Osteoporos Int https://doi.org/10.1007/s00198-017-4294-7
ORIGINAL ARTICLE
A meta-analysis of the association between body mass index and risk of vertebral fracture A. D. Kaze 1 & H. N. Rosen 2 & J. M. Paik 1,3,4
Received: 27 August 2017 / Accepted: 27 October 2017 # International Osteoporosis Foundation and National Osteoporosis Foundation 2017
Abstract Summary We conducted a meta-analysis of prospective studies to assess the association between BMI and incident vertebral fracture. We found that as body mass index (BMI) increases, the risk of vertebral fracture decreases in men, but not in women, suggesting possible gender differences in the relationship of BMI with risk of vertebral fracture. Introduction Recent evidence suggests that the relationship between BMI and fracture risk may be site-specific. We conducted a systematic review and meta-analysis of prospective studies to investigate the association between BMI and risk of incident vertebral fracture.
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00198-017-4294-7) contains supplementary material, which is available to authorized users. * A. D. Kaze
[email protected] H. N. Rosen
[email protected] J. M. Paik
[email protected] 1
Division of Renal Medicine, Brigham and Women’s Hospital, Boston, MA, USA
2
Division of Endocrinology, Diabetes, and Metabolism, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
3
Channing Division of Network Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA
4
Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, MA, USA
Methods PubMed and Embase were searched for relevant articles published from inception through February 15, 2017. Extracted relative risks (RR) from the prospective studies were pooled using random-effects meta-analysis. Results Six studies were included, with a total of 105,129 participants followed for 3 to 19 years. The pooled RR (95% confidence interval [CI]) for vertebral fracture per each standard deviation increase in BMI was 0.94 (95% CI = 0.80–1.10) with significant heterogeneity (I2 = 88.0%, p < 0.001). In subgroup analysis by gender, we found a significant inverse association between BMI and risk of vertebral fracture in men (RR = 0.85, 95% CI = 0.73–0.98, n = 25,617 participants) but not in women (RR = 0.98, 95% CI = 0.81–1.20, n = 79,512 participants). Across studies of women not adjusting for bone mineral density (BMD), there was no significant association between BMI and risk of vertebral fracture (RR = 0.91, 95% CI = 0.80–1.04, p = 0.18, n = 72,755 participants). However, BMI was associated with an increased risk of vertebral fracture in studies of women that adjusted for BMD (RR = 1.28, 95% CI = 1.17–1.40, p < 0.001, n = 6757 participants). Substantial heterogeneity was found among studies of women (I2 = 90.1%, p < 0.001), which was partly explained by the adjustment for BMD (adjusted R2 = 61%). We found no evidence of publication bias (p = 0.40). Conclusions In conclusion, our findings suggest that there might be gender differences in the relationship of BMI with risk of vertebral fracture. Further research is needed, including the assessment of other measures of adiposity, such as visceral adiposity, on the risk of vertebral fracture. Keywords BMI . Meta-analysis . Prospective cohort studies . Risk . Systematic review . Vertebral fracture
Osteoporos Int
Introduction
Search strategy
The vertebra is the most common fracture site in adults, making vertebral fractures a major public health problem with 700,000 new cases each year, accounting for half of all osteoporotic fractures in the USA [1]. Vertebral fractures are associated with significant disability [2], poorer quality of life [3], high cost [4], increased morbidity [5], and mortality [6] and are predictive of future fracture risk [7]. Twenty-five percent of all postmenopausal women in the USA are estimated to have a vertebral fracture [8], and the prevalence increases with age [9]. However, we currently have a limited understanding of modifiable risk factors for vertebral fracture. In fact, most vertebral fractures are not due to trauma but precipitated by routine everyday activities [10]. Risk factors for vertebral fractures may also be inherently different than for fractures at other sites, such as hip or wrist fractures, due to different biomechanics [11], microarchitecture [12], and compressive loading [13]. Obesity has traditionally been thought to be protective for fractures due to the higher bone mineral density (BMD) associated with obesity and the protective effect of soft tissue padding during falls [14]. However, emerging thinking is challenging the notion that obesity is protective for all fractures [14–16], especially given our evolving understanding of the complex pathophysiology underlying the relationship between obesity and fractures. The relation between obesity and fractures appears to be fracture site-specific [14–17]. In a recent metaanalysis of 398,610 women followed for an average of 5.7 years, after adjusting for BMD, high body mass index (BMI) was protective for hip fractures and was a risk factor for osteoporotic fractures, fractures of the tibia and fibula, distal forearm, and upper arm [18]. However, this study did not investigate the relationship between BMI and risk of vertebral fracture. Although there have been individual studies reporting on the association between obesity and the risk of vertebral fracture, the findings from these studies have been conflicting; the Global Longitudinal Study of Osteoporosis in Women (GLOW) study reported an inverse association between BMI and incident clinical vertebral fracture [19], whereas a positive association with prevalent morphometric vertebral fracture was reported in two cross-sectional studies [20, 21], and no association was found in a large population-based study in Spain [22]. Therefore, our goal was to perform a systematic review and meta-analysis of prospective studies to summarize the available evidence on the association between BMI and incident vertebral fracture.
