- Email: [email protected]
Effect of Obesity on Bone Healing After Foot and Ankle Long Bone Fractures
The Journal of Foot & Ankle Surgery xxx (2016) 1–5
Contents lists available at ScienceDirect
The Journal of Foot & Ankle Surgery journal homepage: www.jfas.org
Original Research
Effect of Obesity on Bone Healing After Foot and Ankle Long Bone Fractures Jakob C. Thorud, DPM, MS, AACFAS 1, 2, Spencer Mortensen, DPM, AACFAS 3, 4, 5, Jennifer L. Thorud, MPH 6, Naohiro Shibuya, DPM, MS, FACFAS 1, 2, 7, Yolanda Munoz Maldonado, PhD 8, Daniel C. Jupiter, PhD 9 1
Podiatrist, Central Texas Veterans Affairs Health Care System, Temple, TX Podiatrist, Baylor Scott & White Health, Temple, TX Resident, Central Texas Veterans Affairs Health Care System, Temple, TX 4 Resident, Baylor Scott & White Health, Temple, TX 5 Resident, Texas A&M Health Science Center, College of Medicine, Bryan, TX 6 Research Associate, Baylor Scott & White Health, Temple, TX 7 Professor, Texas A&M Health Science Center, College of Medicine, Bryan, TX 8 Biostatistician, Baylor Scott & White Health, Temple, TX 9 Assistant Professor, Department of Preventive Medicine and Community Health, University of Texas Medical Branch, Galveston, TX 2 3
a r t i c l e i n f o
a b s t r a c t
Level of Clinical Evidence: 3
As obesity has become more common, fractures in the obese population have become more frequent. Concern exists regarding alterations in bone health and healing in obese patients. A matched case-control study was performed at 1 institution to evaluate whether an association exists between nonunion and a high body mass index in metatarsal and ankle fractures. A total of 48 patients with nonunion were identified, and control patients matched 2 to 1 (n ¼ 96) were selected. The control patients were matched for age, sex, and fracture type. No association was identified between nonunion and the continuous body mass index (p ¼ .23) or morbid obesity, with a body mass index of 40 kg/m2 (p ¼ .51). However, the results from both univariate and multivariate analysis suggested that patients with a current alcohol problem or a history of an alcohol problem might have a greater risk of nonunion. The odds ratio of a patient with a history of alcohol use experiencing nonunion was 2.7 (95% confidence interval 1.2 to 6.2). Further studies are warranted to confirm these findings. Ó 2016 by the American College of Foot and Ankle Surgeons. All rights reserved.
Keywords: alcohol BMI body mass index fracture healing metatarsal fracture
Obesity is a growing problem in the United States. In 2009 to 2010, the Centers for Disease Control and Prevention National Center for Health Statistics found that 35.5% of men were considered obese, a significant increase from the 27.5% in 1999 to 2000. These numbers were obtained using body mass index (BMI), a widely accepted method of grossly assessing obesity in large populations (1). Increased body weight has traditionally been thought to be protective against fractures owing to the increased bone density (2). However, recent studies have questioned this supposition. Cohen et al (3) found that trunk fat is associated with poor bone quality, decreased trabecular bone volume, decreased stiffness, greater cortical porosity, and decreased bone formation in premenopausal women with a normal
Financial Disclosure: None reported. Conflict of Interest: None reported. Address correspondence to: Jakob C. Thorud, DPM, MS, AACFAS, 3922 Mercy Drive, McHenry, IL 60050. E-mail address: [email protected] (J.C. Thorud).
