
  15. 15. Lee MH, Perl DP, Steiner J, Pasternack N, Li W, Maric D, et al. Neurovascular injury with complement activation and inflammation in COVID-19. Brain. 2022;145(7):2555–68. pmid:35788639
  16. 16. Stein JA, Kaes M, Smola S, Schulz-Schaeffer WJ. Neuropathology in COVID-19 autopsies is defined by microglial activation and lesions of the white matter with emphasis in cerebellar and brain stem areas. Front Neurol. 2023;14:1229641. pmid:37521293
  17. 17. Baker J, Incognito AV, Wilson RJA, Raj SR. Syncope and silent hypoxemia in COVID-19: Implications for the autonomic field. Auton Neurosci. 2021;235:102842. pmid:34246957
  18. 18. Douaud G, Lee S, Alfaro-Almagro F, Arthofer C, Wang C, McCarthy P, et al. SARS-CoV-2 is associated with changes in brain structure in UK Biobank. Nature. 2022;604(7907):697–707. pmid:35255491
  19. 19. Ziff OJ, Ashton NJ, Mehta PR, Brown R, Athauda D, Heaney J, et al. Amyloid processing in COVID-19-associated neurological syndromes. J Neurochem. 2022;161(2):146–57. pmid:35137414
  20. 20. Proal AD, VanElzakker MB, Aleman S, Bach K, Boribong BP, Buggert M, et al. SARS-CoV-2 reservoir in post-acute sequelae of COVID-19 (PASC). Nat Immunol. 2023;24(10):1616–27. pmid:37667052
  21. 21. Yin K, Peluso MJ, Luo X, Thomas R, Shin M-G, Neidleman J, et al. Long COVID manifests with T cell dysregulation, inflammation and an uncoordinated adaptive immune response to SARS-CoV-2. Nat Immunol. 2024;25(2):218–25. pmid:38212464
  22. 22. Gutman EG, Salvio AL, Fernandes RA, Duarte LA, Raposo-Vedovi JV, Alcaraz HF, et al. Long COVID: plasma levels of neurofilament light chain in mild COVID-19 patients with neurocognitive symptoms. Mol Psychiatry. 2024;29(10):3106–16. pmid:38678084
  23. 23. Marottoli RA, Cooney LM Jr, Wagner R, Doucette J, Tinetti ME. Predictors of automobile crashes and moving violations among elderly drivers. Ann Intern Med. 1994;121(11):842–6. pmid:7978696
  24. 24. Stutts JC, Stewart JR, Martell C. Cognitive test performance and crash risk in an older driver population. Accid Anal Prev. 1998;30(3):337–46. pmid:9663293
  25. 25. Curry AE, Yerys BE, Metzger KB, Carey ME, Power TJ. Traffic Crashes, Violations, and Suspensions Among Young Drivers With ADHD. Pediatrics. 2019;143(6):e20182305. pmid:31110164
  26. 26. Grabowski DC, Campbell CM, Morrisey MA. Elderly licensure laws and motor vehicle fatalities. JAMA. 2004;291(23):2840–6. pmid:15199034
  27. 27. Krumholz A, Fisher RS, Lesser RP, Hauser WA. Driving and epilepsy. A review and reappraisal. JAMA. 1991;265(5):622–6. pmid:1987412
  28. 28. Dong X, Xie K, Yang H. How did COVID-19 impact driving behaviors and crash Severity? A multigroup structural equation modeling. Accid Anal Prev. 2022;172:106687. pmid:35500416
  29. 29. de Paula JJ, Paiva RERP, Souza-Silva NG, Rosa DV, Duran FL de S, Coimbra RS, et al. Selective visuoconstructional impairment following mild COVID-19 with inflammatory and neuroimaging correlation findings. Mol Psychiatry. 2023;28(2):553–63. pmid:35701598
  30. 30. Hu Y, Ye B, Ma T, Feng Z, Chen X, Huang D, et al. Exposure to COVID-19 and aggression: the mediating role of anxiety and the moderating role of rumination. Curr Psychol. 2023:1–9. pmid:37359682
  31. 31. Monfort SS, Cicchino JB, Patton D. Weekday bicycle traffic and crash rates during the COVID-19 pandemic. J Transp Health. 2021;23:101289. pmid:34745883
