This is one of the few representative clinical cohort studies examining the effect of schizophrenia on the risk of site-specific skeletal fractures by using a nationwide population-based data set over a 10-year follow-up period. The major finding of this study was the significantly high incidence and risk of hip and vertebral fractures among patients with schizophrenia compared with controls.
As shown in Table 1, the patients with schizophrenia had more comorbidities than the controls, which is consistent with the findings of previous reports; compared with the general population, patients with schizophrenia have increased physical problems, including diabetes, coronary heart disease, cerebrovascular disease, chronic obstructive pulmonary disease, and osteoporosis [7, 18]. As shown in Tables 2–4, after adjusting for the confounding factors, the HRs for hip and vertebral fractures were 1.78 and 1.40, respectively, and both values were significant. These results suggested that the schizophrenia cohort had significantly high risks of hip and vertebral fractures compared with the control cohort. Furthermore, Cox proportional-hazards regression analysis revealed that, in addition to schizophrenia diagnosis, other risk factors for fractures include age; sex; and the comorbidities of hypertension, coronary artery disease, and osteoarthritis. These results are compatible with those in previous reports [11, 19, 20].
Few studies have explored the correlation between schizophrenia and fractures, and they have reported inconsistent results. Bolton et al. revealed that schizophrenia is associated with fractures (odds ratio [OR]: 2.17, 95% confidence interval [CI]: 1.75–2.69) [21]. Moreover, Howard et al. reported that hip fracture is associated with both schizophrenia (OR: 1.73; 95% CI: 1.32–2.28) and the use of prolactin-raising antipsychotics (OR: 2.6; 95% CI: 2.43–2.78); however, in their multivariate analysis, schizophrenia diagnosis did not have an independent association with fracture risk [22]. Sorensen et al. conducted a national cohort study and found that, after adjusting for the established risk factors for fracture, schizophrenia diagnosis was associated with an increased risk of hip fracture, but the effect was nonsignificant after adjustment for psychotropics [23]. In a case–control study using data from the Taiwan NHIRD, Wu et al. reported a similar result that current antipsychotic use is associated with an increased risk of hip fracture in patients with schizophrenia [24]. However, Tsai et al. conducted a nationwide cohort study based on data from the same Taiwan NHIRD, which revealed that the proportion of days with psychiatric medication does not have a dose–response relationship with fracture risk [25]. Bishop et al. conducted a study in female patients and concluded that fractures in the past 12 months were more likely in the schizophrenia cohort than in the control cohort [26]; Kelly et al. reported similar results in the schizophrenia and control groups [27]. In a comparative meta-analysis of eight studies, Stubbs et al. revealed that patients with schizophrenia have a higher fracture risk compared with the control population (incidence rate ratio: 1.72; 95% CI: 1.24–2.39) [28]. The present study revealed similar findings and further analyzed the risk of fractures at different sites. The strengths of this study are discussed as follows. First, this was a nationwide representative study. Second, this was a cohort study with an average follow-up of 8 years. Third, the study used data from the NHIRD, which enabled a comparison of a range of relevant clinical variables between patients with schizophrenia and controls. Fourth, the study included three fracture sites for a more comprehensive analysis. Fifth, we considered both sexes and performed a sex-specific analysis. Therefore, the results of this nationwide population-based cohort study on the influence of fractures on patients with schizophrenia may provide useful contributions to the literature.
The possible explanations for increased fracture risk among patients with schizophrenia are multifactorial and may include (1) the influence of psychotropic medications, (2) schizophrenia-related poor lifestyle, and (3) some common schizophrenia comorbidities. First, the short-term side effects of antipsychotics, antidepressants, and benzodiazepines, which include sedation, extrapyramidal symptoms, somnolence, and orthostatic hypotension, may affect patient balance and gait, leading to an increased risk of fall and fracture [29-31]. Some studies have investigated the influence of prolactin-increasing antipsychotics on the bone mineral density of patients with schizophrenia. In theory, hyperprolactinemia is associated with decreased bone mineral density, which may increase fracture risk [32, 33]. Sustained hyperprolactinemia may accelerate bone turnover and increase bone resorption more than bone formation, contributing to an osteopenic effect mediated by the hypothalamic–pituitary–gonadal axis [34-36]. However, no study has reported a consistent association between hyperprolactinemia and reduced bone mineral density in patients with schizophrenia [37].