bone mineral density. Likewise, a study across 3 age groups of women found that the relative bone strength decreased with an increasing fat mass, and this relationship strengthened with increasing age (4). Cao (5) suggested that increased adiposity could have detrimental effects on bone by increasing inflammatory cytokine levels, which have negative effects on bone metabolism. Although most studies have found increased bone mineral density with increased weight (6), inferior bone quality is more likely to negatively affect bone healing after fracture. An association with nonunion and obesity has been reported; however, these studies had small sample sizes and did not account for possible confounders (7–11). This was largely because of the difficulty in obtaining a large sample of patients with nonunion. Because, fortunately, nonunions are relatively uncommon, a large population of patients with fracture must be reviewed to find the relatively few cases of nonunion. In the present study, we first reviewed the data from all patients with ankle and metatarsal fractures to find those with a diagnosis of nonunion within the first 18 months after the
1067-2516/$ - see front matter Ó 2016 by the American College of Foot and Ankle Surgeons. All rights reserved. http://dx.doi.org/10.1053/j.jfas.2016.11.010
2
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
Table 1 Definitions used for medical record review Definition Radiographic union: meets all 3 of the following criteria Presence of 2 cortical continuations or bridging at fusion site Absence of hardware failure (breakage or lucent region surrounding hardware) Absence of lytic gapping of fracture site Delayed union: failure to meet criteria for union between 3 and 6 mo Nonunion: failure to meet criteria for union after 6 mo or failure to meet criteria with correctional surgery before 6 mo after fracture diagnosis Smoking history Active: smoked within previous 2 wk or returned to smoking during postoperative period Quit: stopped smoking >2 wk before surgery and did not restart during postoperative period Never: never smoked Unknown: smoking status not recorded Alcohol history: history or current documented concern of alcohol consumption
fracture. We then selected matched controls to obtain a large cohort of patients with nonunion. The present study compared the BMI of patients with long bone fractures of the foot and ankle that had progressed to nonunion and those that had healed uneventfully. To determine whether the BMI was significantly different between these 2 groups, a retrospective medical record review was performed to compare patients with radiographic nonunion with those with radiographic union. Patients and Methods The present study was a retrospective review of patients aged 18 to 90 years, selected using the International Classification of Diseases, 9th Revision, Clinical
Modification (ICD-9-CM), diagnosis codes for fractures of the ankle and metatarsals, presenting from January 1, 2006 to January 1, 2012 to Baylor Scott & White Health Memorial Hospital. We included both patients who had undergone operative and nonoperative treatment of fracture. Patients who died before 18 months after the fracture diagnosis were excluded. Pathologic fractures were excluded from the present study. The diagnosis of nonunion (ICD-9-CM code 733.82) within 18 months after the fracture diagnosis (ICD-9-CM codes 824 open or closed ankle fracture, 825.25 metatarsal fracture, closed, and 825.35 metatarsal fracture, open) were used to identify the possible patients with nonunion. A medical record review was then undertaken to confirm the diagnosis of nonunion using the criteria listed in Table 1. An example of a nonunion identified using the criteria is shown in the Fig. Patients without sufficient radiographic data to determine the presence of nonunion were excluded from the study (i.e., patients without enough follow-up data available). Laboratory data were recorded only if they were collected within 90 days of the fracture date. After identification and confirmation of all nonunion cases, matched control patients without a diagnosis of nonunion were selected using the same parameters and from the same patient population. The control patients were matched for age, sex, and fracture type in a 2:1 ratio to those with nonunion. Comparisons of the continuous variables were performed using the average values for the controls and a paired t test was used for the comparison between the cases and controls. McNemar’s tests were used to compare the clinical and radiologic outcomes for each of the controls against the respective case. A conditional logistic regression model was fit to the data to assess for possible significant factors for nonunion. The candidate predictors in the model were alcohol use, diabetes, tobacco history, renal disease, and BMI as a categorical variable. Forward, backward, and stepwise regression all selected alcohol use as the only significant predictor. The final model included alcohol use and BMI coded as a categorical variable as predictors. BMI was included in the final model because that was the aim of the present study. All analyses were completed using SAS, version 9.4 (SAS Institute, Cary, NC).
Results A total of 48 patients with nonunion were identified, and 96 control patients were matched. The mean age of the case patients with nonunion and control patients with union was 50.8 14.5 years
Fig. Radiographic evidence of nonunion. Ankle radiographs of nonunion in a 51-year-old female with a body mass index of 40 kg/m2 with a medial malleolar fracture originally fixated >8 months earlier. The parallel lines indicate the absence of cortical continuations. The bracket indicates gapping of the fracture fragments. The white arrow indicates the lucency surrounding screws.
