Global, regional, and national burden of lower-limb fractures, 1990 to 2021, with health inequalities and projections to 2040: An observational analysis of Global Burden of Disease Study 2021 data

Abstract

Lower-limb fractures are important contributors to disability worldwide, but their long-term burden, Sociodemographic Index (SDI)-related inequalities, and future trajectories remain incompletely characterized. This study assessed the global, regional, and national burden of lower-limb fractures from 1990 to 2021 and projected trends to 2040 using data from the Global Burden of Disease Study 2021. Data on years lived with disability due to lower-limb fractures were obtained from the Global Burden of Disease Study 2021 for 204 countries and territories. Four anatomical subtypes were analyzed: fracture of the hip (FH); fracture of the patella, tibia, fibula, or ankle (FPTFA); fracture of the foot bones excluding the ankle (FFB); and fracture of the femur excluding the femoral neck (FF). Age-standardized rates (ASRs) and YLD numbers were assessed by sex, age, year, location, and SDI. Joinpoint regression, decomposition analysis, the Slope Index of Inequality, Concentration Index, and Bayesian age-period-cohort models were used to evaluate temporal trends, drivers of absolute burden, SDI-related inequalities, and projections to 2040. In 2021, FPTFA had the largest global YLD burden, with 16.58 million years lived with disability, followed by FH with 3.10 million, FF with 1.18 million, and FFB with 0.41 million. From 1990 to 2021, global YLD numbers increased for all 4 subtypes, by 75.2% for FH, 43.0% for FPTFA, 35.5% for FFB, and 56.7% for FF. In contrast, global ASRs declined by 19.7%, 23.1%, 26.5%, and 18.2%, respectively. Slope Index of Inequality values decreased across all subtypes, indicating narrowing absolute SDI-related differences, whereas Concentration Index values changed less markedly, suggesting persistent relative distributional inequality. Joinpoint regression showed declining global ASRs, with average annual percentage changes of −0.71 for FH, −0.83 for FPTFA, −0.99 for FFB, and −0.65 for FF. Bayesian age-period-cohort projections suggested broadly stable or slightly declining global ASRs through 2040, with sex-specific patterns differing by subtype. Although global ASRs for lower-limb fracture-related disability declined from 1990 to 2021, the absolute YLD burden increased across all 4 subtypes. These findings support subtype- and context-specific strategies, including osteoporosis management, fall prevention, fracture liaison services, and rehabilitation in aging high-SDI settings, alongside strengthened trauma care, timely orthopedic treatment, rehabilitation, and financial protection in lower-SDI settings.

1. Introduction

Skeletal fractures are a major public health problem and an important cause of disability worldwide. Lower-limb fractures are particularly important because they directly impair mobility, independence, and quality of life, and often require surgery, rehabilitation, and long-term care.^[1,2] As population aging accelerates, the burden of fragility fractures related to osteoporosis, falls, and frailty is expected to increase further.^[1,3–5] Because lower-limb fractures are more often disabling than fatal, years lived with disability (YLDs) provide a clinically relevant measure of their population health impact.^[2]

Previous studies have described the burden of fractures at global, regional, or national levels, but several gaps remain. First, many analyses have focused on incidence, prevalence, or all-site fractures, with less emphasis on YLDs attributable to specific lower-limb fracture subtypes.^[3,6,7] Second, the drivers of changes in the absolute number of YLDs remain unclear because increasing YLDs may reflect population growth, population aging, epidemiological changes, or their combined effects.^[4,8] Third, inequalities across the Sociodemographic Index (SDI) spectrum have not been fully characterized using measures that distinguish absolute from relative inequality.^[2,9] Finally, updated projections are needed to guide prevention, trauma care, rehabilitation planning, and resource allocation.

To address these gaps, we used data from the Global Burden of Disease Study 2021 (GBD 2021) to assess the global, regional, and national burden of lower-limb fractures from 1990 to 2021 and to project trends to 2040. Four anatomical subtypes were analyzed: fracture of the hip (FH); fracture of the patella, tibia, fibula, or ankle (FPTFA); fracture of the foot bones excluding the ankle (FFB); and fracture of the femur excluding the femoral neck (FF). Joinpoint regression was used to evaluate temporal trends in age-standardized YLD rates. Decomposition analysis was used to quantify the contributions of population growth, population aging, and epidemiological changes, represented by changes in age-specific YLD rates, to changes in the absolute number of YLDs. The Slope Index of Inequality (SII) and Concentration Index were used to assess absolute and relative SDI-related inequalities, respectively. Bayesian age-period-cohort (BAPC) models were applied to project future age-standardized YLD rates through 2040.

2. Methods

2.1. Data sources and definitions

Data were obtained from GBD 2021 through the Global Health Data Exchange GBD Results Tool. GBD 2021 provides standardized estimates of disease and injury burden by age, sex, year, and location for 204 countries and territories from 1990 to 2021.^[10]

We extracted estimates for lower-limb fractures from 1990 to 2021. Four anatomical subtypes were included: FH, FPTFA, FFB, and FF. Estimates were retrieved by sex, age group, year, country or territory, GBD region, and SDI region.

2.2. Outcome measures and stratification

Because lower-limb fractures are more often disabling than fatal, YLDs were selected as the primary outcome. We analyzed both the absolute number of YLDs and age-standardized YLD rates. The absolute number of YLDs reflects the total disability burden in a population, whereas age-standardized rates (ASRs) allow comparisons across locations and over time after accounting for differences in population age structure. ASRs were expressed per 100,000 population. GBD-derived YLD numbers and ASRs were reported with 95% uncertainty intervals.^[10]

Analyses were stratified by sex, age group, year, country or territory, GBD region, and SDI region. SDI is a composite measure of sociodemographic development based on lag-distributed income per capita, mean educational attainment among individuals aged 15 years or older, and total fertility rate among individuals younger than 25 years; it ranges from 0 to 1, with higher values indicating higher development.^[11] Locations were grouped into low-, low-middle-, middle-, high-middle-, and high-SDI categories.

2.3. Statistical analysis overview

The analytical framework was designed to address 4 objectives: describing temporal trends, identifying drivers of changes in absolute burden, assessing SDI-related inequalities, and projecting future burden. Joinpoint regression was used to evaluate changes in age-standardized YLD rates over time and to identify periods with significant shifts in trends. Decomposition analysis was used to quantify the contributions of population growth, population aging, and epidemiological changes to changes in the absolute number of YLDs. The SII and Concentration Index were used to assess absolute and relative SDI-related inequalities, respectively. BAPC models were used to project future burden through 2040. Mathematical details of the decomposition analysis, inequality metrics, Joinpoint-derived trend estimates, and BAPC model specification are provided in the Supplementary Methods 1 to 5, Supplemental Digital Content 1.

