Accumulating Fragility Fractures Associate with Worsening Physical Capability and Self-Rated Health in Postmenopausal Women: A 20-Year Cohort Study

1. Introduction

Fracture risk increases with age and is influenced by several well-established factors, including osteoporosis or osteopenia, low body mass index (BMI), female sex, prior fractures, and increased fall propensity (Schini et al., 2024; Shevroja et al., 2021). Fragility fractures typically result from reduced bone density and are often triggered by low-energy incidents such as falls from standing height. Frequently affected anatomical sites include the wrist, humerus, hip, spine, tibia, and fibula (Johnell and Kanis, 2006). Globally, an estimated 37 million osteoporosis-related fractures occur each year in individuals aged 55 years and older, translating to approximately 70 fractures per minute (Wu et al., 2021). A prior fracture increases the risk of sustaining another by 86%, with most refractures occurring within 1–2 years (Kanis et al., 2023; Söreskog et al., 2020). The combination of post-menopausal estrogen loss, higher fracture incidence, impaired healing mechanisms, and greater disability burden makes women both the highest risk population and the most clinically relevant target for osteoporosis research (Ortona et al., 2023).

Fragility fractures are associated with a decline in physical capability. Physical capability is often assessed using standardized tools such as the Instrumental Activities of Daily Living (IADL) scale. Over a 10-year period, declines in the Timed Up and Go (TUG) test and IADL scores have been associated with fragility fractures, suggesting that these measures may help predict fracture risk. Individuals with fragility fractures typically show reduced functional status, increased dependency, higher absenteeism, greater productivity loss, and a higher reliance on both paid and unpaid assistance (Yeh et al., 2023). Moreover, the negative effects of fragility fractures on physical activity are long lasting, with functional deficits persisting for at least four years after injury (Fischer et al., 2017).

Hip fractures are well documented to substantially elevate mortality rates and healthcare costs while diminishing quality of life (Dyer et al., 2016; Haleem et al., 2023). Similarly, older adults with pelvic fractures have reported marked declines in self-rated health (SRH), and early, intensive rehabilitation appears essential for recovery (Hack et al., 2019). Vertebral fractures also contribute to long-term mobility impairments (Huang et al., 1996; Shetty et al., 2020). Regarding upper extremity injuries, recent meta-analytic evidence indicates that wrist fractures may cause clinically significant functional decline and increased healthcare burden for up to 12 years post fracture in middle aged adults (Babatunde et al., 2021). Men over age 65 with distal radius fractures are similarly more likely to experience post fracture disability regardless of radiographic findings (Egund et al., 2020). Evidence for ankle fractures remains mixed; a meta-analysis found that mobility restrictions may persist up to 24 months following fracture, though other studies have reported more limited long-term functional impact (Anderson et al., 2008; Beckenkamp et al., 2014).

Osteoporosis and sarcopenia frequently co-occur and share overlapping pathophysiological mechanisms. Loss of skeletal bone and muscle mass has been associated with frailty and frailty fractures. These changes can further add to the risk of developing chronic pain conditions. Chronic pain is common in the elderly who suffer from osteoporosis with lower back pain being the most prominent finding (Paolucci et al., 2016). Chronic pain and temporary physical limitations resulting from fractures can reinforce subjective feelings of frailty or aging, which in turn may lead to reduced physical activity and accelerate progression toward frailty (Sale et al., 2024).

While numerous longitudinal studies have examined the effects of individual fragility fractures on physical capability and SRH, the cumulative impact of multiple fractures has been less thoroughly investigated. Recent evidence has highlighted the substantial negative impact of fracture accumulation on overall healthcare costs in elderly women (Sund et al., 2026). The aim of the present study was to assess the longitudinal association between cumulative fracture burden and changes in self-reported physical capability (PC) and self-rated health (SRH) over a 20 year follow-up.

