Fragility fractures are defined as new fractures caused by low trauma (fall from a standing height or less) and mostly affected the hip, spine, distal radius, and proximal humerus (1, 2). Previous data showed that fragility fractures increase the incidence of related complications and even death. Patients with hip fractures have a mortality range of 8.4%–36.0% during the first year following the injury, and nearly 50% are disabled or require long-term home care (25%) (3, 4). High medical costs cause a heavy economic burden for individuals and society. Related risk factors further screened are essential for the prevention of fragility fractures.

In addition to age, gender, bone mineral density, and other indicators in the most widely used fracture risk assessment tool (FRAX) risk prediction system, the relationship between metabolic diseases and related risk factors and fragility fractures is of growing concern. Studies on the association between lipid metabolic disorders and fragility fractures have not shown consistent results. Remnant cholesterol (RC) is a type of cholesterol content that is rich in triglyceride lipoproteins and has been of extensive interest for it is associated with adverse cardiovascular events such as atherosclerosis (5), ischemic heart disease (6), and coronary heart disease (7). A longitudinal cohort study performed in China indicated a superior independent association of increased RC levels with new-onset carotid plaque compared with other conventional lipid parameters (8). Considering a close link between lipid metabolism and bone remodeling, we speculate that the level of RC may be associated with the risk of fragility fractures.

Most previous studies on the effect of lipids on fragility fractures were based on a single lipid level. However, the effect of risk factors on adverse outcome events depends not only on the dose of exposure but also on the duration of exposure. A study found that using multiple blood pressure recordings from patients’ electronic health records showed stronger associations with incident cardiovascular disease than a single blood pressure measurement (9). Therefore, to examine the effect of RC on fragility fractures, the current study was conducted in the Kailuan cohort (registration number: chicTR-TNRC-11001489) to analyze the association between RC and the risk of fragility fractures using cumulative RC (cumRC).

This study provides the report of a U-shaped relationship between the cumRC level and the risk of developing fragility fractures. Both too high and too low cumRC levels were associated with a greater risk of fragility fractures, with the effect being more significant in young and middle-aged people (age <65 years). To the best of our knowledge, this is the first study to investigate the association between cumRC exposure and fragility fractures, and our findings further confirm that dyslipidemia increases the risk of fragility fractures.

Following adjustment for potential confounders, compared with the Q2 group with the lowest event rate, the risk of new-onset fragility fractures increased by 61% and 38% in the Q1 and Q4 groups, respectively, and maybe the result was affected by the single RC measurement but remained independent of it. Further analysis using RCS regression showed a non-linear relationship between cumRC exposure and the risk of fragility fractures. While no similar studies have been conducted to date, the effect of traditional lipid indicators on fragility fractures has been reported. A meta-analysis showed that each 50 mg/dl increase in TC was associated with a 15% increase in fracture risk (11). In contrast, a prospective cohort study in Austria revealed that women with higher TC had a reduced risk of hip fracture (12). A nested case–control study from Denmark showed that lower LDL-C levels were associated with an increased risk of fracture, as compared with LDL-C measurements between 3.039 mmol/L and 5.959 mmol/L. LDL-C measurements equal to or higher than 3 mmol/L were protective against fragility fractures among patients with diabetes (13). Different from the previous observation of a single blood lipid level, the present study evaluated the risk of fragility fractures using the cumulative exposure value calculated from three RC measurements over 6 years. The results reflected the association between long-term RC burden and the risk of new-onset fragility fractures, and the sensitivity analyses were supportive of the main analysis.

However, when further stratified by age (<65 years and ≥65 years), the association between cumRC and fragility fractures was significant in the age group <65 years, but not in the age group ≥65 years. According to research, the primary risk factors for osteoporosis in individuals younger than the age of 55 are modifiable, while non-modifiable risk factors such as reduced hormone levels, chronic pain, and sleep disorders become crucial in individuals older than 55 (14). It is suggested that lowering exposure to risk factors may be more beneficial for the primary prevention of osteoporosis in younger populations. In older individuals, the risk of cardiovascular disease, chronic inflammation, glucose, and lipid disorders increases. Although a single disease may not significantly affect bone density, the simultaneous occurrence of multiple diseases is associated with the accumulation of physiological system imbalances and a decrease in bone strength. Moreover, the decline in muscle mass caused by a lack of physical activity in the elderly may increase the risk of falls and decrease bone density. Therefore, the effect of cumRC exposure on fragility fractures may be overlaid by aging in the elderly population. Although the incidence of fracture is lower in younger populations, the risk of secondary fractures may be increased after the first fracture. Our research results may provide insights into strategies for preventing fragility fractures in younger populations. Statins are the commonly used lipid-lowering agents, but the effects of statins on osteoblast differentiation and bone formation remain controversial (15). Although the HR value did not change appreciably in either the Q1 or Q4 groups compared with model 2 after further adjusting for lipid-lowering therapy in model 3, in a sensitivity analysis excluding people taking statins, the HR value decreased in the Q1 group, suggesting that statin lipid-lowering drugs may increase the risk of fragility fractures caused by low RC. Therefore, not only the lipid-lowering effect of statins but also the increased risk of fragility fractures should be considered when they are prescribed for those with lower levels of RC.