We conducted a comprehensive search of PubMed and Embase from inception through February 15, 2017 using a combination of terms related to obesity, BMI, and vertebral fracture, without language restriction. Two reviewers (ADK and JMP) independently identified articles and sequentially screened them for inclusion, starting with titles and abstracts, then full-text review. Additionally, reference lists of identified studies were manually scanned and citing references screened through the ISI Web of Knowledge database, for possible additional eligible studies.
Materials and methods
Statistical analysis
The present systematic review and meta-analysis was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [23].
The main exposure of interest was BMI evaluated as an increase of 1 standard deviation (SD) at baseline. The outcome was the risk of incident vertebral fracture. We included
Eligibility criteria and study selection Studies were considered eligible if they met the following criteria: (1) the study design was a prospective cohort study, (2) BMI was measured at baseline in kilograms per square meter, (3) the outcome of interest was the occurrence of any vertebral fracture, and (4) the relative risk (RR) or hazard ratio (HR) and its corresponding 95% confidence interval (CI) (or data to calculate them) were reported. We excluded studies conducted in people with vertebral deformities which are considered to be a risk factor for vertebral fracture, as well as cross-sectional and retrospective studies of vertebral fractures. Data extraction Two investigators (ADK and JMP) independently abstracted data from eligible studies and conducted quality assessment. The information extracted included the first author’s name, publication year, country of study origin, number of participants, mean age, age range, sex of the participants, method used to assess BMI, study duration, the maximally adjusted HR or RR, and its 95% confidence limits, along with the variables included in the maximally adjusted model, where available. Assessment of study quality We assessed the quality of studies using the Newcastle-Ottawa Scale (NOS) for cohort studies. The NOS for cohort studies is formulated by assigning a maximum of nine stars to studies of the highest quality according to three parameters: selection of study groups, comparability of groups, and ascertainment of the outcome of interest [24].
Osteoporos Int
Results Review process The study selection process is summarized in Fig. 1. The searches identified 323 citations. After screening abstracts and removing duplicates, 46 records were selected for fulltext review. Of these, six studies were included in the final meta-analysis. Two studies reported data separately by gender. These were treated as distinct studies, leading to eight data contributions to our main analyses. Characteristics of studies The characteristics of the included studies are summarized in Table 1. Data from 105,129 participants were included in the meta-analysis, with a median age of 63.1 years (25–75th percentiles = 55.1–68.5 years). The studies were published between 2003 and 2014 and conducted in the following countries: Japan [33], Canada [34], USA [35], and Sweden [36] (one study each). One study, the GLOW, was conducted in ten countries (Australia, Belgium, Canada, France, Germany, Italy, Netherlands, Spain, UK, and USA) [19], and the European Prospective Osteoporosis Study (EPOS) was conducted in 36 centers across Europe [27]. The median length of follow-up in the studies was 3.8 years (interquartile range = 3.35–10.85 years). The sample size of the included cohorts ranged from 1614 to 52,629 participants. Two studies presented results separately for men and women [27, 36], and
323 records identified through database searches 204 from Medline 108 from Embase 11 from references and other sources
Screening
305 records screened after duplicates removed
259 records excluded
Eligibility
46 Full-text articles assessed for eligibility
Included
Fig. 1 Selection of articles for inclusion in the meta-analysis
Identification
studies that reported vertebral fractures defined by selfreport, clinical, or radiographic assessment. We used the maximally adjusted RR from each study, estimated the pooled RR, and associated 95% CIs using randomeffects model meta-analysis. The random-effects model is the most conservative approach because it makes allowances for within and between-study heterogeneity. The z statistic was used to test the null hypothesis (that BMI is not associated with the outcome). We also conducted stratified analyses by gender. We assessed the heterogeneity between studies using Cochran’s Q statistic, H, and the I2 statistics [25]. The I2 statistic estimates the percentage of total variation across studies due to true between-study difference rather than chance, assuming that I2 values of 25, 50, and 75% represent low, medium, and high heterogeneity, respectively. Whenever significant heterogeneity was found, we performed subgroup and meta-regression analyses examining the following prespecified variables: mean age, sample size, average follow-up period, year of publication, and adjustment for BMD. We evaluated the robustness of our results by performing influence analysis where each individual study was omitted one at a time and the effect on the pooled estimate was assessed. Publication bias was assessed by visual inspection of funnel plots and formal testing with the Egger’s test and Begg’s test [26]. All tests were two-sided and statistical significance was defined as p < 0.05. All analyses were performed using Stata software (StataCorp V.14, TX, USA).