2. Methods

We used data from the population-based Kuopio Osteoporosis Risk Factor and Prevention (OSTPRE) study. The original OSTPRE cohort consisted of 14,220 women born between 1932 and 1941 and residing in the Northern Savonia region of Finland in 1989. Initial postal questionnaires yielded responses from 13,100 participants, corresponding to a 92.1% response rate. The 25-year follow-up questionnaire was sent to all 10,785 women alive with an existing address in the population registry at that time, and of these 72.0% (N = 7,767) answered. That follow-up study resulted in a representative sample of women aged 72 to 82 years old living in the region. For the purposes of the current study, we focused on the population who had valid values for physical capability (PC) in the 25-year follow-up questionnaire as well as in the 5-year questionnaire, which was the first containing questions on PC and self-rated health (SRH). We ended up with the cohort consisting of 6,612 women, who were alive in 2014 and had 20-years of backward follow-up data on PC and SRH from the questionnaires in the years 2004 (10-year) and 1994 (baseline) (Figure 1).

Figure 1

Physical capability was assessed in the questionnaires with the question “What is your current level of physical capability?” and response options 1) “Fully mobile”, 2) “Unable to run”, 3) “Can move less than 1,000 meters independently”, 4) “Can move less than 100 meters independently”, 5) “Can only move indoors”, 6) “Temporarily immobile”, 7) “Permanently immobile”. To maintain adequate category sizes, responses 1 and 2 were combined as “good physical capability,” while responses 4–7 were grouped as “poor physical capability.” Response 3 was categorized as “intermediate.” The validity of this self-reported PC scale has been previously established through a subsample of the OSTPRE cohort. Self-reported scores demonstrated significant correlations with objective physical performance metrics, including grip strength, knee extension force, the ability to perform a full squat, and a 10-second unipedal stance test (Juopperi et al., 2021). While these objective assessments were not available for the entire cohort, the high correlation supports the use of the self-reported scale as a reliable proxy.

Self-rated health (SRH) was assessed with the question: “How would you rate your own health compared to others of the same age?” with the response categories 1) “Excellent”, 2) “Good”, 3) “Average”, 4) “Poor”, and 5) “Very poor”. For analysis, responses 1 and 2 were combined into “good”, response 3 was categorized as “intermediate”, and responses 4 and 5 into “poor”. For clarity, only the prevalence of “good” PC and SRH is presented in the main text, while full distributions (good/intermediate/poor) are available in Supplementary Table 1.

For the cumulative fragility fractures we followed the definition in (Sund et al., 2026) and included fractures of wrist, hip, humerus, ankle, and spine. Fractures were identified from validated self-reports, register data and radiological reports (Honkanen et al., 1999; Nissinen et al., 2023, 2025). By combining self-reported fractures that were verified using medical records with fracture cases identified from register data, the study ensures that the majority of fracture events are captured throughout the entire follow-up (Sund et al., 2026).

For descriptive analyses, the study population was divided into groups based on their cumulative number of fractures at the endpoint: no fractures (N = 4,674), one fracture (N = 1,268), two fractures (N = 456), or three or more fractures (N = 214). Participants in the control group could have sustained minor fractures not included in the list; given their low frequency, they were considered unlikely to influence PC or SRH outcomes. Detailed fracture distributions for the one and two fracture groups are reported in Supplementary Figure 2.

3. Results

3.3 Physical capability and self-rated health

The control group (N = 4674) consistently reported the highest physical capability (PC) throughout the entire study. At baseline, the control group reported good PC at 93.9%, which remained relatively stable at 95.2% at the 10-year follow-up (FU) but notably declined to 79.0% at the endpoint (Figure 2).

Figure 2

Women with one fracture (N = 1268) throughout the study reported a PC of 93.7% at baseline, similar to the control group. The two-fractures group (N = 456) also reported good PC at 93.2% at baseline, and the three or more fractures group (N = 214) reported an initial PC of 90.7%. Physical capability remained largely similar in all groups up to the 10-year FU, after which a clear decline was reported in all groups examined.