Statistical analyses and visualizations were performed using R software (version 4.4.1; R Foundation for Statistical Computing) and the Joinpoint Regression Program (version 5.1.0; Statistical Research and Applications Branch, Surveillance Research Program, National Cancer Institute). Statistical significance was defined as a two-sided P value < .05, where applicable.

2.4. Cross-country inequality analysis

To evaluate inequalities in fracture-related disability across the SDI spectrum, we assessed age-standardized YLD rates across 204 countries and territories using 2 complementary measures: the SII and the Concentration Index.^[12] The SII captures the absolute difference in burden across the SDI distribution, whereas the Concentration Index describes the relative distribution of age-standardized YLD rates among countries and territories ranked by SDI. Considering both measures allowed us to distinguish changes in absolute gaps from persistent relative patterning across levels of sociodemographic development.

For the SII analysis, countries and territories were ordered by SDI and assigned a relative rank according to the midpoint of their cumulative population distribution. Age-standardized YLD rates were then modeled as a function of this SDI-related rank using linear regression. Positive SII values indicate higher age-standardized YLD rates in higher-SDI settings, whereas negative values indicate higher rates in lower-SDI settings. The magnitude of the SII reflects the degree of absolute inequality.

Relative inequality was further examined using the Concentration Index. Empirical concentration curves were constructed by plotting the cumulative fraction of age-standardized YLD rates against the cumulative fraction of the population ranked by SDI. The index was calculated from the area between the concentration curve and the line of equality. A value of 0 represents no relative inequality, and larger absolute values indicate stronger relative inequality. The sign of the index indicates whether the burden is relatively concentrated toward the lower or higher end of the SDI distribution. SII and Concentration Index values were presented as point estimates.

2.5. Decomposition analysis

Decomposition analysis was used to identify the drivers of changes in the absolute number of YLDs from 1990 to 2021. This analysis partitioned the net change in YLDs into 3 components: population growth, population aging, and epidemiological changes.^[13] Population growth represents changes attributable to variation in total population size; population aging represents changes attributable to shifts in the age structure; and epidemiological changes represent changes in age-specific YLD rates after accounting for population size and age structure.

The analysis was conducted for each lower-limb fracture subtype and by SDI region. By separating demographic from epidemiological components, this approach allowed us to determine whether the increasing absolute YLD burden was mainly driven by population expansion, aging, or changes in age-specific fracture-related disability rates.

2.6. Joinpoint regression analysis

Joinpoint regression analysis was used to characterize temporal trends in age-standardized YLD rates from 1990 to 2021 and to identify periods in which trends changed significantly. This method was selected because it allows nonlinear temporal patterns to be summarized as connected log-linear segments rather than assuming a single constant trend over the entire study period.^[14]

For each lower-limb fracture subtype, Joinpoint models were fitted to annual age-standardized YLD rates at the global and SDI-stratified levels. The annual percentage change was calculated for each trend segment, and the average annual percentage change (AAPC) was calculated to summarize the overall trend from 1990 to 2021. A decreasing trend was defined as an AAPC with a 95% confidence interval (CI) below 0, whereas an increasing trend was defined as an AAPC with a 95% CI above 0. Trends with 95% CIs crossing 0 were considered statistically stable. Analyses were performed using the Joinpoint Regression Program, version 5.1.0.^[15]

2.7. BAPC projection

BAPC models were used to project age-standardized YLD rates from 2022 to 2040. This approach was selected because it incorporates age, period, and cohort patterns and allows uncertainty in future estimates to be quantified within a Bayesian framework.^[16] Age-specific YLD numbers and corresponding population denominators from 1990 to 2021 were used as model inputs. Projected population data for 2022 to 2040 were used to generate future estimates. Separate models were fitted for males, females, and both sexes.

Model implementation was performed using the BAPC and integrated nested Laplace approximation packages in R. Random-walk smoothing priors were applied to model temporal effects, and parameter estimation was performed using the integrated nested Laplace approximation.^[17] Projected age-specific YLD rates were standardized using the GBD standard population to obtain projected age-standardized YLD rates per 100,000 population. Projected estimates were reported with 95% credible intervals (CrIs), derived from the 2.5th and 97.5th percentiles of the posterior distribution.

Projection estimates were generated under the assumption that historical age-, period-, and cohort-specific patterns were informative for future trends.

2.8. Ethics approval and consent to participate

This study used publicly available, aggregated, and deidentified data from GBD 2021. The GBD study was reviewed and approved by the Institutional Review Board of the University of Washington, and informed consent was waived because no identifiable individual-level data were used. Therefore, no additional ethical approval or individual informed consent was required for the present analysis.

3. Results

3.1. Global, regional, and national burden in 2021

In 2021, lower-limb fractures accounted for a substantial global YLD burden across all 4 anatomical subtypes. FPTFA had the largest absolute burden, with 16.58 million YLDs globally, followed by FH with 3.10 million YLDs, FF with 1.18 million YLDs, and FFB with 0.41 million YLDs (Tables 1–4).

Table 1.

YLDs and age-standardized YLD rates for fracture of the hip in 1990 and 2021, with percentage changes from 1990 to 2021, by location.

