Introduction

Road-traffic injuries are responsible for approximately 1.19 million deaths globally each year and a substantially larger non-fatal disability burden, with low- and middle-income countries bearing the disproportionate share.(1,2) India alone contributes nearly one-tenth of all global road-traffic deaths and has the largest absolute road-injury burden of any single country.(1,3,4) The United Nations Decade of Action for Road Safety 2021-2030 has set a target of halving global road-traffic deaths and serious injuries by 2030, with explicit prioritisation of vulnerable-road-user (VRU) protection.(5)

Within this context, vulnerable road users - pedestrians, cyclists and motorcyclists - constitute a heterogeneous and rapidly growing risk pool whose protection requires distinct policy levers. Pedestrian protection depends on infrastructure design, speed management and crossing engineering; cyclist protection depends on segregated infrastructure, route planning and visibility; motorcyclist protection depends on helmet compliance, licensing, anti-lock braking adoption and post-crash care.(6-9) These levers are not interchangeable, and the relative weight of investment between them should be guided by the modal composition of road-injury burden in each setting.(9,10)

The Indian evidence base, however, has historically been dominated by aggregate road-death counts from administrative systems, particularly the Ministry of Road Transport and Highways and the National Crime Records Bureau.(11,12) Aggregate counts conceal mode-specific composition and capture only the fatal end of the injury distribution; they cannot inform the relative emphasis between motorcyclist, pedestrian and cyclist policy. Existing sub-national disability-adjusted life-year (DALY) analyses have summarised total road-injury burden but have not, to our knowledge, characterised state-level modal composition through the COVID-19 pandemic recovery period of 2020 to 2023.(3,4,13)

The COVID-19 pandemic period offers a natural quasi-experiment for understanding state-level road-injury dynamics. National lockdowns in 2020 sharply reduced motorised mobility and overall road-injury exposure, with partial relaxation through 2021 and progressive normalisation by 2023.(14,15) Whether the post-lockdown trajectory was uniform across states and modes, whether high-burden states rebounded disproportionately, and whether the pandemic equalised or preserved state burden positions are open questions of direct relevance to vulnerable-road-user policy.(16,17)

We therefore conducted an ecological cross-sectional analysis with three pre-specified objectives. First, to quantify mode-specific road-injury DALY burden across all 32 Indian state and union-territory analytical units for 2020 to 2023, with cross-validation against administrative road-death rankings where available. Second, to characterise the pandemic-to-recovery trajectory by state and mode using a state-level rebound regression model that tests whether high-baseline-burden states changed disproportionately. Third, to quantify mode-specific inter-state inequality through Gini coefficients and Lorenz curves, in order to identify the modal stratum and the state cluster where VRU policy intensification is most needed.

Material And Methods

Study design and data sources

We conducted an ecological cross-sectional analysis using publicly available aggregated secondary data. The primary analytical source was the Global Burden of Disease (GBD) 2021 subnational extract for Indian states and state-equivalent units, accessed through the Global Burden of Disease Results Tool of the Institute for Health Metrics and Evaluation.(1,3,4,18) Cross-validation was conducted against state road-death counts from the Ministry of Road Transport and Highways report Road Accidents in India 2023 for the ten high-burden states for which calendar-year 2023 counts are reported.(11) The National Crime Records Bureau report Accidental Deaths and Suicides in India 2023 was used to provide context on accidental death rates.(12) Reporting follows the STROBE statement for cross-sectional observational studies.(19)

Time frame and analytical units

The analytical panel comprised 32 state and state-equivalent units in India for the four calendar years 2020 to 2023. Geographic units were harmonised across GBD, Ministry of Road Transport and Highways and National Crime Records Bureau through a fixed crosswalk. Jammu and Kashmir and Ladakh remained combined as a single analytical unit, consistent with the GBD 2021 subnational scheme; smaller territories that are not separately enumerated in all sources were combined into an "Other Union Territories" unit. This choice prioritised internal consistency across data sources over exact reproduction of every current administrative boundary.

Mode-specific outcomes

We extracted age-standardised DALY rates and age-standardised mortality rates per 100,000 population for both sexes combined and all ages, for the following GBD road-injury sub-categories: pedestrian road injuries, cyclist road injuries, motorcyclist road injuries, motor-vehicle road injuries, other road injuries, and the aggregate "road injuries" category. Vulnerable-road-user (VRU) burden was defined a priori as the sum of pedestrian, cyclist and motorcyclist DALY rates, consistent with World Health Organization terminology.(1,5)

Pandemic-to-recovery rebound analysis

For each mode, we computed the per-state change in age-standardised DALY rate between 2020 (lockdown year) and 2023 (recovery year) as the absolute difference and as the percent change. To examine whether high-baseline-burden states changed disproportionately, we fitted ordinary-least-squares regression models within each mode of the form change_in_rate_2023_2020 = beta_0 + beta_1 * baseline_rate_2020 + epsilon, with normal-approximation 95% confidence intervals. A negative slope (beta_1 < 0) would indicate convergence (high-baseline states rebounded less or fell more); a positive slope would indicate divergence; a slope statistically indistinguishable from zero would indicate that state burden positions persisted through the pandemic.