40 full-text articles excluded 26 no data on the exposure 7 not prospective 4 no data on outcome 2 Different categories of BMI 1 No hazard ratio 6 articles providing 8 data contributions included in metaanalysis
Malmo Project, 10,902 Women 48 Sweden
SOF, USA
Holmberg et al. 2006 [29]
Nevitt et al. 2005 [30] Papaioannou et al. 2005 [31] Men
3173
3402
EPOS, Europe
EPOS, Europe
3
3.7
Women 62.2 (50–79) 3.8
63.1 (50–79) 3.8
Women 68
Women 71 (≥ 65)
CaMos, Canada 5143
5822
15
19
6.7
3
Assessment of vertebral fractures
Physical Radiographic examination Structured Self-reported, questionnaire medical report verification when available Physical Radiographic examination Physical Radiographic examination
Physical Clinical and examination radiographic
Physical Clinical and examination radiographic Physical Clinical and examination radiographic
Structured Self-reported questionnaire
Average follow- BMI up (years) assessment
144
80
34
181
160
168
335
442
No. of events Variables included in maximally adjusted model
0.83 Age, prior spine fracture, prior fracture (0.76–0.92) of another bone, unintentional weight loss of > 5 kg, asthma, general health, and physical activity 1.28 Age, DM, prior fracture, BMD, (1.17–1.40) back pain, and treatment 0.89 Age, resting pulse, diabetes, (0.77–1.03) serum TG, serum cholesterol, serum creatinine, GGT, smoking, poor self-rated health 0.89 Age, diastolic BP, resting pulse, diabetes, (0.76–1.04) serum TG, GGT, ESR, smoking, poor self-rated health 1.13 Age, clinic (0.97–1.35) 0.99 Age, current height, height loss, (0.32–3.09) current weight, weight loss, BMI, BMD, prevalent vertebral fracture status 0.76 Age, center (0.60–0.97) 0.85 Age, center (0.71–1.02)
Effect estimate (95% CI)
9
9
8
9
9
9
9
8
NOS quality score
BMD bone mineral density, BMI body mass index, BP blood pressure, CaMos Canadian Multicentre Osteoporosis Study, DM diabetes mellitus, EPOS European Prospective Osteoporosis Study, ESR erythrocyte sedimentation rate, GGT gamma-glutamyl transferase, GLOW Global Longitudinal Study of Osteoporosis in Women, NOS Newcastle-Ottawa Scale, SOF Study of Osteoporotic Fractures, TG triglycerides
Roy et al. 2003 [32] Roy et al. 2003 [32]
Nagano Cohort, 1614 Women 63 (≥ 50) Japan Malmo Project, 22,444 Men 44 Sweden
Tanaka et al. 2013 [28] Holmberg et al. 2006 [29]
52,629 Women 69 (≥ 55)
GLOW, 10 countries
Compston et al. 2014 [27]
Total Gender Mean age n (range), years
Study name, country
Characteristics of the eight data contributions included in the meta-analysis
Author, year
Table 1
Osteoporos Int
Osteoporos Int
four studies included only female participants [19, 33–35]. The degree of covariate adjustment varied across studies, with all studies adjusting for age. Other factors that were adjusted for included history of prior fracture, BMD, height, weight, weight loss, height loss, smoking, and poor self-rated health. A detailed list of all the variables that were adjusted for in the studies is described in Table 1.
p = 0.02, n = 25,617 participants, Fig. 2). However, there was no significant association observed between BMI and risk of vertebral fracture in women (RR = 0.98, 95% CI = 0.81–1.20, p = 0.87, n = 79,512 participants, Fig. 2). There was no evidence of statistical heterogeneity in the male subgroup (I2 = 17.7%, p = 0.27), but substantial heterogeneity was found among studies of female participants (I2 = 90.1%, p < 0.001).