Location	YLDs in 1990 (95% UI)	ASR in 1990, per 100,000 (95% UI)	YLDs in 2021 (95% UI)	ASR in 2021, per 100,000 (95% UI)	Percentage change in YLDs, 1990–2021 (%)	Percentage change in ASR, 1990–2021 (%)
Global	1,770,882 (1,250,541–2,346,921)	46.91 (33.29–62.39)	3,101,921 (2,176,548–4,193,623)	37.65 (26.41–50.88)	75.2	−19.7
High SDI	576,102 (402,763–774,719)	52.62 (36.8–70.8)	1,107,778 (741,585–1,526,040)	48.33 (32.58–66.25)	92.3	−8.2
High-middle SDI	405,521 (286,881–537,891)	43.34 (30.63–57.67)	516,518 (355,853–708,506)	27.43 (18.94–37.61)	27.4	−36.7
Middle SDI	410,874 (286,284–547,486)	39.33 (27.72–52.36)	723,521 (506,519–979,082)	29.76 (20.8–40.36)	76.1	−24.3
Low-middle SDI	274,577 (190,720–369,517)	44.56 (31.37–59.44)	539,444 (381,221–722,720)	41.19 (29.14–54.96)	96.5	−7.6
Low SDI	101,585 (66,224–151,808)	38.66 (26.5–52.81)	211,979 (144,073–303,323)	39.63 (27.45–53.73)	108.7	2.5
Andean Latin America	7718 (5202–10,574)	31.3 (21.41–42.34)	12,322 (8686–16,335)	20.68 (14.65–27.29)	59.7	−33.9
Australasia	11,438 (7785–15,881)	51 (34.76–71.15)	39,837 (27,159–55,146)	66.96 (45.27–92.75)	248.3	31.3
Caribbean	8401 (5839–11,189)	33 (23.08–44)	20,486 (13,810–28,990)	38.46 (25.88–55.28)	143.9	16.5
Central Asia	13,114 (9094–17,594)	25.17 (17.62–33.7)	15,418 (10,703–20,494)	17.98 (12.5–23.82)	17.6	−28.6
Central Europe	84,903 (58,888–113,717)	62.15 (43.08–83.84)	64,318 (43,231–89,023)	29.62 (19.84–41.33)	−24.2	−52.3
Central Latin America	50,349 (34,612–67,565)	53.84 (37.56–71.85)	69,230 (48,731–92,413)	28.31 (19.96–37.84)	37.5	−47.4
Central Sub-Saharan Africa	8344 (5495–12,108)	28.47 (19.53–38.11)	19,125 (12,616–28,458)	29.53 (20.03–40.94)	129.2	3.7
East Asia	278,684 (197,343–374,505)	33.2 (23.5–44.65)	429,804 (292,003–598,544)	22.23 (15.03–31.03)	54.2	−33.0
Eastern Europe	102,116 (72,097–134,562)	38.36 (27–50.75)	74,158 (52,038–101,349)	23.25 (16.11–32.11)	−27.4	−39.4
Eastern Sub-Saharan Africa	35,346 (21,192–59,929)	32.05 (21.58–46.24)	55,557 (36,180–84,808)	28.82 (19.77–39.8)	57.2	−10.1
High-income Asia Pacific	81,189 (57,336–108,390)	42.15 (29.82–56.45)	172,695 (115,730–239,038)	32.66 (21.92–45)	112.7	−22.5
High-income North America	149,036 (100,751–205,577)	41.72 (28.31–57.48)	379,011 (252,791–526,745)	55.46 (37.13–77.18)	154.3	32.9
North Africa and Middle East	84,108 (54,822–124,291)	37.92 (25.77–52.7)	125,913 (86,459–178,913)	25.98 (18.19–35.57)	49.7	−31.5
Oceania	859 (601–1159)	30.84 (21.4–41.13)	2692 (1852–3621)	40.76 (28.69–54.25)	213.4	32.2
South Asia	280,075 (196,015–375,463)	52.9 (37.02–70.67)	675,606 (479,784–903,431)	52 (36.62–69.2)	141.2	−1.7
Southeast Asia	100,489 (67,408–139,382)	37.39 (25.88–49.73)	169,712 (117,385–232,395)	29.17 (20.35–39.6)	68.9	−22.0
Southern Latin America	26,974 (19,202–35,386)	60.74 (43.35–79.58)	35,568 (25,483–46,804)	40.51 (29.14–53.17)	31.9	−33.3
Southern Sub-Saharan Africa	9281 (6331–12,312)	28.78 (19.99–37.76)	9921 (6873–13,065)	15.89 (11.1–20.91)	6.9	−44.8
Tropical Latin America	46,528 (32,808–61,791)	48.98 (34.62–64.94)	89,832 (64,115–119,393)	36.13 (25.74–47.99)	93.1	−26.2
Western Europe	369,103 (256,361–500,061)	62.98 (43.81–85.3)	588,982 (399,750–815,049)	54.39 (36.73–74.92)	59.6	−13.6
Western Sub-Saharan Africa	22,828 (15,789–30,567)	25.36 (17.52–33.77)	51,734 (36,116–69,281)	26 (18.08–34.76)	126.6	2.5

Data are presented as point estimates with 95% UIs. ASR indicates age-standardized YLD rate per 100,000 population. Percentage change was calculated as $\$(\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{2021}-\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990})/\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990}\times 100\%\$$.

ASR = age-standardized rate, UI = uncertainty interval, YLDs = years lived with disability.

Table 4.

YLDs and age-standardized YLD rates for fracture of femur, other than femoral neck in 1990 and 2021, with percentage changes from 1990 to 2021, by location.