Inter-state inequality

Inter-state inequality in age-standardised DALY rates was quantified for each year and mode using the coefficient of variation, the Gini coefficient (computed across the 32 state-level rates), the top-to-bottom ratio (maximum divided by minimum), and the share of total mode-specific DALY burden held by the top 20% of states. We additionally generated Lorenz curves of cumulative state-share against cumulative burden-share for the major modes in 2023.(20)

Cross-validation against administrative data

For the ten high-burden states for which Ministry of Road Transport and Highways calendar-year 2023 road-death counts are reported, we computed the Spearman rank correlation between GBD road-injury DALY rank and administrative road-death rank, and the Pearson correlation between absolute counts. Discordance was interpreted qualitatively given differences in year of measurement, definition (police-recorded fatalities vs modelled DALYs) and reporting construct.(21)

Quality control and software

All numeric results were independently recomputed from the analytical panel prior to manuscript preparation; all rounded values reported below were derived from the recomputed analytical files. Analyses were performed in Python with the pandas, numpy and scipy libraries; figures were produced with matplotlib. No individual-level data were used; institutional review board approval and informed consent were not applicable.

Results

National burden by mode and year

Across the 32-unit panel, motorcyclist injuries accounted for the largest median state-level age-standardised DALY rate in 2023 (362.9 per 100,000), followed by motor-vehicle injuries (227.3), pedestrian injuries (208.7), cyclist injuries (58.3) and other road injuries (6.9) (Table 1, Figure 1). The aggregate road-injury median DALY rate in 2023 was 868.7 per 100,000 (state-level range 474 to 1752). Vulnerable-road-user mechanisms together contributed a median 74.7% of state-level road-injury DALYs in 2023 (versus 74.2% in 2020), with motorcyclists alone contributing 42.2%, pedestrians 23.5% and cyclists 7.2% at the median state level.

Figure 1: Median state-level age-standardised road-injury DALY rate by mode, India, 2020-2023. Lines show the state-level median across the 32 analytical state and state-equivalent units. Motorcyclist injuries dominate the modal composition throughout. DALY: disability-adjusted life year.

Pandemic-to-recovery rebound by mode

Between 2020 and 2023, modal trajectories diverged (Table 2, Figure 2). Cyclist DALY rates rose in 30 of 32 states (median +5.4%), motorcyclist rates rose in 25 of 32 states (median +5.4%), and other road injuries rose in 25 of 32 states (median +2.8%). In contrast, pedestrian rates declined in 26 of 32 states (median -2.3%) and motor-vehicle rates declined in 21 of 32 states (median -1.1%). The aggregate road-injury rate rose marginally (median +1.9%) and increased in 21 of 32 states.

Figure 2: State-level vulnerable-road-user injury DALY rates by mode, India, 2023. Each panel shows the 32 state-level age-standardised DALY rates for one vulnerable-road-user mode, ranked from lowest to highest. The right-hand panel (motorcyclist) shows the largest absolute spread.

State-level rebound regression of the 2023 minus 2020 change on the 2020 baseline rate did not yield statistically significant slopes for any vulnerable-road-user mode: pedestrian beta = -0.0176 (95% CI -0.0807 to 0.0455; p = 0.584); cyclist beta = 0.0252 (95% CI -0.0139 to 0.0642; p = 0.207); motorcyclist beta = 0.0031 (95% CI -0.0502 to 0.0564; p = 0.909). State burden positions therefore persisted through the pandemic with no statistically detectable convergence or divergence.

Inter-state inequality by mode

Inequality across states was greatest for motorcyclist injuries (Gini 0.238; coefficient of variation 8.3-fold range; top 20% of states held 31.4% of total motorcyclist burden), followed by cyclist injuries (Gini 0.253; range 6.1-fold) and pedestrian injuries (Gini 0.189; range 5.0-fold) (Table 3, Figure 3,4). Motor-vehicle injury rates were the most evenly distributed (Gini 0.169; range 3.7-fold). Lorenz curves confirmed that motorcyclist burden was disproportionately concentrated in a small group of high-burden states, while motor-vehicle burden tracked closer to the line of equality.