Potential sources of bias and methodological quality All the studies had a NOS score of 8 or higher, indicating that the included studies were of high quality (Supplementary Table 1). BMI and incident vertebral fracture The main meta-analysis results are shown in Fig. 2. The pooled RR of vertebral fracture per 1-SD increase in BMI across the eight data contributions (n = 105,129 participants) was 0.94 (95% CI = 0.80–1.10, p = 0.45, Fig. 2). We also performed a subgroup meta-analysis by gender (Fig. 2). We observed an inverse association between BMI and risk of vertebral fracture in men (RR = 0.85, 95% CI = 0.73–0.98,
Fig. 2 Forest plot showing the association between BMI and vertebral fracture risk overall and by gender. Relative risks are per 1-SD increase in BMI. Black boxes represent the effect estimates and the horizontal bars are for the 95% confidence intervals (CIs). The diamond is for the pooled
Subgroup analyses across studies of female participants Table 2 shows the subgroup analyses to investigate the sources of heterogeneity across the prospective studies among women. The relative risk of vertebral fracture did not vary by mean age, sample size, average duration of follow-up, and year of publication (Table 2); substantial heterogeneity remained evident in subgroup analyses with all I2 > 80% and p values < 0.05, except for publication year (Table 2). In subgroup analysis defined by adjustment for BMD, no association between BMI and vertebral fracture was observed in studies that did not adjust for BMD (RR = 0.91, 95% CI = 0.80–1.04, p = 0.18, n = 72,755 participants), whereas a positive association between BMI and vertebral fracture was
effect estimate and 95% CI and the dotted vertical line centered on the diamond has been added to assist visual interpretation. BMI, body mass index; CI, confidence interval; RR, relative risk; SD, standard deviation
0.99 0.70 0.76
Sources of heterogeneity across the studies of female participants were further investigated using randomeffect meta-regression analysis. Of the study-level characteristics, the inclusion of BMD in the model explained 61% of the between-study heterogeneity. To assess the influence of individual studies on the summary estimate among women, we performed an influence analysis, in which the pooled RR was computed omitting one study at a time. No study was found to have a strong effect on the pooled estimate. We found no major asymmetrical appearance in the funnel plot (Supplementary Fig. 1) and no evidence of publication bias from the Egger’s test (p = 0.40) or Begg’s test (p = 0.90) (Fig. 3).
Bold values represent significant results, at p < 0.05
BMD bone mineral density, BMI body mass index, CI confidence interval, N number, RR relative risk, SD standard deviation
1.6 (1.0–3.0) 4.8 (3.2–7.1) 61.4 (0–89) 95.6 (90–98) 0.89 0.91 0.98 (0.77–1.25) 0.98 (0.73–1.33) 14,367 65,145 3 3 Before median (2006) After median (2006) Year of publication
Mean follow-up
Meta-regression, sensitivity analysis, and publication bias
0.08 < 0.01
0.86
0.56
0.17 0.36 0.20 0.07 2.8 (1.7–4.7) 2.2 (1.3–4.0) 2.2 (1.3–4.0) 3.6 (2.3–5.7) Below median (5483) Above median (5483) Below median (3.8) Above median (3.8)
Mean age at baseline
Sample size
3 3 3 3
10,159 69,353 63,594 15,918
1.05 (0.72–1.51) 0.93 (0.78–1.12) 0.96 (0.73–1.27) 1.00 (0.75–1.33)
0.81 0.45 0.78 0.98
87.4 (64–96) 80.1 (37–94) 80.1 (37–94) 92.2 (80–97)
< 0.01 < 0.01 < 0.01 < 0.01
0.86
1.9 (1.1–3.1) – 3.6 (2.3–5.7) 2.2 (1.3–4.0) 70.9 (17–90) 0.0 92.2 (80–97) 80.1 (37–94) No Yes Below median (65.5) Above median (65.5) Adjustment for BMD
4 2 3 3
72,755 6757 15,918 63,594
0.91 (0.80–1.04) 1.28 (1.17–1.40) 1.00 (0.75–1.33) 0.96 (0.73–1.27)
0.18 <0.01 0.98 0.78
H (95% CI) Subgroup
N studies
Sample size
RR (95% CI)
p effect estimate
I2 (95% CI)
0.02 0.66 < 0.01 < 0.01
0.32 – 0.07 0.20
0.08
found in studies that adjusted for BMD (RR = 1.28, 95% CI = 1.17–1.40, p < 0.001, n = 6757 participants, Table 2).