Location	YLDs in 1990 (95% UI)	ASR in 1990, per 100,000 (95% UI)	YLDs in 2021 (95% UI)	ASR in 2021, per 100,000 (95% UI)	Percentage change in YLDs, 1990–2021 (%)	Percentage change in ASR, 1990–2021 (%)
Global	754,039 (494,852–1,083,057)	17.28 (11.38–25.01)	1,181,398 (779,928–1,699,881)	14.14 (9.34–20.38)	56.7	−18.2
High SDI	207,656 (134,843–302,216)	20.23 (13.18–29.47)	275,278 (179,555–400,940)	15.32 (9.9–22.44)	32.6	−24.3
High-middle SDI	209,999 (137,318–305,506)	20.63 (13.51–30.05)	275,415 (179,433–399,000)	15.7 (10.27–22.74)	31.2	−23.9
Middle SDI	192,033 (126,331–273,972)	14.89 (9.78–21.65)	354,486 (232,934–509,094)	13.79 (9.12–19.86)	84.6	−7.4
Low-middle SDI	105,672 (70,069–151,128)	13.82 (9.11–19.94)	196,562 (129,199–284,707)	12.99 (8.55–18.67)	86.0	−6.0
Low SDI	37,635 (25,769–54,434)	11.61 (7.9–16.7)	78,442 (54,593–110,875)	11.55 (7.98–16.34)	108.4	−0.5
Andean Latin America	3625 (2455–5228)	12.4 (8.34–17.58)	6909 (4574–9867)	10.95 (7.22–15.59)	90.6	−11.7
Australasia	4938 (3262–7153)	22.27 (14.8–32.23)	8020 (5329–11,549)	17.15 (11.37–24.96)	62.4	−23.0
Caribbean	3845 (2536–5538)	13.16 (8.65–18.92)	7932 (5297–11,337)	15.3 (10.23–21.89)	106.3	16.3
Central Asia	9165 (5981–13,379)	15.89 (10.38–22.86)	11,675 (7710–16,636)	12.83 (8.45–18.2)	27.4	−19.3
Central Europe	43,927 (28,586–63,678)	32.41 (21.1–46.97)	40,907 (26,818–59,385)	23.08 (15.03–33.61)	−6.9	−28.8
Central Latin America	24,393 (16,361–34,711)	20.69 (13.69–29.65)	36,858 (24,377–52,630)	14.45 (9.56–20.65)	51.1	−30.2
Central Sub-Saharan Africa	3666 (2524–5221)	10.12 (6.92–14.28)	8408 (5833–12,011)	10.02 (6.93–14)	129.4	−1.0
East Asia	134,793 (88,474–196,430)	13.74 (9.03–20.16)	270,189 (175,630–391,233)	13.97 (9.22–20.23)	100.4	1.7
Eastern Europe	78,395 (51,468–113,385)	30.72 (20.02–44.45)	67,144 (44,160–97,230)	23.5 (15.34–33.93)	−14.4	−23.5
Eastern Sub-Saharan Africa	13,851 (8944–20,386)	10.41 (7.12–14.88)	22,057 (15,001–31,602)	8.77 (6.09–12.25)	59.2	−15.8
High-income Asia Pacific	34,380 (22,332–50,244)	17.61 (11.44–25.69)	37,986 (24,893–55,149)	10.48 (6.74–15.32)	10.5	−40.5
High-income North America	65,536 (42,558–95,116)	20.08 (13.1–29.15)	98,701 (63,603–143,834)	17.12 (11–24.83)	50.6	−14.7
North Africa and Middle East	47,094 (31,736–68,477)	18.26 (12.38–26.31)	86,668 (60,321–121,710)	15.66 (10.88–22.01)	84.0	−14.2
Oceania	420 (273–609)	10.55 (6.85–15.44)	1126 (746–1626)	12.27 (8.2–17.83)	168.1	16.3
South Asia	93,779 (61,236–135,545)	14.24 (9.37–20.69)	211,337 (138,743–306,133)	14.5 (9.52–21.02)	125.4	1.8
Southeast Asia	44,291 (29,810–63,903)	13.15 (8.73–18.77)	71,774 (47,820–103,280)	10.97 (7.32–15.75)	62.1	−16.6
Southern Latin America	6680 (4385–9671)	14.26 (9.34–20.63)	10,638 (6970–15,515)	13.27 (8.7–19.43)	59.3	−6.9
Southern Sub-Saharan Africa	5998 (3983–8674)	16.02 (10.58–23.35)	7513 (4888–10,863)	10.64 (6.96–15.45)	25.3	−33.6
Tropical Latin America	23,643 (15,416–34,553)	20.36 (13.41–29.86)	41,641 (27,430–60,141)	16.6 (10.97–24.04)	76.1	−18.5
Western Europe	101,867 (65,682–147,943)	20.13 (13.05–29.28)	111,539 (73,403–162,420)	13.82 (8.99–20.25)	9.5	−31.3
Western Sub-Saharan Africa	9755 (6511–14,176)	8 (5.31–11.57)	22,375 (14,816–32,063)	7.98 (5.35–11.5)	129.4	−0.2

Data are presented as point estimates with 95% UIs. ASR indicates age-standardized YLD rate per 100,000 population. Percentage change was calculated as $\$(\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{2021}-\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990})/\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990}\times 100\%\$$.

ASR = age-standardized rate, UI = uncertainty interval, YLDs = years lived with disability.

Table 2.

YLDs and age-standardized YLD rates for fracture of patella, tibia, fibula, or ankle in 1990 and 2021, with percentage changes from 1990 to 2021, by location.

Location	YLDs in 1990 (95% UI)	ASR in 1990, per 100,000 (95% UI)	YLDs in 2021 (95% UI)	ASR in 2021, per 100,000 (95% UI)	Percentage change in YLDs, 1990–2021 (%)	Percentage change in ASR, 1990–2021 (%)
Global	11,592,216 (7,539,516–16,739,747)	255.77 (166.75–367.91)	16,577,060 (10,892,513–23,822,303)	196.61 (129.22–282.32)	43.0	−23.1
High SDI	3,447,422 (2,230,533–4,966,513)	341.43 (220.28–492.95)	4,266,016 (2,767,768–6,132,517)	267.22 (172.45–384.13)	23.7	−21.7
High-middle SDI	3,364,720 (2,184,049–4,899,964)	320.73 (208.51–466.57)	4,065,248 (2,655,654–5,876,523)	237.79 (154.35–344.07)	20.8	−25.9
Middle SDI	2,759,189 (1,800,982–3,980,589)	199.69 (130.93–286.05)	4,681,482 (3,075,081–6,740,803)	174.31 (114.65–250.47)	69.7	−12.7
Low-middle SDI	1,542,160 (1,009,622–2,227,482)	184.86 (121.69–264.2)	2,614,043 (1,722,455–3,765,434)	156.74 (103.63–224.23)	69.5	−15.2
Low SDI	462,911 (309,117–664,828)	140.69 (94.62–200.59)	933,380 (640,783–1,333,341)	126.33 (86.31–179.27)	101.6	−10.2
Andean Latin America	52,962 (35,520–76,617)	186.76 (125.33–268.01)	111,313 (73,316–160,157)	172.75 (113.86–247.74)	110.2	−7.5
Australasia	105,629 (67,971–152,197)	478.46 (307.49–691.1)	166,581 (106,515–238,542)	413.3 (262.11–594.39)	57.7	−13.6
Caribbean	51,940 (33,960–74,929)	171.76 (112.67–247.11)	91,362 (60,812–129,355)	177.37 (118.11–251.54)	75.9	3.3
Central Asia	163,660 (106,361–237,928)	283.9 (185.63–410.28)	207,868 (136,253–301,403)	221.58 (145.28–320.51)	27.0	−22.0
Central Europe	679,808 (444,113–983,884)	489.01 (318.82–708.23)	592,514 (390,302–852,875)	363.21 (236.81–521.79)	−12.8	−25.7
Central Latin America	388,301 (251,167–558,865)	315.18 (205.34–452.64)	551,776 (358,400–788,596)	211.8 (137.6–302.34)	42.1	−32.8
Central Sub-Saharan Africa	41,944 (28,227–59,866)	117.68 (79.24–167.97)	96,251 (66,026–136,323)	108.65 (74.43–154.66)	129.5	−7.7
East Asia	1,867,914 (1,205,983–2,710,697)	174.58 (113.2–252.9)	3,323,004 (2,165,832–4,814,323)	170.47 (110.78–248.12)	77.9	−2.4
Eastern Europe	1,272,196 (831,824–1,844,663)	494.63 (323.35–717.09)	1,079,501 (710,354–1,566,105)	384.06 (251.78–555.48)	−15.1	−22.4
Eastern Sub-Saharan Africa	136,643 (94,090–194,046)	113.25 (77.25–161.96)	259,843 (177,893–371,515)	95.63 (65.49–136.23)	90.2	−15.6
High-income Asia Pacific	662,132 (429,006–960,999)	337.79 (218.52–489.67)	690,713 (448,720–989,332)	230.07 (147.89–334.15)	4.3	−31.9
High-income North America	963,904 (628,789–1,391,141)	300.37 (195.4–433.55)	1,308,417 (860,131–1,886,762)	239.62 (157.28–344.81)	35.7	−20.2
North Africa and Middle East	623,369 (417,910–894,643)	252.68 (169.02–361.43)	1,211,689 (817,952–1,730,366)	209.43 (141.47–298.32)	94.4	−17.1
Oceania	5721 (3710–8325)	129.76 (84.74–185.52)	15,295 (10,001–22,201)	144.89 (95.17–208.37)	167.3	11.7
South Asia	1,406,974 (908,368–2,037,163)	180.48 (117.56–259.07)	2,682,909 (1,753,441–3,850,045)	161.73 (105.82–231.58)	90.7	−10.4
Southeast Asia	585,718 (389,482–843,951)	166.83 (111.1–238.97)	962,796 (640,376–1,381,774)	135.98 (90.5–194.44)	64.4	−18.5
Southern Latin America	143,375 (93,585–207,923)	301.81 (197.15–437.1)	219,399 (143,021–315,374)	281.51 (183.43–405.62)	53.0	−6.7
Southern Sub-Saharan Africa	66,526 (44,171–96,069)	180.01 (119.74–257.54)	88,488 (58,670–127,174)	123.25 (82.15–177.07)	33.0	−31.5
Tropical Latin America	421,108 (271,188–607,351)	341.5 (221.68–488.63)	633,203 (411,746–908,665)	249.41 (161.8–357.41)	50.4	−27.0
Western Europe	1,823,785 (1,172,612–2,628,272)	382.69 (245.12–552.88)	1,990,720 (1,276,494–2,853,037)	307.68 (193.71–446.49)	9.2	−19.6
Western Sub-Saharan Africa	128,610 (84,528–185,351)	100.82 (66.51–143.94)	293,415 (194,029–423,912)	93.88 (62.48–133.91)	128.1	−6.9