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
Table 2 Demographic characteristics (N ¼ 144) Characteristic
Nonunion (n ¼ 48)
Control (n ¼ 96)
Total
p Value
Age (yr) Sex Female Male Race White Black Other BMI (kg/m2) BMI group BMI <40 kg/m2 BMI 40 kg/m2 Diabetes Peripheral neuropathy Malnutrition Vitamin D deficiency Renal disease Alcohol history Tobacco use Active Quit Never smoked
50.8 14.5
51.3 15.3
51.1 15.0
.48* .99y
30 (62.5) 18 (37.5)
58 (60.4) 38 (39.6)
88 (61.1) 56 (38.9)
36 (75.0) 4 (8.3) 8 (16.7) 32.3 7.1
70 (72.9) 10 (10.4) 16 (16.7) 29.6 7.0
106 (73.6) 14 (9.7) 24 (16.7) 31.4 7.1
40 8 15 8 2 5 8 18
84 12 21 11 2 8 8 19
124 20 36 19 4 13 16 37
.92y
(83.3) (16.7) (31.3) (16.7) (4.2) (10.4) (16.7) (37.5)
12 (26.1) 13 (28.3) 21 (45.7)
(87.5) (12.5) (21.9) (11.5) (2.1) (8.3) (8.3) (19.8)
17 (18.1) 13 (13.8) 64 (68.1)
(86.1) (13.9) (25.0) (13.2) (2.8) (9.0) (11.1) (25.7)
.23* .51y .17y .37y .49y .68y .13y .02y .06y
29 (20.7) 26 (18.6) 85 (60.7)
Abbreviation: BMI, body mass index. Data presented as n (%) for categorical variables and mean standard deviation for continuous variables. * Age and BMI continuous data were analyzed with a paired t test pairing the case against the average value of the controls. y Categorical data were analyzed using McNemar’s test.
and 51.3 15.3 years, respectively. The demographic characteristics for the cohort are listed in Table 2. The combined group of cases and controls experienced 75 ankle fractures and 69 metatarsal fractures. The effect of the BMI, measured as both categorical (<40 versus 40 kg/m2) and continuous data, was similar for the cases and matched controls. Of the 48 patients with nonunion and 96 controls with union, 8 and 12, respectively, were categorized as morbidly obese (p ¼ .23). The mean BMI of those with nonunion and those with union was 32.3 7.1 kg/m2 and 29.6 7.0 kg/m2, respectively (p ¼ .51). No initial surgical operation was needed in most cases; 57 of all 144 patients (cases and controls; 39.6%) required surgery overall. No difference was seen in the initial operations performed between the 2 groups, with 40 of the 96 controls and 17 of the 48 patients with nonunion undergoing an initial operation (p ¼ .67). No significant difference was found between the cases and controls in terms of age, sex, or race. Only a history of alcohol use was a significant factor for a patient to experience nonunion (p ¼ .02). The laboratory values for the cases and controls are listed in Table 3. No association was found between the calcium, albumin, blood urea nitrogen, or creatinine levels and nonunion. Vitamin D had too few recorded values to be analyzed, and parathyroid hormone was not analyzed owing to the small sample size. In the conditional
3
logistic regression model, no association was found for calcium, albumin, vitamin D, blood urea nitrogen, or creatinine levels and nonunion. The adjusted odds ratio estimates for alcohol history and BMI as a categorical variable are listed in Table 4. The odds ratio for patients with a history of alcohol use experiencing nonunion was 2.7 (95% confidence interval 1.2 to 6.2) and the odds ratio for patients with a BMI of 40 kg/m2 was 1.5 (95% confidence interval 0.6 to 3.9). Discussion The association of nonunion with obesity in patients has been documented previously; however, these studies were small and reported conflicting results. In a retrospective study of 344 cats with fractures, 18 developed nonunion. Cats with nonunion weighed significantly more than did the cats with union (9). Jupiter et al (7) found 16 cases of nonunion in 21 obese patients with long bone fractures of the upper and lower extremities. Decomas and Kaye (10) found that 7 of 19 patients (37%) with nonunion of the humerus were obese. However, no control group was used to compare the rates of obesity in these studies. In a retrospective study of 38 nonoperatively treated diaphyseal humeral fractures, 15 developed nonunion (11). The investigators stated that obesity was significantly associated with nonunion; however, they did not provide the number of obese patients within each group. Tucker et al (12) prospectively followed up 113 patients with femoral shaft fractures treated with intramedullary nailing. Much controversy exists regarding bone healing and vitamin D deficiencies. In a study of 35 patients with nonunion compared with 35 patients with bony union of closed fractures of the extremities, 