  32. 32. Brodeur A, Cook N, Wright T. On the effects of COVID-19 safer-at-home policies on social distancing, car crashes and pollution. J Environ Econ Manage. 2021;106:102427. pmid:33583969
  33. 33. Lee J, Liu H, Abdel-Aty M. Changes in traffic crash patterns: Before and after the outbreak of COVID-19 in Florida. Accid Anal Prev. 2023;190:107187. pmid:37364361
  34. 34. Adanu EK, Brown D, Jones S, Parrish A. How did the COVID-19 pandemic affect road crashes and crash outcomes in Alabama?. Accid Anal Prev. 2021;163:106428. pmid:34649013
  35. 35. Stiles J, Kar A, Lee J, Miller HJ. Lower Volumes, Higher Speeds: Changes to Crash Type, Timing, and Severity on Urban Roads from COVID-19 Stay-at-Home Policies. Transp Res Rec. 2023;2677(4):15–27. pmid:37153167
  36. 36. Kabiri N, Abbasi A, Pashazadeh F, Hajebrahimi S, Soleimanpour H. The Impact of The COVID-19 Pandemic on Hospital Admissions Due to Road Traffic Crashes; a Systematic Review and Meta-Analysis. Arch Acad Emerg Med. 2023;12(1):e12. pmid:38162384
  37. 37. Patwary AL, Khattak AJ. Crash harm before and during the COVID-19 pandemic: Evidence for spatial heterogeneity in Tennessee. Accid Anal Prev. 2023;183:106988. pmid:36724654
  38. 38. Hughes JE, Kaffine D, Kaffine L. Decline in Traffic Congestion Increased Crash Severity in the Wake of COVID-19. Transp Res Rec. 2023;2677(4):892–903. pmid:37153182
  39. 39. Redelmeier DA, Wang J, Thiruchelvam D. COVID Vaccine Hesitancy and Risk of a Traffic Crash. Am J Med. 2023;136(2):153–162.e5. pmid:36470796
  40. 40. Doucette ML, Tucker A, Auguste ME, Watkins A, Green C, Pereira FE, et al. Initial impact of COVID-19’s stay-at-home order on motor vehicle traffic and crash patterns in Connecticut: an interrupted time series analysis. Inj Prev. 2021;27(1):3–9. pmid:33115707
  41. 41. Egeli A, Adıgüzel S, Kapıcı Y, Güc B, Yetkin Tekin A, Tekin A. Risk-Taking Behavior in Recovered COVID-19 Patients. Psychiatr Danub. 2021;33(1):107–13. pmid:33857054
  42. 42. Thomasson M, Voruz P, Cionca A, Jacot de Alcântara I, Nuber-Champier A, Allali G, et al. Markers of limbic system damage following SARS-CoV-2 infection. Brain Commun. 2023;5(4):fcad177. pmid:37415776
  43. 43. Lukiw WJ, Pogue A, Hill JM. SARS-CoV-2 Infectivity and Neurological Targets in the Brain. Cell Mol Neurobiol. 2022;42(1):217–24. pmid:32840758
  44. 44. Taskiran-Sag A, Yazgi H. Limbic system damage following SARS-CoV2 infection. Brain Commun. 2023;5(6):fcad340. pmid:38116142
  45. 45. Tsai CI, Zeng Y. Risky but alluring: Severe COVID-19 pandemic influence increases risk taking. J Exp Psychol Appl. 2021;27(4):679–94. pmid:35073131
  46. 46. Alhajyaseen WKM, Almukdad A, Hussain Q, Almallah M, Al Malki MA, Singaravelu J, et al. Road safety status during COVID-19 pandemic: exploring public and road safety expert’s opinions. Int J Inj Contr Saf Promot. 2022;29(2):135–51. pmid:34392810
  47. 47. Kabbush O, Almannaa M, Alarifi SA, Alghamdi A. Assessing the Effect of COVID-19 on the Traffic Safety of Intercity and Major Intracity Roads in Saudi Arabia. Arab J Sci Eng. 2023:1–19. pmid:37361468
  48. 48. Ehsani JP, Kinnear N. Young driver licensing and COVID-19. Inj Prev. 2021;27(6):574–6. pmid:33574128