Group
Table 2
Subgroup meta-analyses of the association between BMI and risk of vertebral fracture across studies of female participants
p heterogeneity
p Egger test
p diff subgroups
Osteoporos Int
Discussion The present meta-analysis of six prospective cohort studies including 105,129 individuals showed that each SD increase in BMI is associated with a 15% reduction in risk of vertebral fracture among men (n = 25,617); however, no significant association was found between BMI and risk of vertebral fracture among women (n = 79,512). Additionally, in studies of female participants that did not adjust for BMD, the association between BMI and risk of vertebral fracture remained null, whereas a positive relationship was observed in studies of women that adjusted for BMD. To the best of our knowledge, our meta-analysis is the first to evaluate the relationship of BMI with risk of vertebral fracture. Our analysis revealed an inverse association between BMI and risk for vertebral fracture in men. These results are consistent with a previous meta-analysis which found that metabolic syndrome was associated with a reduction in risk of bone fractures in men but not in women [37]. However, it is difficult to make any definitive conclusion from our meta-analysis about the association between BMI and risk of vertebral fracture in men since only two studies provided data for male participants and these studies did not adjust for important factors such as BMD, mobility limitations, and walking pace. In a large prospective study of osteoporotic fractures in men, the risk of osteoporotic fractures was increased in obese subjects, but the association was weaker and nonsignificant after adjusting for mobility limitations and walking pace [38]. In contrast to the inverse association in men, we found no significant association between BMI and risk of vertebral fracture across the six studies of women. One possible explanation could be the modifying effect of age on the relationship between body composition and BMI in women. Additionally,
Osteoporos Int
Fig. 3 Forest plot showing the association between BMI and risk of vertebral fracture in women, stratified by the inclusion of BMD in the model. Relative risks are per 1-SD increase in BMI. Black boxes represent the effect estimates and the horizontal bars are for the 95% confidence intervals (CIs). The diamond is for the pooled effect estimate and
95% CI and the dotted vertical line centered on the diamond has been added to assist visual interpretation. BMD, bone mineral density; BMI, body mass index; CI, confidence interval; RR, relative risk; SD, standard deviation
BMI as a measure of adiposity has been shown to be less valid in older people due to changes in body composition associated with aging [39]. In most of the included studies, the mean age of the female participants was greater than 50 years, except in one study with a mean age of 48 years [36]. Another explanation for our null finding could be that the occurrence of osteoporosis in postmenopausal women perfectly balances the potential benefits of obesity on bone [40]. Indeed, obesity has been shown to have two discordant effects on bone in postmenopausal women [40]. First, aromatization of androgens in the adipose tissue tends to increase estrogen level and thus protects bone. On the other hand, the secretion of inflammatory cytokines by visceral adipocytes enhances bone resorption. Consequently, these two effects may nullify the association between obesity and risk of fracture in postmenopausal women [40]. There is a need to consider alternative measures of adiposity, such as body fat distribution or visceral adiposity and their effect on risk of vertebral fracture. Interestingly, across studies reporting a BMD-adjusted estimate, we found that higher BMI was associated with an increased risk of incident vertebral fracture in women. It is worth mentioning that studies not adjusting for BMD had much larger sample sizes than studies that adjusted for BMD. Our results are consistent with findings from a recent meta-analysis, which showed that upon adjustment for BMD, high BMI was a significant risk factor for osteoporotic