86% of the patients had vitamin D deficiencies (defined as <30 ng/mL) in both the union and the nonunion groups. Although no significant difference was found, the study had a small sample size (13). Brinker et al (14) reviewed the data from 37 patients with nonunion of the upper and lower extremities, without obvious causes such as inadequate fixation, who had been referred to an endocrinologist from a large set of 683 consecutive patients with nonunion. Of the 37 patients, 83.8% had a newly diagnosed metabolic or endocrine abnormality, of which vitamin D deficiency was the most common, occurring in 25 patients (14). It is unknown how many of the patients without nonunion also had a vitamin D deficiency. More recently, 30 cases of unexplained tibial fracture nonunion were matched to 32 cases of tibial fractures with normal healing. These were matched by surgical treatment, age, sex, and BMI, and those with nonunion had significantly lower vitamin D levels than the levels of the controls (25.8 20.4 nmol/L versus 49.03 26.9 nmol/L) (15). In a study comparing ovariectomized and vitamin D-deficient rats (mimicking the fracture healing commonly found in postmenopausal women) against control rats, Melhus et al (16) found no difference in healing rates, mean bone mineral density of the callus, or mechanical strength
Table 3 Laboratory values stratified by group (N ¼ 144) Serum Laboratory Data
Nonunion (n ¼ 48)
Control (n ¼ 96)
Total
p Value
Calcium (mg/dL) Albumin (g/dL) 25(OH)D (ng/mL) PTH (pg/mL) BUN (mg/dL) Creatinine (mg/dL)
21; 9.1 0.6 20; 3.8 0.6 7; 39 (24 to 72) 5; 12 (3 to 73) 22; 17 (8 to 72) 24; 1.0 (0.6 to 3.3)
44; 9.0 0.5 43; 3.8 0.5 7; 33 (25 to 55) 2; 150 (107 to 193) 47;15 (6 to 84) 48; 0.9 (0.5 to 4.1)
65; 9.0 0.5 63; 3.8 0.5 14; 36 (24 to 72) 7; 38 (3 to 193) 69; 16 (6 to 84) 72; 0.9 (0.5 to 4.1)
.10* .37* NA NA .88y .87y
Abbreviations: BUN, blood urea nitrogen; 25(OH)D, 25-hydroxyvitamin D; NA, not applicable (not tested because of small sample size); PTH, parathyroid hormone. Data presented as n and mean standard deviation or n and median (range), as appropriate. * Calcium and albumin were normally distributed and analyzed using the paired t test. y BUN and creatinine were not normally distributed and the Wilcoxon sign rank test was used.
4
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
Table 4 Adjusted odds ratio estimates (N ¼ 144) Effect
Point Estimate
95% CI
Alcohol history (yes versus no) BMI (40 versus <40)
2.69 1.47
1.16 6.21 0.55 3.92
Abbreviations: BMI, body mass index; CI, confidence interval.
of bone, except for stiffness of the right femoral neck. In contrast, in a randomized, double-blind, placebo-controlled study of 30 osteoporotic women with humeral fractures, callus formation determined by regional bone mineral density at the fracture site increased in patients taking vitamin D and calcium compared with controls (17). Vitamin D deficiency and obesity appear to be linked. Up to 90% of obese individuals have been found to have vitamin D deficiencies before gastric bypass surgery (18). Also, a strong inverse correlation between vitamin D levels and BMI has been observed (19). In a comparison of Roux-en-Y gastric bypass and sleeve gastrectomy, investigators found that weight loss played a larger role in improving the vitamin D level than did the effect of malabsorption on decreasing the vitamin D level (20). Our study found no association between BMI or laboratory values and the development of nonunion. However, only 7 control and 7 nonunion cases had vitamin D data available within 90 days of the fracture. Therefore, although the median values and ranges were comparable between the groups, the rest of the statistics are unreliable. Although vitamin D deficiency has been in the spotlight of bone healing complications in recent years, one should not overlook the other factors that might be more strongly associated with nonunion. The modification of such factors might be more important for successful bone healing than vitamin D monitoring and supplementation. In an umbrella systematic review of vitamin D and health outcomes, little evidence has suggested that supplementation of vitamin D improves any of the 137 outcomes (21). Our study found that a documented current or historical concern of alcohol consumption was significantly associated with progression to nonunion (odds ratio 2.7, 95% confidence interval 1.2 to 6.2). The association