Cross-validation against Ministry of Road Transport and Highways data

For the ten high-burden states with comparable 2023 administrative road-death counts, the Spearman rank correlation between the GBD road-injury DALY rank and the Ministry of Road Transport and Highways road-death rank was 0.65 (p = 0.043); the Pearson correlation between absolute counts was 0.78 (p = 0.008) (Table 4). Concordance was strong at the extremes (Uttar Pradesh ranked first in both systems) but moderate in the middle of the distribution (Bihar, second by GBD modelled DALYs, ranked seventh in administrative road-death counts), consistent with under-registration of rural road-injury fatalities and survivors with severe disability in administrative systems and supporting the use of GBD modelled estimates as the analytical primary.

Figure 3: State-level VRU DALY rates, 2020 vs 2023, by mode. Each point is one state. Points above the dashed no-change line indicate increased burden in 2023. Cyclist DALY rates rose in 30 of 32 states; motorcyclist rates rose in 25 of 32 states; pedestrian rates declined in most states.

Figure 4: Lorenz curves of state-level DALY rate concentration by mode, India, 2023. Curves further from the line of equality indicate greater concentration of total mode-specific DALY burden in a small group of states. Motorcyclist injury shows the greatest concentration; motor-vehicle injury is distributed closest to equality.

Discussion

This is, to our knowledge, the first sub-national mode-stratified analysis of road-injury DALY burden across all Indian states and union territories through the COVID-19 pandemic recovery period. Three findings emerge.

First, vulnerable-road-user mechanisms dominate the road-injury burden in India and have not changed in modal composition through the pandemic. The median state-level VRU share of 74.7% in 2023 is closely consistent with the World Health Organization Global Status Report on Road Safety, which estimates that vulnerable road users account for more than half of road-traffic deaths globally and a higher fraction in low- and middle-income countries.(1,5,9) The dominance of motorcyclist injury (median 42.2% of state-level burden) reflects the rapid two-wheeler motorisation of India over the past two decades and is consistent with prior Indian and South Asian estimates that placed motorcyclist deaths at one-third or more of all road fatalities.(4,9,17,22) These findings support a fundamental reframing of national road-safety investment around motorcyclist-specific interventions: helmet enforcement and quality, anti-lock braking system mandates for new motorcycles, lane discipline and dedicated lane infrastructure where feasible, structured rider licensing, and post-crash care directed at the high-energy crash physics characteristic of two-wheeler injury.(6,7,9,23,24)

Second, the cyclist trajectory is anomalous and policy-relevant. Cyclist DALY rates rose in 30 of 32 states between 2020 and 2023 (median +5.4%), opposite in direction to pedestrian and motor-vehicle burden. This pattern is consistent with global observations that cycling exposure rose during the pandemic and did not retreat fully to pre-pandemic levels in the recovery phase, while road infrastructure remained calibrated to motorised traffic.(25,26) For India this implies that state surveillance must capture cyclist injury, that segregated cycling infrastructure should be funded as part of the broader push for non-motorised transport in cities, and that mixed-traffic exposure assessments should be incorporated into urban transport planning.(26,27)

Third, motorcyclist injury rates are the most unequally distributed across Indian states (Gini 0.238; 8.3-fold range), and state burden positions persisted through the pandemic. The combination of high modal contribution and high inter-state inequality identifies motorcyclist injury as the modal stratum where state-stratified policy can be expected to deliver the largest national gain. The non-significant rebound regression slopes indicate that the pandemic did not equalise state burden, providing direct evidence against a passive "natural recovery" hypothesis. State-level VRU policy intensification, rather than uniform national targets, is therefore required to meet the United Nations Decade of Action 2021-2030 ambition of halving road-traffic deaths and serious injuries by 2030.(5)

Cross-validation against the Ministry of Road Transport and Highways data showed moderate-to-strong concordance with the GBD modal ranking (Spearman rho = 0.65, p = 0.043; Pearson r = 0.78). Discordance in the middle of the distribution, particularly for Bihar, is consistent with prior Indian and global findings that police-recorded road-death systems systematically underweight rural fatalities and non-fatal injury sequelae.(21,28)

The study has limitations. First, it is ecological and cannot support individual-level causal inference. Second, GBD 2021 estimates are modelled and depend on input data availability, cause attribution and disability weights; the modal composition reported here should be interpreted accordingly. Third, the 2020 baseline corresponds to a national lockdown year and therefore reflects sharply reduced motorised exposure rather than typical pre-pandemic burden; the rebound regression should be read in that light. Fourth, the cross-validation against Ministry of Road Transport and Highways data is restricted to ten states for which calendar-year 2023 counts are publicly reported. Fifth, the analytical units include a combined Jammu and Kashmir and Ladakh entity and an aggregated "Other Union Territories" unit, neither of which corresponds exactly to current administrative boundaries.