fractures in women [18]. However, the earlier meta-analysis did not investigate the specific relationship between BMI and risk of vertebral fracture. The mechanisms whereby BMI affects vertebral fracture are not entirely clear and could be related to the complex interplay between fat mass, bone metabolism, and fracture risk [41]. Importantly, bone and fat cells are derived from the same progenitor cells and adipose tissue is now recognized as a metabolically active endocrine organ which can exert complex effects on bone mass, strength, and quality. First, adipose tissue can exert hormonal effects on bone through the increased production of adipokines which can promote bone resorption, inhibit bone formation [42], and result in an increased risk of incident fracture [43]. Second, obesity can exert its effects on bone through the vitamin D/ parathyroid hormone (PTH) axis. Individuals with a higher BMI have higher PTH levels compared to individuals with a lower BMI [44]. Third, the relationship between BMI and risk of vertebral fracture may be related to impaired biomechanical factors [28]. Other potential pathways could be related to sarcopenia, osteoarthritis, and the associated increased risk of falls [16]. Key areas of research related to obesity and risk of vertebral fracture include the assessment of the impact of fat distribution and risk of vertebral fracture. The site of fat tissue may exert differing effects on bone (e.g., visceral fat versus subcutaneous
Osteoporos Int
fat) [31]. Several studies have found that higher abdominal fat is associated with lower BMD [31]. In a prospective study of 1126 Australian participants aged 50 or older with an average follow-up of 5 years, increased abdominal fat mass was associated with an increased risk of clinical vertebral fracture [30]. Additionally, visceral abdominal fat could also have detrimental effects on bone Bquality.^ A recent study of healthy premenopausal women found that those in the highest tertile of trunk fat had inferior bone quality, as evidenced by lower trabecular bone volume fraction, fewer and thinner trabeculae, lower trabecular stiffness, and higher cortical porosity, as well as decreased bone formation [29]. Moreover, it is unclear how the association between adiposity and vertebral fracture risk is modified in subgroups that are already vulnerable to developing fractures and have an increased fracture risk, such as people with diabetes [32] and chronic kidney disease [45]. Our meta-analysis has several limitations. First, we included prospective observational studies; hence, it is possible that the summary estimates were subject to residual confounding or other potential biases. Second, some heterogeneity was observed across the included studies, which was only partly explained in subgroup and metaregression analyses. Part of the heterogeneity is probably related to variability in the definition of vertebral fractures, as well as the demographic diversity of the included cohorts, or the variability of BMI levels within each study. Additionally, it is possible that the underestimation of vertebral fractures in some studies might have biased the results toward the null. Third, the primary studies lacked data on important covariates, as well as consistent adjustment across studies. Furthermore, we were unable to investigate a detailed dose-response relationship between BMI and vertebral fracture because of the limited data provided by the primary studies. These limitations notwithstanding, our review has several strengths. First, only prospective studies in which BMI was assessed before the development of the outcome were included. Second, we made considerable efforts to identify all the available evidence by searching multiple electronic databases. Third, we appraised the quality of studies with a standard quality assessment tool for cohort studies. Finally, our meta-analysis is quantitative and highlights the magnitude of the association between BMI and risk of vertebral fracture. In summary, our meta-analysis demonstrated that higher BMI is associated with a significant reduction in risk of vertebral fracture in men (although studies did not adjust for BMD) but not in women. However, after adjustment for BMD, higher BMI is associated with a significant increase in risk of vertebral fracture in women. Future research is warranted to confirm our findings and explore the impact of body fat distribution on risk of vertebral fracture.
Acknowledgements This research was supported by the National Institute of Health grant K23DK100447. Compliance with ethical standards Conflicts of interest None.
References 1.
2.
3. 4.
5. 6. 7.
8. 9.
10. 11.
12.
13.
14. 15.
16. 17.
18.
19.