between the effect of alcohol and impaired fracture healing is unclear. Previous data have shown an association between excessive alcohol consumption and osteoporosis (22), an increased fracture incidence (23), inhibition of osteoblast cell activity (24,25), impaired callus formation (26), and the development of Charcot arthropathy from alcohol-induced peripheral neuropathy (27). However, few studies have evaluated the effect of alcohol abuse in fracture healing and as a cause of nonunion. In a series of 25 cases of mandibular fracture nonunion, 18 patients were identified as having drug and alcohol abuse problems (28). In 1995, Foulk and Szabo (11) found that alcohol abuse was associated with nonunion in 38 humeral fractures; however, the number of patients with alcohol abuse in each of the case and control groups was not clear. In a study of 122 patients with femoral neck fractures, osseous complications were associated with a history of alcohol abuse (29). In that study, the osseous complications included 9 cases of nonunion, 14 cases of avascular necrosis, and 16 cases of a loss of fixation. Other investigators have suggested that because of confounding comorbidities and the methodologic shortcomings of these studies, alcohol might be seen artifactually to play an important role in bone healing (30). In the present study, we accounted for many of these comorbidities, including smoking, diabetes, age, malnutrition, and peripheral neuropathy, which were identified by both medical record review and a review of the laboratory values. We found that only alcohol affected the outcome of interest as an independent predictor (Tables 2 and 3). The present study had several limitations. First, we performed a retrospective medical record review that was limited by the amount
of data documented in the medical records. Second, we relied on radiographic union criteria for the assessment of union, because relying on medical record review of clinical assessment of union was deemed inadequate. Although many studies have used both clinical and radiographic evaluations to determine the presence of nonunion (31), it has been suggested that radiographic and clinical union data cannot be used interchangeably (32). The present study also used very specific criteria for categorization of union and nonunion to allow for a clear interpretation of groups. In conclusion, as obesity becomes more common, fractures in the obese population will become more frequent. Concerns have been raised about the alteration of bone health and healing in obese patients. Our study found no association between obesity and nonunion, but our findings were suggestive of a connection between patients with a history of an alcohol problem or a current alcohol problem might have a greater risk of nonunion. Further studies are warranted to confirm these findings. Acknowledgments We would like to thank Yang Liu, MS, for assistance in the analysis and Lori Murdoch, RN, for assistance with data collection. References 1. Centers for Disease Control and Prevention. Defining Overweight and Obesity 2012. Available at: http://www.cdc.gov/obesity/adult/defining.html. Accessed March 30 2013. 2. Edelstein SL, Barrett-Connor E. Relation between body size and bone mineral density in elderly men and women. Am J Epidemiol 138:160–169, 1993. 3. Cohen A, Dempster DW, Recker RR, Lappe JM, Zhou H, Zwahlen A, Muller R, Zhao B, Guo X, Lang T, Saeed I, Liu XS, Guo XE, Cremers S, Rosen CJ, Stein EM, Nickolas TL, McMahon DJ, Young P, Shane E. 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, 2013. 4. Xu L, Nicholson P, Wang QJ, Wang Q, Alen M, Cheng S. Fat mass accumulation compromises bone adaptation to load in Finnish women: a cross-sectional study spanning three generations. J Bone Miner Res 25:2341–2349, 2010. 5. Cao JJ. Effects of obesity on bone metabolism. J Orthop Surg Res 6:30, 2011. 6. Felson DT, Zhang Y, Hannan MT, Anderson JJ. Effects of weight and body mass index on bone mineral density in men and women: the Framingham study. J Bone Miner Res 8:567–573, 1993. 7. Jupiter JB, Ring D, Rosen H. The complications and difficulties of management of nonunion in the severely obese. J Orthop Trauma 9:363–370, 1995. 8. Lee RJ, Hsu NN, Lenz CM, Leet AI. Does obesity affect fracture healing in children? Clin Orthop Relat Res 471:1208–1213, 2013. 