  49. 49. Vaca FE, Li K. Do the Youngest Drivers Crash the Most?. JAMA Netw Open. 2022;5(4):e228785. pmid:35467738
  50. 50. Walshe EA, Romer D, Wyner AJ, Cheng S, Elliott MR, Zhang R, et al. Licensing Examination and Crash Outcomes Postlicensure in Young Drivers. JAMA Netw Open. 2022;5(4):e228780. pmid:35467733
  51. 51. Michelaraki E, Sekadakis M, Katrakazas C, Ziakopoulos A, Yannis G. One year of COVID-19: Impacts on safe driving behavior and policy recommendations. J Safety Res. 2023;84:41–60. pmid:36868670
  52. 52. Alsofayan YM, Alghnam SA, Alkhorisi AM, Almalki HA, Alsaihani MD, Almazroa MA, et al. Epidemiology of Traffic Injuries before, during and 1 Year after the COVID-19 Pandemic Restrictions: National Findings from the Saudi Red Crescent Authority. Saudi J Med Med Sci. 2022;10(2):111–6. pmid:35602391
  53. 53. Kılıç C, Yıldız Mİ, Emekli E, Gülşen G, Alp A. Psychological factors responsible for low adherence to mask-wearing measures during the COVID-19 pandemic. BJPsych Open. 2022;8(6):e203. pmid:36416230
  54. 54. Reuter M, Rigó M, Formazin M, Liebers F, Latza U, Castell S, et al. Occupation and SARS-CoV-2 infection risk among 108 960 workers during the first pandemic wave in Germany. Scand J Work Environ Health. 2022;48(6):446–56. pmid:35670286
  55. 55. Zhang M. Estimation of differential occupational risk of COVID-19 by comparing risk factors with case data by occupational group. Am J Ind Med. 2021;64(1):39–47. pmid:33210336
  56. 56. Lin L, Shi F, Li W. Assessing inequality, irregularity, and severity regarding road traffic safety during COVID-19. Sci Rep. 2021;11(1):13147. pmid:34162894
  57. 57. Vanlaar WGM, Woods-Fry H, Barrett H, Lyon C, Brown S, Wicklund C, et al. The impact of COVID-19 on road safety in Canada and the United States. Accid Anal Prev. 2021;160:106324. pmid:34371287
  58. 58. Mungmunpuntipantip R, Wiwanitkit V. The COVID-19 pandemic and traffic accidents. S Afr Med J. 2021;111(3):13227. pmid:33944735
  59. 59. Atique S, Asim M, El-Menyar A, Mathradikkal S, Hammo A-A, Baykuziyev T, et al. Motorcycle-related crashes before and during the COVID-19 pandemic: A comparative retrospective observational study from the Middle East. Injury. 2024;55(3):111343. pmid:38309084
  60. 60. Ford ND, Agedew A, Dalton AF, Singleton J, Perrine CG, Saydah S. Notes from the Field: Long COVID Prevalence Among Adults - United States, 2022. MMWR Morb Mortal Wkly Rep. 2024;73(6):135–6. pmid:38359012
  61. 61. CTDOT. Connecticut Crash Data Repository (CTCDR) Available from: https://www.ctcrash.uconn.edu/QueryTool2.action
  62. 62. DOT I. ICAT - Iowa Crash Analysis Tool. Available from: https://icat.iowadot.gov
  63. 63. FLHSMV. Crash Dashboard. Available from: https://www.flhsmv.gov/traffic-crash-reports/crash-dashboard/
  64. 64. GDOT. Georgia Crash Data. Available from: https://gdot.aashtowaresafety.net/crash-data#/
  65. 65. MassDOT. IMPACT. Available from: https://apps.impact.dot.state.ma.us/cdp/home
  66. 66. NHTSA. National Center for Statistics and Analysis (NCSA) Motor Vehicle Traffic Crash Data. Available from: https://cdan.dot.gov/query
  67. 67. Patrol WS. Collision Analysis Tool Available from: https://fortress.wa.gov/wsp/collisionanalysistool/Query/SearchCriteria
  68. 68. TxDOT. TxDOT Crash Query Tool. Available from: https://cris.dot.state.tx.us/public/Query/app/home