Cauley JA, Hochberg MC, Lui LY, Palermo L, Ensrud KE, Hillier TA, Nevitt MC, Cummings SR (2007) Long-term risk of incident vertebral fractures. JAMA 298:2761–2767 Nevitt MC, Ettinger B, Black DM, Stone K, Jamal SA, Ensrud K, Segal M, Genant HK, Cummings SR (1998) The association of radiographically detected vertebral fractures with back pain and function: a prospective study. Ann Intern Med 128:793–800 Gold DT (1996) The clinical impact of vertebral fractures: quality of life in women with osteoporosis. Bone 18:185S–189S Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A (2007) Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Miner Res Off J Am Soc Bone Miner Res 22:465–475 Lr M (2003) Adverse outcomes of osteoporotic fractures in the general population. J Bone Miner Res 18:1139–1141 Lr M, Kallmes D (2006) Epidemiology of vertebral fractures: implications for vertebral augmentation. Acad Radiol 13:538–545 Delmas P, Genant H, Crans G, Stock J, Wong M, Siris E, Adachi J (2003) Severity of prevalent vertebral fractures and the risk of subsequent vertebral and nonvertebral fractures: results from the MORE trial. Bone 33:522–532 Melton LJ 3rd (1997) Epidemiology of spinal osteoporosis. Spine (Phila Pa 1976) 22:2S–11S Melton LJ 3rd, Kan SH, Frye MA, Wahner HW, O'Fallon WM, Riggs BL (1989) Epidemiol Vertebral Fract Women. Am J Epidemiol 129:1000–1011 Cummings SR, Melton LJ (2002) Epidemiology and outcomes of osteoporotic fractures. Lancet 359:1761–1767 Christiansen BA, Bouxsein ML (2010) Biomechanics of vertebral fractures and the vertebral fracture cascade. Curr Osteoporos Rep 8: 198–204 Yeni YN, Zinno MJ, Yerramshetty JS, Zauel R, Fyhrie DP (2011) Variability of trabecular microstructure is age-, gender-, race- and anatomic site-dependent and affects stiffness and stress distribution properties of human vertebral cancellous bone. Bone 49:886–894 Duan Y, Seeman E, Turner CH (2001) The biomechanical basis of vertebral body fragility in men and women. J Bone Miner Res Off J Am Soc Bone Miner Res 16:2276–2283 Compston J (2015) Obesity and fractures in postmenopausal women. Curr Opin Rheumatol 27:414–419 Compston JE, Watts NB, Chapurlat R et al (2011) Obesity is not protective against fracture in postmenopausal women: GLOW. Am J Med 124:1043–1050 Cauley JA (2017) Osteoporosis: fracture epidemiology update 2016. Curr Opin Rheumatol 29:150–156 Lacombe J, Cairns BJ, Green J, Reeves GK, Beral V, Armstrong ME (2016) The effects of age, adiposity, and physical activity on the risk of seven site-specific fractures in postmenopausal women. J Bone Miner Res Off J Am Soc Bone Miner Res 31:1559–1568 Johansson H, Kanis JA, Oden A et al (2014) A meta-analysis of the association of fracture risk and body mass index in women. J Bone Miner Res Off J Am Soc Bone Miner Res 29:223–233 Compston JE, Flahive J, Hosmer DW et al (2014) Relationship of weight, height, and body mass index with fracture risk at different
Osteoporos Int
20.
21.
22.
23.
24.
25. 26.
27.
28.
29.
30.
31.
sites in postmenopausal women: the Global Longitudinal study of Osteoporosis in Women (GLOW). J Bone Miner Res Off J Am Soc Bone Miner Res 29:487–493 Pirro M, Fabbriciani G, Leli C, Callarelli L, Manfredelli MR, Fioroni C, Mannarino MR, Scarponi AM, Mannarino E (2010) High weight or body mass index increase the risk of vertebral fractures in postmenopausal osteoporotic women. J Bone Miner Metab 28:88–93 Laslett LL, Just Nee Foley SJ, Quinn SJ, Winzenberg TM, Jones G (2012) Excess body fat is associated with higher risk of vertebral deformities in older women but not in men: a cross-sectional study. Osteoporos Int J established Result Coop Eur Found Osteoporos Natl Osteoporos Found U S A 23:67–74 Prieto-Alhambra D, Premaor MO, Fina Aviles F, Hermosilla E, Martinez-Laguna D, Carbonell-Abella C, Nogues X, Compston JE, Diez-Perez A (2012) The association between fracture and obesity is site-dependent: a population-based study in postmenopausal women. J Bone Miner Res Off J Am Soc Bone Miner Res 27:294– 300 Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 6:e1000097 Zeng X, Zhang Y, Kwong JS, Zhang C, Li S, Sun F, Niu Y, Du L (2015) The methodological quality assessment tools for preclinical and clinical studies, systematic review and meta-analysis, and clinical practice guideline: a systematic review. J Evid-Based Med 8:2– 10 Higgins JP, Thompson SG (2002) Quantifying heterogeneity in a meta-analysis. Stat Med 21:1539–1558 Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ (Clin Res Ed) 315:629–634 Roy DK, O'Neill TW, Finn JD et al (2003) Determinants of incident vertebral fracture in men and women: results from the European Prospective Osteoporosis Study (EPOS). Osteoporos Int 14:19–26 Bredella MA, Lin E, Gerweck AV, Landa MG, Thomas BJ, Torriani M, Bouxsein ML, Miller KK (2012) Determinants of bone microarchitecture and mechanical properties in obese men. J Clin Endocrinol Metab 97:4115–4122 Cohen A, Dempster DW, Recker RR et al (2013) Abdominal fat is associated with lower bone formation and inferior bone quality in healthy premenopausal women: a transiliac bone biopsy study. J Clin Endocrinol Metab 98:2562–2572 Yang S, Nguyen ND, Center JR, Eisman JA, Nguyen TV (2013) Association between abdominal obesity and fracture risk: a prospective study. J Clin Endocrinol Metab 98:2478–2483 Bredella MA, Torriani M, Ghomi RH, Thomas BJ, Brick DJ, Gerweck AV, Harrington LM, Breggia A, Rosen CJ, Miller KK (2011) Determinants of bone mineral density in obese premenopausal women. Bone 48:748–754
32.