Data are presented as point estimates with 95% UIs. ASR indicates age-standardized YLD rate per 100,000 population. Percentage change was calculated as $\$(\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{2021}-\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990})/\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990}\times 100\%\$$.

ASR = age-standardized rate, UI = uncertainty interval, YLDs = years lived with disability.

Table 3.

YLDs and age-standardized YLD rates for fracture of foot bones except ankle in 1990 and 2021, with percentage changes from 1990 to 2021, by location.

Location	YLDs in 1990 (95% UI)	ASR in 1990, per 100,000 (95% UI)	YLDs in 2021 (95% UI)	ASR in 2021, per 100,000 (95% UI)	Percentage change in YLDs, 1990–2021 (%)	Percentage change in ASR, 1990–2021 (%)
Global	302,545 (180,094–476,917)	6.63 (3.95–10.42)	410,057 (244,113–637,840)	4.87 (2.9–7.58)	35.5	−26.5
High SDI	113,151 (67,104–181,159)	11.33 (6.71–18.22)	128,457 (76,630–202,311)	8.3 (4.95–13.25)	13.5	−26.7
High-middle SDI	82,601 (48,975–130,626)	7.87 (4.67–12.43)	96,184 (57,194–152,838)	5.7 (3.38–9.06)	16.4	−27.6
Middle SDI	61,214 (36,427–95,536)	4.36 (2.59–6.77)	102,633 (61,110–160,305)	3.83 (2.28–5.97)	67.7	−12.2
Low-middle SDI	33,281 (19,802–52,469)	3.9 (2.35–6.12)	56,744 (34,013–89,053)	3.38 (2.04–5.3)	70.5	−13.3
Low SDI	11,946 (6989–18,934)	3.37 (2.04–5.23)	25,652 (15,277–40,653)	3.32 (2.02–5.23)	114.7	−1.5
Andean Latin America	1231 (744–1942)	4.16 (2.49–6.57)	2432 (1437–3777)	3.77 (2.24–5.85)	97.6	−9.4
Australasia	3931 (2309–6369)	17.99 (10.56–29.15)	5675 (3356–9022)	14.66 (8.59–23.64)	44.4	−18.5
Caribbean	1150 (688–1797)	3.77 (2.26–5.88)	2166 (1306–3352)	4.24 (2.56–6.61)	88.3	12.5
Central Asia	3555 (2067–5695)	6.1 (3.54–9.78)	4545 (2709–7206)	4.83 (2.88–7.63)	27.8	−20.8
Central Europe	14,041 (8199–22,488)	10.14 (5.9–16.24)	12,314 (7283–19,601)	7.61 (4.47–12.2)	−12.3	−25.0
Central Latin America	8830 (5271–14,061)	6.98 (4.14–11.07)	12,307 (7380–19,360)	4.73 (2.83–7.42)	39.4	−32.2
Central Sub-Saharan Africa	1062 (658–1658)	2.79 (1.73–4.3)	2674 (1639–4274)	2.91 (1.8–4.58)	151.8	4.3
East Asia	40,835 (23,602–63,578)	3.8 (2.21–5.93)	69,859 (40,681–110,185)	3.6 (2.08–5.66)	71.1	−5.3
Eastern Europe	27,573 (16,086–43,604)	10.75 (6.26–16.99)	23,330 (13,642–36,920)	8.37 (4.86–13.25)	−15.4	−22.1
Eastern Sub-Saharan Africa	4237 (2477–6877)	3.09 (1.86–4.91)	8033 (4764–13,182)	2.87 (1.71–4.72)	89.6	−7.1
High-income Asia Pacific	23,464 (13,763–37,862)	12.03 (7.04–19.41)	22,623 (13,292–36,200)	7.91 (4.67–12.85)	−3.6	−34.2
High-income North America	35,320 (20,700–55,987)	11.16 (6.53–17.76)	43,256 (25,967–68,180)	8.27 (4.96–12.99)	22.5	−25.9
North Africa and Middle East	14,717 (8835–23,435)	5.65 (3.48–8.89)	29,941 (18,517–47,672)	5.11 (3.18–8.11)	103.4	−9.6
Oceania	120 (69–190)	2.69 (1.55–4.3)	326 (190–515)	3.06 (1.8–4.83)	171.7	13.8
South Asia	29,463 (17,206–46,419)	3.73 (2.21–5.81)	56,554 (33,795–88,048)	3.4 (2.04–5.28)	91.9	−8.8
Southeast Asia	13,372 (8021–21,073)	3.68 (2.23–5.87)	21,853 (13,290–33,761)	3.07 (1.87–4.76)	63.4	−16.6
Southern Latin America	5797 (3438–9250)	12.14 (7.21–19.34)	8616 (5137–13,686)	11.21 (6.67–17.87)	48.6	−7.7
Southern Sub-Saharan Africa	1692 (1024–2628)	4.49 (2.71–6.97)	2219 (1338–3463)	3.06 (1.85–4.78)	31.1	−31.8
Tropical Latin America	8607 (5060–13,868)	6.94 (4.13–11.1)	13,193 (7872–20,827)	5.22 (3.12–8.24)	53.3	−24.8
Western Europe	60,525 (35,765–96,713)	12.99 (7.64–20.93)	60,847 (35,735–96,659)	9.83 (5.72–15.91)	0.5	−24.3
Western Sub-Saharan Africa	3024 (1816–4734)	2.3 (1.38–3.58)	7293 (4508–11,191)	2.24 (1.36–3.46)	141.2	−2.6