Second, sedentary lifestyle, staying indoors, and poor nutritional intake are frequently observed in patients with schizophrenia, and these behaviors contribute to reduced muscle strength, low sun exposure, and vitamin D deficiency, followed by reduced bone mineral density and increased fall and fracture risk [38, 39]. Studies have demonstrated that low physical activity in patients with schizophrenia is significantly correlated with negative symptoms [40]. Muscle weakness has been reported to be a potential key predictor for falls and fractures [41], and muscular fitness is impaired in patients with schizophrenia, which may contribute to their reduced walking capacity [42]. Therefore, a multidisciplinary team, including physiotherapists and exercise physiologists, may aid in reducing fracture risk in patients with schizophrenia [43].
Third, certain comorbidities of schizophrenia, including diabetes mellitus, smoking, and excessive alcohol consumption, contribute to fracture risk. Patients with schizophrenia have an increased risk of diabetes mellitus, which is the key risk factor for fractures in the general population [44]. Furthermore, smoking and alcohol use disorder are common in patients with schizophrenia, both of which are established risk factors for osteoporosis and osteoporotic fracture [45, 46].
Some studies have focused on the influence of sex on fracture risk in patients with schizophrenia, and the results have been somewhat unclear. Howard et al. [22] indicated that hip fracture risk did not increase in male or female patients, whereas Jung et al. [47] concluded that the risk of any fracture increased in patients of both sexes. Moreover, some studies have reported that female patients with schizophrenia are more likely to experience a fracture than controls [26], whereas others have reported no such difference [27]. In the present study, a sex-based subgroup analysis revealed that the risk of hip fracture remained significantly higher in female patients with schizophrenia than in the female controls. However, after inclusion of the interaction term between sex and schizophrenia in the model, no effect of the association between sex and schizophrenia on the risk of fractures was observed. Considering differences in the designs of the abovementioned studies, a more comprehensive investigation is required to determine the correlation between sex and fracture risk among patients with schizophrenia.
Several limitations of this study, which are inherent to claims database, must be considered. First, the causes of fractures are usually complex and vary depending on the individual. The potential risk factors of osteoporotic fractures include age, sex, body mass index, tobacco use , alcohol consumption, bone mineral density, physical inactivity, and depression [48-50]. Many psychological and environmental factors can also contribute to fracture risk. Several important demographic variables, such as alcohol consumption, tobacco use, lifestyle, body mass index, and physical activity, are unavailable in the NHIRD. Second, psychotropic drugs have been suggested to be an important confounder that increases the incidence of fracture [51-53]. However, we did not include data related to psychotropics in this study because of the complexity of pharmacological prescriptions, such as polypharmacy, frequent switching of psychotropics, self-medication, and the high possibility of poor drug adherence, in patients with schizophrenia. This limitation is unavoidable in a retrospective study based on NHIRD data. Further prospective studies that take into account psychotropics are required. Third, the follow-up duration in our study may have been insufficient for detecting late onset fractures. Thus, future studies with longer follow-up periods are required to more thoroughly elucidate the long-term risk of fractures in patients with schizophrenia. Fourth, the sample size of some subgroups was small, especially the subgroup of patients with schizophrenia with wrist fractures (n = 18), which may have contributed to a low statistical power in these groups. Given the scarcity of cohort studies on this topic, additional studies must be conducted to obtain evidence clarifying the association between schizophrenia and fractures.
In conclusion, this nationwide cohort study revealed increased risks of hip and vertebral fractures in patients with schizophrenia during a 10-year follow-up. Hip fracture risk was significantly higher in female patients with schizophrenia than in female controls. Considering the high morbidity and mortality in patients with schizophrenia who experienced a fracture, strategies must be developed to improve lifestyle and bone health, which may reduce the possibility of subsequent fractures. To confirm our findings, prospective studies, particularly those with additional patient-level data, are warranted.