Strengths include the use of fully reproducible publicly available data; the first sub-national mode-stratified GBD analysis through the pandemic recovery period for India; the explicit linkage of state-level burden patterns to actionable VRU policy levers; the inclusion of state-level rebound regression and Lorenz-based inequality analysis; and the cross-validation against the principal Indian administrative road-death source.

Implications for policy and practice are direct. State-stratified VRU policy should prioritise (i) motorcyclist protection in the high-burden mortality-prominent cluster identified here, with helmet enforcement, anti-lock braking system adoption, and motorcyclist-aware emergency-response capacity; (ii) cyclist surveillance and infrastructure expansion across all states given the broad-based rise in cyclist injury; and (iii) pedestrian-environment design in the urban states where pedestrian DALY burden remains disproportionately high. Routine state-level road-safety reports should publish modal composition alongside aggregate counts to enable accountability against the Decade of Action targets.(5,9,29,30)

Conclusion

Vulnerable-road-user injury, dominated by motorcyclist burden, accounts for approximately three-quarters of state-level road-injury DALYs in India. Cyclist injury rose in nearly all states between 2020 and 2023, opposite in direction to motor-vehicle and pedestrian burden. Inter-state inequality is greatest for motorcyclist injury, and state burden positions persisted through the pandemic. State-stratified VRU policy targeting motorcyclist protection in high-burden states, cyclist infrastructure nationally, and pedestrian-environment redesign in urban states is required to meet the Decade of Action 2021-2030 targets.

Acknowledgements

The author acknowledges the Institute for Health Metrics and Evaluation, the Ministry of Road Transport and Highways, and the National Crime Records Bureau for making the underlying datasets publicly available.

Data availability: All source datasets are publicly available from the Global Burden of Disease Results Tool (Institute for Health Metrics and Evaluation), the Ministry of Road Transport and Highways (Road Accidents in India 2023) and the National Crime Records Bureau (Accidental Deaths and Suicides in India 2023). Processed analytical files and analysis code that support the findings of this study are available from the corresponding author on reasonable request.

Prior publication: This work has not been published previously and is not under consideration for publication elsewhere, in whole or in part.

Use of artificial intelligence: Generative artificial-intelligence tools were used solely for language polishing and formatting checks. The author verified all data, calculations, interpretations and references and bears full responsibility for the scientific content of the manuscript.