Data are presented as point estimates with 95% UIs. ASR indicates age-standardized YLD rate per 100,000 population. Percentage change was calculated as $\$(\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{2021}-\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990})/\mathrm{v}\mathrm{a}\mathrm{l}\mathrm{u}{\mathrm{e}}_{1990}\times 100\%\$$.

ASR = age-standardized rate, UI = uncertainty interval, YLDs = years lived with disability.

From 1990 to 2021, global YLD numbers increased for all 4 subtypes, whereas global ASRs declined. The increase in YLDs was 75.2% for FH, 43.0% for FPTFA, 35.5% for FFB, and 56.7% for FF. Over the same period, ASRs decreased by 19.7% for FH, 23.1% for FPTFA, 26.5% for FFB, and 18.2% for FF (Tables 1–4).

Regional patterns varied by subtype. In 2021, high-SDI regions accounted for large absolute YLD burdens for FH and FPTFA, while populous regions such as South Asia and East Asia also contributed substantially to the total burden. Western Europe had the highest ASR for FH, whereas Australasia had the highest ASR for FPTFA. For FFB, Australasia and Western Europe showed among the highest ASRs, while Central Europe and Eastern Europe had relatively high ASRs for FF (Tables 1–4). National distributions of YLD numbers in 2021 are shown in Figure 1.

Figure 1.

National distribution of YLDs due to lower-limb fracture subtypes in 2021. (A) Fracture of the hip. (B) Fracture of the patella, tibia, fibula, or ankle. (C) Fracture of the foot bones excluding the ankle. (D) Fracture of the femur excluding the femoral neck. Values represent the number of YLDs in 2021. Insets show selected regions with small geographic areas. YLDs = years lived with disability.

3.2. SDI-related inequalities

SDI-related inequality analyses showed that ASRs were unevenly distributed across the SDI spectrum for all 4 subtypes (Fig. 2). The SII decreased between 1990 and 2021 for all subtypes, indicating a narrowing of absolute differences in ASRs across countries and territories ranked by SDI. For FH, the SII decreased from 15.77 in 1990 to 9.42 in 2021. Similar decreases were observed for FPTFA, from 250.07 to 198.89; FFB, from 7.53 to 5.42; and FF, from 10.48 to 5.72.

Figure 2.

SDI-related inequalities in age-standardized YLD rates for lower-limb fracture subtypes in 1990 and 2021. Panels (A), (C), (E), and (G) show SII analyses for FH, FPTFA, FFB, and FF, respectively. Panels (B), (D), (F), and (H) show Concentration Index analyses for FH, FPTFA, FFB, and FF, respectively. In the SII plots, countries and territories are ranked by SDI on the x-axis, and the y-axis shows ASRs per 100,000 population. Bubble size represents population size. In the concentration curves, the x-axis represents the cumulative fraction of the population ranked by SDI, and the y-axis represents the cumulative fraction of age-standardized YLD rates. SII reflects absolute SDI-related inequality, whereas the Concentration Index reflects relative distributional inequality. Values of SII and the Concentration Index are presented as point estimates. ASR = age-standardized rate, FF = fracture of the femur excluding the femoral neck, FFB = fracture of the foot bones excluding the ankle, FH = fracture of the hip, FPTFA = fracture of the patella, tibia, fibula, or ankle, SDI = Sociodemographic Index, SII = Slope Index of Inequality, YLDs = years lived with disability.

The Concentration Index showed smaller changes over the same period. For FH, the Concentration Index changed from −0.10 in 1990 to −0.07 in 2021. For FPTFA, it changed from −0.21 to −0.19; for FFB, from −0.25 to −0.21; and for FF, from −0.13 to −0.07. Overall, the SII decreased more markedly than the Concentration Index, suggesting smaller absolute gaps but persistent relative distributional inequality across the SDI spectrum (Fig. 2).

3.3. Decomposition analysis and temporal trends

Decomposition analysis of changes in the absolute number of YLDs showed that population growth contributed positively to the increasing global burden for all 4 lower-limb fracture subtypes from 1990 to 2021 (Figs. 3A–6A). For FH, population growth, population aging, and epidemiological changes all contributed positively to the global increase in YLDs. For FPTFA and FFB, population growth was the main positive contributor in most SDI regions, while population aging contributed less and was negative in some lower-SDI settings. For FF, population growth contributed positively across SDI regions, whereas the contribution of population aging varied by SDI level.

Figure 3.

Decomposition analysis, temporal trends, and projections for FH. (A) Decomposition of changes in the absolute number of YLDs from 1990 to 2021 into contributions from population growth, population aging, and epidemiological changes. (B) Joinpoint regression of ASRs from 1990 to 2021 at the global and SDI-stratified levels. Segment-specific APCs are shown in the panel, and overall AAPCs are provided in Table S1, Supplemental Digital Content 6. (C) BAPC projections of global ASRs from 2022 to 2040, stratified by sex. The vertical dashed line indicates the transition from observed to projected estimates. Shaded areas represent 95% credible intervals. AAPC = average annual percentage change, APC = annual percentage change, ASR = age-standardized rate, BAPC = Bayesian age-period-cohort, FH = fracture of the hip, SDI = Sociodemographic Index, YLDs = years lived with disability.