  69. 69. Services USDoHH. COVID-19 Vaccinations in the United States,Jurisdiction. Available from: https://catalog.data.gov/dataset/covid-19-vaccinations-in-the-united-statesjurisdiction-07bc5
  70. 70. Services USDoHH. COVID Data Tracker. Available from: https://covid.cdc.gov/covid-data-tracker/#datatracker-home
  71. 71. Parkins MD, Lee BE, Acosta N, Bautista M, Hubert CRJ, Hrudey SE, et al. Wastewater-based surveillance as a tool for public health action: SARS-CoV-2 and beyond. Clin Microbiol Rev. 2024;37(1):e0010322. pmid:38095438
  72. 72. Saber LB, Kennedy SS, Yang Y, Moore KN, Wang Y, Hilton SP, et al. Correlation of SARS-CoV-2 in Wastewater and Individual Testing Results in a Jail, Atlanta, Georgia, USA. Emerg Infect Dis. 2024;30(13):S21–7. pmid:38561638
  73. 73. Robertson L. Did “long COVID” increase road deaths in the USA?. Inj Prev. 2025;31(1):18–20. pmid:38485228
  74. 74. Dwyer M. United States: Electric vehicles and hybrids surpass 16% of total 2023 U.S. light-duty vehicle sales. In: Administration USEI, editor. 2024.
    • 75. Eitches E, Crain V. Using gasoline data to explain inelasticity. In: Statistics USBoL, editor. 2016.
      • 76. Blomberg RD, Peck RC, Moskowitz H, Burns M, Fiorentino D. The Long Beach/Fort Lauderdale relative risk study. J Safety Res. 2009;40(4):285–92. pmid:19778652
      • 77. Akhtar S, Aldhafeeri E, Alshammari F, Jafar H, Malhas H, Botras M, et al. A proportional odds model of risk behaviors associated with recurrent road traffic crashes among young adults in Kuwait. BMC Med Res Methodol. 2022;22(1):19. pmid:35026988
      • 78. Sekadakis M, Katrakazas C, Michelaraki E, Kehagia F, Yannis G. Analysis of the impact of COVID-19 on collisions, fatalities and injuries using time series forecasting: The case of Greece. Accid Anal Prev. 2021;162:106391. pmid:34525414
      • 79. Shimada R, Kibayashi K. Changes in the number of traffic collisions during the various waves of COVID-19 infection in Japan. PLoS One. 2022;17(12):e0278941. pmid:36520824
      • 80. Fischer K, Sternfeld I, Melnick DS. Impact of population density on collision rates in a rapidly developing rural, exurban area of Los Angeles County. Inj Prev. 2013;19(2):85–91. pmid:22773021
      • 81. Robertson LS. Daily motor vehicle traffic volume and other risk factors associated with road deaths in U.S. counties. J Safety Res. 2024;90:43–7. pmid:39251297
      • 82. Knipling RR. Crash Heterogeneity: Implications for Naturalistic Driving Studies and for Understanding Crash Risks. Transportation Research Record: Journal of the Transportation Research Board. 2017;2663(1):117–25.
      • 83. Arbula S, Pisanu E, Bellavita G, Menichelli A, Lunardelli A, Furlanis G, et al. Insights into attention and memory difficulties in post-COVID syndrome using standardized neuropsychological tests and experimental cognitive tasks. Sci Rep. 2024;14(1):4405. pmid:38388708
      • 84. Maffei M, Mazzatenta A, Origlia N. Editorial: Loss of taste and smell in COVID-19 patients: A prognostic tool and a starting point to investigate the action of SARS-CoV-2 in the central nervous system. Front Cell Neurosci. 2023;17:1191227. pmid:37078006
      • 85. Sharetts R, Moein ST, Khan R, Doty RL. Long-Term Taste and Smell Outcomes After COVID-19. JAMA Netw Open. 2024;7(4):e247818. pmid:38652477