9. Nolte DM, Fusco JV, Peterson ME. Incidence of and predisposing factors for nonunion of fractures involving the appendicular skeleton in cats: 18 cases (19982002). J Am Vet Med Assoc 226:77–82, 2005. 10. Decomas A, Kaye J. Risk factors associated with failure of treatment of humeral diaphyseal fractures after functional bracing. J La State Med Soc 162:33–35, 2010. 11. Foulk DA, Szabo RM. Diaphyseal humerus fractures: natural history and occurrence of nonunion. Orthopedics 18:333–335, 1995. 12. Tucker MC, Schwappach JR, Leighton RK, Coupe K, Ricci WM. Results of femoral intramedullary nailing in patients who are obese versus those who are not obese: a prospective multicenter comparison study. J Orthop Trauma 21:523–529, 2007. 13. Boszczyk AM, Zakrzewski P, Pomianowski S. Vitamin D concentration in patients with normal and impaired bone union. Pol Orthop Traumatol 78:1–3, 2013. 14. Brinker MR, O’Connor DP, Monla YT, Earthman TP. Metabolic and endocrine abnormalities in patients with nonunions. J Orthop Trauma 21:557–570, 2007. 15. Pourfeizi HH, Tabriz A, Elmi A, Aslani H. Prevalence of vitamin D deficiency and secondary hyperparathyroidism in nonunion of traumatic fractures. Acta Med Iran 51:705–710, 2013. 16. Melhus G, Solberg LB, Dimmen S, Madsen JE, Nordsletten L, Reinholt FP. Experimental osteoporosis induced by ovariectomy and vitamin D deficiency does not markedly affect fracture healing in rats. Acta Orthop 78:393–403, 2007. 17. Doetsch AM, Faber J, Lynnerup N, Watjen I, Bliddal H, Danneskiold-Samsoe B. The effect of calcium and vitamin D3 supplementation on the healing of the proximal humerus fracture: a randomized placebo-controlled study. Calcif Tissue Int 75:183–188, 2004. 18. Goldner WS, Stoner JA, Thompson J, Taylor K, Larson L, Erickson J, McBride C. Prevalence of vitamin D insufficiency and deficiency in morbidly obese patients: a comparison with non-obese controls. Obes Surg 18:145–150, 2008. 19. Lin E, Armstrong-Moore D, Liang Z, Sweeney JF, Torres WE, Ziegler TR, Tangpricha V, Gletsu-Miller N. Contribution of adipose tissue to plasma 25hydroxyvitamin D concentrations during weight loss following gastric bypass surgery. Obesity (Silver Spring) 19:588–594, 2011.
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
20. Coupaye M, Breuil MC, Riviere P, Castel B, Bogard C, Dupre T, Msika S, Ledoux S. Serum vitamin D increases with weight loss in obese subjects 6 months after Roux-en-Y gastric bypass. Obes Surg 23:486–493, 2013. 21. Theodoratou E, Tzoulaki I, Zgaga L, Ioannidis JP. Vitamin D and multiple health outcomes: umbrella review of systematic reviews and meta-analyses of observational studies and randomised trials. BMJ 348:g2035, 2014. 22. Friday KE, Howard GA. Ethanol inhibits human bone cell proliferation and function in vitro. Metabolism 40:562–565, 1991. 23. Kristensson H, Lunden A, Nilsson BE. Fracture incidence and diagnostic roentgen in alcoholics. Acta Orthop Scand 51:205–207, 1980. 24. Dyer SA, Buckendahl P, Sampson HW. Alcohol consumption inhibits osteoblastic cell proliferation and activity in vivo. Alcohol 16:337–341, 1998. 25. Chakkalakal DA. Alcohol-induced bone loss and deficient bone repair. Alcohol Clin Exp Res 29:2077–2090, 2005. 26. Lauing KL, Roper PM, Nauer RK, Callaci JJ. Acute alcohol exposure impairs fracture healing and deregulates beta-catenin signaling in the fracture callus. Alcohol Clin Exp Res 36:2095–2103, 2012.
5
27. Shibuya N, La Fontaine J, Frania SJ. Alcohol-induced neuroarthropathy in the foot: a case series and review of literature. J Foot Ankle Surg 47: 118–124, 2008. 28. Mathog RH, Toma V, Clayman L, Wolf S. Nonunion of the mandible: an analysis of contributing factors. J Oral Maxillofac Surg 58:746–753, 2000. 29. Duckworth AD, Bennet SJ, Aderinto J, Keating JF. Fixation of intracapsular fractures of the femoral neck in young patients: risk factors for failure. J Bone Joint Surg Br 93:811–816, 2011. 30. Thevendran G, Younger A, Pinney S. Current concepts review: risk factors for nonunions in foot and ankle arthrodeses. Foot Ankle Int 33:1031–1040, 2012. 31. Corrales LA, Morshed S, Bhandari M, Miclau T III. Variability in the assessment of fracture-healing in orthopaedic trauma studies. J Bone Joint Surg Am 90:1862– 1868, 2008. 32. Thorud JC, Jolley T, Shibuya N, Lew E, Britt M, Butterfield T, Boike A, Hardy M, Brancheau S, Motley T, Jupiter DC. Comparison of hallux interphalangeal joint arthrodesis fixation techniques: a retrospective multicenter study. J Foot Ankle Surg 55:22–27, 2015.