For the control group, good PC declined by 16.2 percentage points (pp) between the 10- year FU and endpoint, decreasing from 95.2% to 79.0%. Similarly, the one-fracture group experienced a decline of 19.5 pp, from 93.7% to 74.2%. The two fractures group reported a decrease in good PC of 17.7 pp from 94.7% to 77.0% in the same period. The three or more fractures group reported a decline in the prevalence of good PC of 23.4 pp from 94.9% to 71.5% (Supplementary table 1, Figure 2).

The control group reported the highest prevalence of good self-rated health (SRH) throughout the study. At baseline, 47.9% of the women in control group reported good SRH. In the one-fracture group 43.4% of women reported good SRH at baseline. Of the two fractures group members, 46.1% reported good SRH at baseline. Women in the three or more fractures group consistently reported the lowest SRH throughout the study. At baseline, 36.5% of the group members reported good SRH. All groups reported slightly increasing proportion of good SRH at the 10-year FU after which there was a steep decline.

For the control group, good SRH declined by 7.2 pp from 49.9% to 42.7% between the 10- year FU and endpoint. For the one fracture group a decline of 7.2 pp from 45.3% to 38.1% was reported. Similarly, women in the two fractures group reported a decline in good SRH of 11.6 pp, from 47.6% to 36.0%. For the three fractures group a reduction of 7.1 pp, from 40.4% to 33.3% was recorded (Supplementary table 1, Figure 3).

Figure 3

A longitudinal ordinal GEE analysis for PC and SRH is presented in Table 2. The unadjusted model shows a clear increase in odds ratios between cumulative fractures and worse PC. Odds ratios (OR) increased from 1.21 (95% CI: 1.07–1.37) for one fracture to 2.08 (95% CI: 1.58–2.74) for three or more fractures. Adjustment for age and BMI in the second model did not change this association significantly, with OR remaining at 1.21 (95% CI: 1.06–1.37) and slightly increasing to 2.11 (95% CI: 1.58–2.81) for one and three or more fractures respectively. The third model, which includes levels of education and comorbidity burden, reduced the association further. A single fracture showed an OR of 1.16 (95% CI: 1.01–1.32), while two fractures showed an OR of 1.53 (95% CI: 1.23–1.90) and three or more fractures an OR of 1.93 (95% CI: 1.43–2.61). Severe comorbidity (CCI severe vs. none) also had a clear association with worse PC in the third model (OR 5.78, 95% CI: 4.39–7.60) while higher education level was associated with better PC (high education vs. basic education (OR 0.58, 95% CI: 0.45–0.74).

Table 2

Longitudinal ordinal regression analysis (GEE with logit link) for PC and SRH. The first model includes only the cumulative fractures (and time point) as a predictor, the second adds BMI and age, the third model adds level of education (basic, intermediate/upper secondary education, advanced/tertiary education) and Charlson comorbidity index (CCI) scores (none—score 0, mild—score 1–2, moderate—score 3–4, severe—score 5+). Estimates are reported as odds ratios with 95% confidence intervals.