Figure 6.

Decomposition analysis, temporal trends, and projections for FF. (A) Decomposition of changes in the absolute number of YLDs from 1990 to 2021 into contributions from population growth, population aging, and epidemiological changes. (B) Joinpoint regression of ASRs from 1990 to 2021 at the global and SDI-stratified levels. Segment-specific APCs are shown in the panel, and overall AAPCs are provided in Table S4, Supplemental Digital Content 8. (C) BAPC projections of global ASRs from 2022 to 2040, stratified by sex. The vertical dashed line indicates the transition from observed to projected estimates. Shaded areas represent 95% credible intervals. AAPC = average annual percentage change, APC = annual percentage change, ASR = age-standardized rate, BAPC = Bayesian age-period-cohort, FF = fracture of the femur excluding the femoral neck, SDI = Sociodemographic Index, YLDs = years lived with disability.

Figure 4.

Decomposition analysis, temporal trends, and projections for FPTFA. (A) Decomposition of changes in the absolute number of YLDs from 1990 to 2021 into contributions from population growth, population aging, and epidemiological changes. (B) Joinpoint regression of ASRs from 1990 to 2021 at the global and SDI-stratified levels. Segment-specific APCs are shown in the panel, and overall AAPCs are provided in Table S2, Supplemental Digital Content 7. (C) BAPC projections of global ASRs from 2022 to 2040, stratified by sex. The vertical dashed line indicates the transition from observed to projected estimates. Shaded areas represent 95% credible intervals. AAPC = average annual percentage change, APC = annual percentage change, ASR = age-standardized rate, BAPC = Bayesian age-period-cohort, FPTFA = fracture of the patella, tibia, fibula, or ankle, SDI = Sociodemographic Index, YLDs = years lived with disability.

Figure 5.

Decomposition analysis, temporal trends, and projections for FFB. (A) Decomposition of changes in the absolute number of YLDs from 1990 to 2021 into contributions from population growth, population aging, and epidemiological changes. (B) Joinpoint regression of ASRs from 1990 to 2021 at the global and SDI-stratified levels. Segment-specific APCs are shown in the panel, and overall AAPCs are provided in Table S3, Supplemental Digital Content 17. (C) BAPC projections of global ASRs from 2022 to 2040, stratified by sex. The vertical dashed line indicates the transition from observed to projected estimates. Shaded areas represent 95% credible intervals. AAPC = average annual percentage change, APC = annual percentage change, ASR = age-standardized rate, BAPC = Bayesian age-period-cohort, FFB = fracture of the foot bones excluding the ankle, SDI = Sociodemographic Index, YLDs = years lived with disability.

Joinpoint regression showed declining global ASRs across all 4 subtypes from 1990 to 2021 (Figs. 3B–6B; Tables S1–S8, Supplemental Digital Content 6). The global AAPC was −0.71 (95% CI: −0.77 to −0.65) for FH, −0.83 (95% CI: −0.90 to −0.76) for FPTFA, −0.99 (95% CI: −1.04 to −0.94) for FFB, and −0.65 (95% CI: −0.74 to −0.56) for FF. At the SDI-stratified level, ASRs declined significantly in most SDI regions. Exceptions were observed in low-SDI regions for FH, FFB, and FF, where trends were statistically stable, with AAPCs of 0.05 (95% CI: −0.17 to 0.27), −0.06 (95% CI: −0.16 to 0.04), and −0.05 (95% CI: −0.46 to 0.36), respectively (Tables S1, S3, and S4, Supplemental Digital Content 6).

3.4. Projections through 2040

BAPC projections suggested that global ASRs would remain broadly stable or decline slightly through 2040 across the 4 subtypes, with uncertainty increasing over the projection horizon (Figs. 3C–6C; Tables S9–S12, Supplemental Digital Content 13). For FH, the projected ASR among both sexes decreased from 37.36 per 100,000 in 2022 to 34.50 per 100,000 in 2040 (95% CrI: 14.40–54.60). A sex-specific difference was observed for FH: in 2040, the projected ASR was higher in females than in males, at 40.85 (95% CrI: 14.74–66.96) versus 23.40 (95% CrI: 9.22–37.59) per 100,000.

The projected patterns differed across the other 3 subtypes. For FPTFA, the global ASR among both sexes remained nearly unchanged, from 197.21 per 100,000 in 2022 to 197.47 per 100,000 in 2040 (95% CrI: 104.20–290.75). In contrast to FH, the projected 2040 ASR for FPTFA was higher in males than in females, at 201.53 (95% CrI: 93.51–309.54) versus 174.77 (95% CrI: 94.18–255.36) per 100,000. For FFB, the projected ASR declined slightly from 4.88 per 100,000 in 2022 to 4.73 per 100,000 in 2040 (95% CrI: 2.65–6.80), with higher projected rates in males than in females in 2040 (5.71 [95% CrI: 2.83–8.59] vs 3.51 [95% CrI: 2.02–5.00] per 100,000). For FF, the projected ASR remained approximately stable, from 14.16 per 100,000 in 2022 to 14.00 per 100,000 in 2040 (95% CrI: 6.74–21.26), with projected 2040 rates of 12.92 (95% CrI: 6.21–19.63) in males and 13.87 (95% CrI: 3.58–24.16) in females. Across all 4 subtypes, the widening CrIs toward 2040 indicated increasing uncertainty with longer projection horizons.

4. Discussion

Using standardized estimates from GBD 2021, this study assessed lower-limb fracture-related disability across 204 countries and territories from 1990 to 2021 and projected global ASRs to 2040. A consistent pattern emerged across all 4 anatomical subtypes: global ASRs declined, yet the absolute number of YLDs continued to increase. This divergence was not uniform across fracture sites. Population growth contributed to the rising absolute burden for all subtypes, whereas population aging had a particularly important role in FH and a more variable role in non-hip fractures. Inequality analyses further showed that absolute SDI-related differences narrowed over time, although relative distributional inequality remained. Looking ahead, projected ASRs were broadly stable or slightly declining, but sex-specific patterns differed by fracture subtype.