Previous studies showed U-shaped or reverse U-shaped relationships between some adverse health events and lipid levels (16, 17), similar to the present study between cumRC and fragility fracture risk, suggesting a possible dual effect of lipid metabolism on bone metabolism. On the one hand, it has been found that increased lipids accumulate beneath the vascular intima and perivascular space in the bones, and the inflammatory bioactive lipids induce bone loss; particularly, oxidized LDL-C plays a significant role in bone loss. Lipid oxidation products can promote arterial calcification by activating osteoblasts but inhibit bone formation in bone tissue (18). Lower Runx2 expression and higher TRAP expression were found in both diet-induced and genetic hyperlipidemia mice, indicating decreased osteoblastic functions and increased osteoclastic functions in these mice (19). Moreover, hyperlipidemia may also induce secondary hyperparathyroidism, further impairing bone regeneration and compromising mechanical strength (20). Finally, the accumulation of fat in the femoral head has been observed to elevate the pressure within the bone marrow microcirculation, resulting in a reduction in bone vascularization. This diminished blood supply causes ischemia and hypoxia in the affected area, impairing the blood supply to the bone. All these factors may contribute to an increased risk of fractures (18). On the other hand, it has been proven that exogenous cholesterol inhibits osteoblast differentiation, while endogenous cholesterol may promote osteogenic differentiation of bone marrow mesenchymal stem cells by activating the Hedgehog signaling pathway (21). Cholesterol may inhibit autophagy during osteoclast differentiation by activating the phosphatidylinositol 3-kinase/AKT/mammalian target of rapamycin signaling pathway and play a role in bone resorption and formation (22). These studies suggest that the effect of cholesterol on osteogenesis is more complex than either “bad” or “good.” Noteworthy, in the present study, the HR for fragility fractures was higher in the Q1 than in the Q4 group. It has been found that cholesterol loading promotes osteogenic differentiation of mesenchymal stem cells and mainly depends on the effect of cholesterol esters, which can promote the formation of mineralized nodules by increasing bone morphogenetic protein 2 and runt-related transcription factor 2 expression (23). The RC particle is larger and carries more cholesterol than other types of cholesterol, and the negative effect of cumRC decrease on osteogenesis may be more obvious, but the specific mechanism remains to be studied.

The main strengths of the present study are the large number of samples and the long follow-up period. Compared with previous studies on the association between single lipid exposure and the risk of fragility fractures, this study used repeated measurement data, which provides a theoretical basis for the application of cumulative exposure. However, there are some limitations. Firstly, the lack of bone mineral density data limits the further exploration of the problems and potential reasons, such as whether the association between cumRC and the risk of fragility fractures is affected by bone mineral density. Second, the large sample was recruited from the data from north China, and whether these results apply to other countries or populations remains unknown. Third, information about potential risk factors such as parental fragility fracture history was not obtained, so there may be residual confounding in the study. Fourth, the data on the level of vitamin D hormone drugs, the usage of steroids, and other bone metabolic markers were not obtained from the participants.

In conclusion, both too high and too low cumRC exposure increased the risk of fragility fractures, and the effect was more significant in young and middle-aged people independent of a single RC measurement. Therefore, much more attention should be devoted to the harm caused by cumRC exposure for both clinicians and individuals and control the RC level in the ideal range as much as possible to reduce the risk of fragility fractures and improve the quality of life.

The data that support the findings of this study are available on request from the corresponding author.

The studies involving humans were approved by the Ethics Committee of Kailuan Medical Group. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

FT and SW designed the study and edited the manuscript. XH developed the manuscript draft and statistical analyses. NZ and LG collected the data and conducted the investigation. YW, MZ, and SC conducted the investigation and methodology. PL, MW, and JL took responsibility for the integrity of the data. All authors contributed to the article and approved the submitted version.