	MODEL 1			MODEL 2			MODEL 3
	ODDS RATIO	95% CI	p-VALUE	ODDS RATIO	95% CI	p-VALUE	ODDS RATIO	95% CI	p-VALUE
Physical capability
No fractures	1	ref.		1	ref.		1	ref.
1 fracture	1.21	1.07–1.37	0.0021	1.21	1.06–1.37	0.0043	1.16	1.01–1.32	0.0292
2 fractures	1.59	1.30–1.94	<0.0001	1.58	1.28–1.95	<0.0001	1.53	1.23–1.90	<0.0001
3 or more fractures	2.08	1.58–2.74	<0.0001	2.11	1.58–2.81	<0.0001	1.93	1.43–2.61	<0.0001
BMI				1.14	1.13–1.16	<0.0001	1.13	1.12–1.14	<0.0001
Age				1.15	1.13–1.17	<0.0001	1.13	1.11–1.15	<0.0001
Education basic							1	ref.
Education intermediate							0.83	0.74–0.93	0.0019
Education high							0.58	0.45–0.74	<0.0001
CCI none							1	ref.
CCI mild							1.79	1.60–2.00	<0.0001
CCI moderate							3.22	2.73–3.80	<0.0001
CCI severe							5.78	4.39–7.60	<0.0001
Self-rated health
No fractures	1	ref.		1	ref.		1	ref.
1 fracture	1.10	0.99–1.22	0.0805	1.09	0.98–1.22	0.1055	1.07	0.96–1.20	0.2059
2 fractures	1.47	1.21–1.79	0.0001	1.47	1.21–1.78	0.0001	1.40	1.16–1.70	0.0006
3 or more fractures	1.31	0.98–1.74	0.0680	1.39	1.04–1.87	0.0270	1.27	0.93–1.73	0.1318
BMI				1.10	1.09–1.11	<0.0001	1.08	1.07–1.09	<0.0001
Age				1.04	1.02–1.05	<0.0001	1.02	1.00–1.04	0.0302
Education basic							1	ref.
Education intermediate							0.64	0.58–0.71	<0.0001
Education high							0.41	0.34–0.50	<0.0001
CCI none							1	ref.
CCI mild							1.89	1.75–2.05	<0.0001
CCI moderate							2.34	2.02–2.71	<0.0001
CCI severe							3.42	2.60–4.49	<0.0001

For SRH, the unadjusted model showed a non-linear association between fracture accumulation and lower SRH. OR increased from 1.10 (95% CI: 0.99–1.22) for one fracture to 1.47 (95% CI: 1.21–1.79) for two fractures but then diminished to 1.31 (95% CI: 0.98–1.74) for three or more fractures. This non-linear association between cumulative fractures and SRH remained in the adjusted models. Model 3 showed that comorbidity burden remained a strong predictor of worse SRH, though the effect sizes were smaller than for PC (CCI severe: OR 3.42, 95% CI: 2.60–4.49). Higher education also demonstrated a protective association with SRH (high education OR 0.41, 95% CI: 0.34–0.50).

4. Discussion

In this study, we investigated the association between cumulative fracture burden and long-term changes in physical capability (PC) and self-rated health (SRH) over a 20-year period in a cohort of similarly aged Finnish women. Our findings indicate a clear association between increasing fracture count and worsening PC, with the most pronounced decline observed among women who experienced three or more fragility fractures. While declines in SRH were more uniform across all groups, women with fractures consistently reported poorer SRH than those without fractures throughout the study period. Comorbidity burden was identified as the strongest independent predictor of deteriorating PC and SRH in the fully adjusted model. Additionally, we describe common patterns of fragility fracture accumulation, showing that an initial fracture is most frequently followed by another fracture at the same site or at the wrist.

Previous studies have reported similar associations between fragility fractures and reductions in both PC and SRH. A 10 year follow-up study of 978 postmenopausal women found that fractures were associated with significantly lower baseline functional status (Pluskiewicz et al., 2023). A multinational short term follow-up study demonstrated that fragility fractures were linked to reductions in Lawton IADL scores of 1.3–2.3 points, comparable to the deficits observed six months after a hip fracture (Yeh et al., 2023). A 20-year registry based follow-up likewise reported that fragility fractures resulted in an 84% increase in activities of daily living (ADL) limitations, compared with a 21% increase in matched controls (Fischer et al., 2017).