A central finding of this study is the divergence between declining ASRs and increasing absolute YLDs. Falling ASRs suggest that age-standardized fracture-related disability has decreased over the past 3 decades, a pattern that may reflect improvements in injury prevention, osteoporosis diagnosis and treatment, acute fracture care, and rehabilitation services in many settings.^[1,18,19] Yet these gains have not translated into a lower absolute burden. Population growth, longer survival, and population aging have expanded the number of individuals exposed to fracture risk, thereby increasing the total volume of disability.^[3,4,8] This pattern underscores a practical challenge for health systems: reductions in standardized rates do not necessarily reduce service demand. Even when individual-level risk declines, hospitals, rehabilitation services, long-term care systems, and community-based support programs may still face a growing number of patients with fracture-related disability.

Patterns identified by the decomposition analysis suggest that lower-limb fractures should not be treated as a single homogeneous disease burden. FH was strongly shaped by demographic aging, which is consistent with its close association with osteoporosis, frailty, falls, and functional decline among older adults.^[20,21] In contrast, FPTFA and FFB showed weaker or even negative contributions from population aging in some lower-SDI settings. These patterns are compatible with a larger contribution from high-energy trauma, occupational exposure, road traffic injuries, and activity-related injuries among younger or working-age populations.^[22,23] FF appeared to lie between these 2 profiles, reflecting its mixed etiology: traumatic femoral fractures in younger adults and frailty-related fractures in older populations.^[2,24] These differences matter because prevention strategies that are effective for hip fractures may not be sufficient for distal lower-limb injuries, and vice versa.

The inequality findings require a nuanced interpretation. Declining SII values indicate that absolute gaps in ASRs across the SDI spectrum became smaller between 1990 and 2021. In practical terms, this represents a narrowing of absolute inequality. However, the Concentration Index changed less markedly, suggesting that relative distributional inequality did not disappear.^[9] Narrower absolute differences, therefore, should not be interpreted as equity having been achieved. A more plausible interpretation is that fracture-related disability is increasingly affecting lower-SDI settings, while the relative ordering of burden across sociodemographic development levels remains uneven. This distinction has policy relevance. High-SDI settings continue to face a substantial burden related to population aging, whereas low- and middle-SDI settings may experience rising fracture volume before trauma care, rehabilitation, and financial protection systems are fully prepared.^[25,26]

Projection results add another layer to this picture. FH was projected to remain higher among females than males in terms of ASR, a pattern that likely reflects postmenopausal bone loss, lower bone mineral density, longer female life expectancy, and survival into older age groups with high fracture risk.^[27,28] By contrast, FPTFA and FFB were projected to remain higher among males, consistent with continued exposure to occupational injuries, road traffic injuries, and other high-energy trauma.^[23,29] Health-system factors may amplify these biological and behavioral differences. Limited access to DXA screening, anti-osteoporosis medications, fracture liaison services, and rehabilitation can increase disability after fracture, particularly among older women and in resource-constrained settings.^[19,27,30] These findings support a more targeted prevention approach: osteoporosis screening and fall prevention should be strengthened for older women, while trauma prevention, road safety, and occupational injury control remain essential for reducing non-hip lower-limb fractures, especially among men.^[29,31]

Policy responses should therefore be tailored to both fracture subtype and development context. In aging high-SDI regions, reducing the future burden of FH will require more than treating osteoporosis after diagnosis. Fall-prevention programs, fracture liaison services, early mobilization, rehabilitation, and long-term functional support are all important components of care.^[20,32] In lower-SDI settings, where population growth remains a major driver of absolute burden, strengthening trauma systems may be equally urgent. Timely fracture surgery, access to orthopedic care, rehabilitation services, and financial risk protection should be integrated into universal health coverage frameworks.^[26] Without these investments, declining ASRs may coexist with increasing disability, particularly in rapidly growing and urbanizing populations.

Several strengths support the interpretation of these findings. By using standardized estimates from GBD 2021, this study assessed 4 lower-limb fracture subtypes across global, regional, and national levels over more than 3 decades.^[8] Combining descriptive burden estimates with Joinpoint regression, decomposition analysis, SDI-related inequality metrics, and BAPC projections allowed us to examine not only how the burden changed, but also why it changed and how it may evolve. Important limitations should also be acknowledged. GBD estimates are partly model-based, and uncertainty may be greater in low-SDI settings where primary data are sparse.^[4] Publicly available GBD outputs do not permit detailed disaggregation by injury mechanism, such as falls, road traffic injuries, occupational injuries, or sports-related trauma. SDI-based inequality analysis is ecological by design and may mask within-country disparities by income, residence, ethnicity, or access to care.^[10] Decomposition analysis separates population growth, population aging, and epidemiological changes for analytical clarity, although these factors may interact in real-world settings. Projections to 2040 should be interpreted as trend-based scenarios rather than deterministic forecasts. They assume that historical age-, period-, and cohort-specific patterns remain informative for future trends and cannot fully capture abrupt changes in prevention policies, trauma systems, osteoporosis treatment coverage, conflicts, pandemics, or other disruptive events. Widening CrIs toward 2040 further indicate increasing uncertainty over longer projection horizons.

5. Conclusion

Using GBD 2021 estimates, this study showed that the global absolute YLD burden of lower-limb fractures increased from 1990 to 2021, despite declining ASRs across all 4 anatomical subtypes. Population growth was a major contributor to the increasing absolute burden, while the roles of population aging and epidemiological changes varied by fracture subtype and SDI level. SDI-related inequality analyses showed narrowing absolute differences but persistent relative distributional inequality. Projections to 2040 suggested broadly stable or slightly declining ASRs, with sex-specific patterns differing across fracture subtypes. These findings support a subtype- and context-specific approach to fracture prevention and care: aging high-SDI settings should strengthen osteoporosis management, fall prevention, fracture liaison services, and rehabilitation, particularly for older women at risk of FH, whereas lower-SDI settings should expand trauma care, timely orthopedic treatment, rehabilitation, and financial protection within universal health coverage frameworks to reduce future fracture-related disability.

Acknowledgments

We sincerely acknowledge the work of the Global Burden of Disease Study collaborators and the Institute for Health Metrics and Evaluation for providing access to the GBD 2021 data.

Author contributions

Data curation: Samuhaer Azhati, Shuning Liu, Zhongxu Hu, Kai Zhao, Dixin Zou, Yutong Wu, Yue Dou, Wei Jiang, Chang Liu.

Formal analysis: Samuhaer Azhati, Shuning Liu, Zhongxu Hu, Kai Zhao, Dixin Zou, Yan Wei, Chang Liu.

Supervision: Wei Jiang, Chang Liu.

Validation: Chang Liu.

Writing – original draft: Samuhaer Azhati.

Writing – review & editing: Samuhaer Azhati, Shuning Liu, Wei Jiang, Chang Liu.