References

1. World Health Organization. Global status report on road safety 2023. Geneva: World Health Organization; 2023. Available from: https://www.who.int/publications/i/item/9789240086517. Accessed May 5, 2026.
2. GBD 2019 Diseases and Injuries Collaborators. Global burden of 369 diseases and injuries in 204 countries and territories, 1990-2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 2020;396:1204-22. doi: 10.1016/S0140-6736(20)30925-9.
3. India State-Level Disease Burden Initiative Road Injury Collaborators. Mortality due to road injuries in the states of India: the Global Burden of Disease Study 1990-2017. Lancet Public Health 2020;5:e86-e98. doi: 10.1016/S2468-2667(19)30246-4.
4. GBD 2021 Diseases and Injuries Collaborators. Global incidence, prevalence, years lived with disability, disability-adjusted life-years, and healthy life expectancy for 371 diseases and injuries in 204 countries and territories and 811 subnational locations, 1990-2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 2024;403:2133-61. doi: 10.1016/S0140-6736(24)00757-8.
5. United Nations General Assembly. Improving global road safety: Resolution A/RES/74/299. New York: United Nations; 2020. Available from: https://digitallibrary.un.org/record/3879711. Accessed May 5, 2026.
6. Liu BC, Ivers R, Norton R, et al. Helmets for preventing injury in motorcycle riders. Cochrane Database Syst Rev 2008;(1):CD004333. doi: 10.1002/14651858.CD004333.pub2.
7. Teoh ER, Campbell M. Role of motorcycle type in fatal motorcycle crashes. J Safety Res 2010;41:507-12. doi: 10.1016/j.jsr.2010.10.005.
8. Retting RA, Ferguson SA, McCartt AT. A review of evidence-based traffic engineering measures designed to reduce pedestrian-motor vehicle crashes. Am J Public Health 2003;93:1456-63. doi:10.1002/14651858.CD004333.pub3
9. Tiwari G, Goel R, Bhalla K. Road safety in India: status report 2023. New Delhi: Transportation Research and Injury Prevention Centre, Indian Institute of Technology Delhi; 2023. Available from: https://tripc.iitd.ac.in/assets/publication/RSI_2023_web.pdf. Accessed May 5, 2026.
10. Peden M, Scurfield R, Sleet D, et al, editors. World report on road traffic injury prevention. Geneva: World Health Organization; 2004. Available from: https://www.who.int/publications/i/item/world-report-on-road-traffic-injury-prevention. Accessed May 5, 2026.
11. Ministry of Road Transport and Highways, Government of India. Road accidents in India 2023. New Delhi: Government of India; 2024. Available from: https://morth.nic.in/en/road-accidents-india-2023. Accessed May 5, 2026.
12. National Crime Records Bureau, Ministry of Home Affairs, Government of India. Accidental deaths and suicides in India 2023. New Delhi: Government of India; 2024. Available from: https://www.data.gov.in/catalog/accidental-deaths-suicides-india-adsi-2023. Accessed May 5, 2026.
13. Hsiao M, Malhotra A, Thakur JS, et al. Road traffic injury mortality and its mechanisms in India: nationally representative mortality survey of 1.1 million homes. BMJ Open 2013;3:e002621. doi: 10.1136/bmjopen-2013-002621.
14. Calderon-Anyosa RJC, Bilal U, Kaufman JS. Variation in non-external and external causes of death in Peru in relation to the COVID-19 lockdown. Yale J Biol Med 2021;94:23-40.
15. Stiles J, Kar A, Lee J, et al. 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;2677:15-27. doi: 10.1177/03611981211044454.
16. Vingilis E, Beirness DJ, Boase P, et al. Coronavirus disease 2019: what could be the effects on road safety? Accid Anal Prev 2020;144:105687. doi: 10.1016/j.aap.2020.105687.
17. Bhalla K, Sriram V, Arora R, et al. The care and transport of trauma victims by layperson emergency medical systems: a qualitative study in Delhi, India. BMJ Glob Health 2019;4:e001963. doi: 10.1136/bmjgh-2019-001963.
18. Institute for Health Metrics and Evaluation. Global Burden of Disease Results Tool. Seattle, WA: IHME; 2024. Available from: https://vizhub.healthdata.org/gbd-results. Accessed May 5, 2026.
19. von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453-7. doi: 10.1016/S0140-6736(07)61602-X.
20. Mackenbach JP, Kunst AE. Measuring the magnitude of socio-economic inequalities in health: an overview of available measures illustrated with two examples from Europe. Soc Sci Med 1997;44:757-71. doi: 10.1016/S0277-9536(96)00073-1.
21. Bhalla K, Naghavi M, Shahraz S, et al. Building national estimates of the burden of road traffic injuries in developing countries from all available data sources: Iran. Inj Prev 2009;15:150-6. doi: 10.1136/ip.2008.020826.
22. Gururaj G. Road traffic deaths, injuries and disabilities in India: current scenario. Natl Med J India 2008;21:14-20.
23. Roy N, Murlidhar V, Chowdhury R, et al. Where there are no emergency medical services-prehospital care for the injured in Mumbai, India. Prehosp Disaster Med 2010;25:145-51. doi: 10.1017/S1049023X00007883.
24. Naci H, Chisholm D, Baker TD. Distribution of road traffic deaths by road user group: a global comparison. Inj Prev 2009;15:55-9. doi: 10.1136/ip.2008.018721.
25. Buehler R, Pucher J. COVID-19 impacts on cycling, 2019-2020. Transp Rev 2021;41:393-400. doi: 10.1080/01441647.2021.1914900.
26. Aldred R, Goodman A. Predictors of the frequency and subjective experience of cycling near misses: findings from the first two waves of the Near Miss Project. Accid Anal Prev 2018;110:161-70. doi: 10.1016/j.aap.2017.09.015.
27. Tiwari G. Progress in pedestrian safety research. Int J Inj Contr Saf Promot 2020;27:35-43. doi: 10.1080/17457300.2020.1720255.
28. Hyder AA, Sugerman DE, Puvanachandra P, et al. Global childhood unintentional injury surveillance in four cities in developing countries: a pilot study. Bull World Health Organ 2009;87:345-52. doi: 10.2471/BLT.08.055798.
29. Mock C, Kobusingye O, Joshipura M, et al. Strengthening trauma and critical care globally. Curr Opin Crit Care 2005;11:568-75. doi: 10.1097/01.ccx.0000186373.49320.65.
30. Reynolds TA, Stewart B, Drewett I, et al. The impact of trauma care systems in low- and middle-income countries. Annu Rev Public Health 2017;38:507-32. doi: 10.1146/annurev-publhealth-032315-021412.