The impact of hip fractures on morbidity, mortality, physical capacity, and quality of life is well documented. A review of 42 studies found that hip fracture survivors experienced substantially poorer mobility, reduced functional independence, and diminished health related quality of life, alongside greater rates of institutionalization compared with their age matched peers. Only 40–60% regained their pre-fracture mobility and IADL function, 40–70% regained basic ADL independence, and 20–60% of previously independent individuals required assistance 1–2 years later, with institutionalization rates of 10–20% (Dyer et al., 2016; Haleem et al., 2023). Even seemingly less severe fractures like wrist fractures have been shown to be associated with functional decline in the mid- to long-term (Babatunde et al., 2021; Egund et al., 2020). Ankle fractures have similarly been shown to impair physical capability for up to two years (Anderson et al., 2008; Beckenkamp et al., 2014).

This study contributes to the literature by demonstrating that cumulative fracture burden is associated with long term deterioration in both PC and SRH using data with a 20-year follow-up period that is substantially longer than in most prior studies. Importantly, the study also provides insights into the role of subsequent fractures, showing that repeated fractures, especially when homotypic, may amplify functional decline.

While we were able to detect the fracture timings accurately and keep up the number of cumulative fractures dynamically, we did not try to account for the proximity of fracture to the questionnaire date even though the fractures occurring closer to a measurement point may have disproportionately affected the reported PC or SRH, because there are numerous other events and conditions that may have similar influence on PC or SRH at the individual level. Keeping all such events included, the potential sudden changes in PC or SRH are likely to blend into the general variation, and average population level associations reflect the situation in the real-world population.

Physical capability in this study was assessed using self-reported measures rather than direct performance-based tests or validated patient reported outcome measures (PROMs). However, prior research has shown strong correlations between the OSTPRE PC question and objective performance measures, including grip strength, knee extension force, ability to squat, and one leg stance time, in a pseudorandom subsample of 3,686 participants (Juopperi et al., 2021). SRH and PC were categorized into three tiers (good/intermediate/poor). For clarity and due to sample size considerations, we presented only the proportion of women reporting good PC and SRH in the main text; full distributions are available in the supplementary materials. It is also notable that the baseline differences in PC actually suggest that the lower functional status may predispose individuals to fractures. On the other hand, higher baseline physical activity levels may increase the risk of certain fractures, particularly wrist fractures, because more active and capable people are likely to have larger exposure to (slippery outdoor) conditions where fractures typically happen. This kind of bidirectional relationship between PC and fractures may complicate the interpretations.

As with all questionnaire-based studies, the current study is susceptible to reporting bias, inaccuracies, and attrition. We tried to mitigate these problems by focusing on the endpoint questionnaire so that our population is a representative cross-sectional sample at that point without dropouts, but we have a long backwards follow-up data, including PC and SRH, for that population. Although this population is a representative sample of the total population of women with similar age in the region for the endpoint, it is a selective subset of the OSTPRE population at baseline. Unsurprisingly, mortality and institutionalization account for most of the selection of later responders compared to earlier. Despite that, both the control and fracture groups showed chronic disease prevalences similar to those of the overall OSTPRE cohort. In any case, our study population naturally consists of a somewhat healthier subset of the original cohort. This should be considered while making interpretations or generalizations of the results. Moreover, this study is limited to Finnish aging women and results may not automatically generalize to men, to contemporary populations or other ethnic and healthcare contexts.

5. Conclusions

We investigated the associations between cumulative fracture burden and long-term physical capability and self-rated health over a 20-year follow-up. Fracture accumulation was associated with deterioration in both PC and SRH, with stronger and more consistent effects observed for PC. Adjusted analyses also identified comorbidity burden as an important independent predictor of diminishing PC and SRH. These findings suggest that both fracture prevention and comorbidity management are important targets for maintaining physical capability and self-rated health in aging women.

Additional Files

The additional files for this article can be found as follows:

Data Accessibility Statement

The datasets generated and/or analyzed during the current study are available from the senior authors upon reasonable request.

Statement on the Use of Artificial Intelligence (AI)

During the preparation of this work the author(s) used Microsoft Copilot in order to improve the readability of the manuscript. No part of the text in the manuscript has been originally generated with AI. After using this tool/service, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the published article.