Bonds DE, Larson JC, Schwartz AV, Strotmeyer ES, Robbins J, Rodriguez BL, Johnson KC, Margolis KL (2006) Risk of fracture in women with type 2 diabetes: the Women's Health Initiative Observational Study. J Clin Endocrinol Metab 91:3404–3410 33. Tanaka S, Kuroda T, Saito M, Shiraki M (2013) Overweight/obesity and underweight are both risk factors for osteoporotic fractures at different sites in Japanese postmenopausal women. Osteoporos Int 24:69–76 34. Papaioannou A, Joseph L, Ioannidis G et al (2005) Risk factors associated with incident clinical vertebral and nonvertebral fractures in postmenopausal women: the Canadian Multicentre Osteoporosis Study (CaMos). Osteoporos Int J Established Result Coop Eur Found Osteoporos Natl Osteoporos Found U S A 16: 568–578 35. Nevitt MC, Cummings SR, Stone KL et al (2005) Risk factors for a first-incident radiographic vertebral fracture in women > or =65 years of age: the study of osteoporotic fractures. J Bone Miner Res Off J Am Soc Bone Miner Res 20:131–140 36. Holmberg AH, Johnell O, Nilsson PM, Nilsson J, Berglund G, Akesson K (2006) Risk factors for fragility fracture in middle age. A prospective population-based study of 33,000 men and women. Osteoporos Int 17:1065–1077 37. Yang L, Lv X, Wei D, Yue F, Guo J, Zhang T (2016) Metabolic syndrome and the risk of bone fractures: a meta-analysis of prospective cohort studies. Bone 84:52–56 38. Nielson CM, Marshall LM, Adams AL, LeBlanc ES, Cawthon PM, Ensrud K, Stefanick ML, Barrett-Connor E, Orwoll ES (2011) BMI and fracture risk in older men: the osteoporotic fractures in men study (MrOS). J Bone Miner Res Off J Am Soc Bone Miner Res 26:496–502 39. Villareal DT, Apovian CM, Kushner RF, Klein S (2005) Obesity in older adults: technical review and position statement of the American Society for Nutrition and NAASO, The Obesity Society. Obes Res 13:1849–1863 40. Reid IR (2010) Fat and bone. Arch Biochem Biophys 503:20–27 41. Cao JJ (2011) Effects of obesity on bone metabolism. J Orthop Surg Res 6:30 42. Aguirre L, Napoli N, Waters D, Qualls C, Villareal DT, Armamento-Villareal R (2014) Increasing adiposity is associated with higher adipokine levels and lower bone mineral density in obese older adults. J Clin Endocrinol Metab 99:3290–3297 43. Johansson H, Oden A, Lerner UH et al (2012) High serum adiponectin predicts incident fractures in elderly men: osteoporotic fractures in men (MrOS) Sweden. J Bone Miner Res 27:1390–1396 44. Bolland MJ, Grey AB, Ames RW, Horne AM, Gamble GD, Reid IR (2006) Fat mass is an important predictor of parathyroid hormone levels in postmenopausal women. Bone 38:317–321 45. Nickolas TL, McMahon DJ, Shane E (2006) Relationship between moderate to severe kidney disease and hip fracture in the United States. J Am Soc Nephrol 17:3223–3232