Contents lists available at ScienceDirect
The Journal of Foot & Ankle Surgery journal homepage: www.jfas.org
Original Research
Effect of Obesity on Bone Healing After Foot and Ankle Long Bone Fractures Jakob C. Thorud, DPM, MS, AACFAS 1, 2, Spencer Mortensen, DPM, AACFAS 3, 4, 5, Jennifer L. Thorud, MPH 6, Naohiro Shibuya, DPM, MS, FACFAS 1, 2, 7, Yolanda Munoz Maldonado, PhD 8, Daniel C. Jupiter, PhD 9 1
Podiatrist, Central Texas Veterans Affairs Health Care System, Temple, TX Podiatrist, Baylor Scott & White Health, Temple, TX Resident, Central Texas Veterans Affairs Health Care System, Temple, TX 4 Resident, Baylor Scott & White Health, Temple, TX 5 Resident, Texas A&M Health Science Center, College of Medicine, Bryan, TX 6 Research Associate, Baylor Scott & White Health, Temple, TX 7 Professor, Texas A&M Health Science Center, College of Medicine, Bryan, TX 8 Biostatistician, Baylor Scott & White Health, Temple, TX 9 Assistant Professor, Department of Preventive Medicine and Community Health, University of Texas Medical Branch, Galveston, TX 2 3
a r t i c l e i n f o
a b s t r a c t
Level of Clinical Evidence: 3
As obesity has become more common, fractures in the obese population have become more frequent. Concern exists regarding alterations in bone health and healing in obese patients. A matched case-control study was performed at 1 institution to evaluate whether an association exists between nonunion and a high body mass index in metatarsal and ankle fractures. A total of 48 patients with nonunion were identified, and control patients matched 2 to 1 (n ¼ 96) were selected. The control patients were matched for age, sex, and fracture type. No association was identified between nonunion and the continuous body mass index (p ¼ .23) or morbid obesity, with a body mass index of 40 kg/m2 (p ¼ .51). However, the results from both univariate and multivariate analysis suggested that patients with a current alcohol problem or a history of an alcohol problem might have a greater risk of nonunion. The odds ratio of a patient with a history of alcohol use experiencing nonunion was 2.7 (95% confidence interval 1.2 to 6.2). Further studies are warranted to confirm these findings. Ó 2016 by the American College of Foot and Ankle Surgeons. All rights reserved.
Keywords: alcohol BMI body mass index fracture healing metatarsal fracture
Obesity is a growing problem in the United States. In 2009 to 2010, the Centers for Disease Control and Prevention National Center for Health Statistics found that 35.5% of men were considered obese, a significant increase from the 27.5% in 1999 to 2000. These numbers were obtained using body mass index (BMI), a widely accepted method of grossly assessing obesity in large populations (1). Increased body weight has traditionally been thought to be protective against fractures owing to the increased bone density (2). However, recent studies have questioned this supposition. Cohen et al (3) found that trunk fat is associated with poor bone quality, decreased trabecular bone volume, decreased stiffness, greater cortical porosity, and decreased bone formation in premenopausal women with a normal
Financial Disclosure: None reported. Conflict of Interest: None reported. Address correspondence to: Jakob C. Thorud, DPM, MS, AACFAS, 3922 Mercy Drive, McHenry, IL 60050. E-mail address: [email protected] (J.C. Thorud).
bone mineral density. Likewise, a study across 3 age groups of women found that the relative bone strength decreased with an increasing fat mass, and this relationship strengthened with increasing age (4). Cao (5) suggested that increased adiposity could have detrimental effects on bone by increasing inflammatory cytokine levels, which have negative effects on bone metabolism. Although most studies have found increased bone mineral density with increased weight (6), inferior bone quality is more likely to negatively affect bone healing after fracture. An association with nonunion and obesity has been reported; however, these studies had small sample sizes and did not account for possible confounders (7–11). This was largely because of the difficulty in obtaining a large sample of patients with nonunion. Because, fortunately, nonunions are relatively uncommon, a large population of patients with fracture must be reviewed to find the relatively few cases of nonunion. In the present study, we first reviewed the data from all patients with ankle and metatarsal fractures to find those with a diagnosis of nonunion within the first 18 months after the
1067-2516/$ - see front matter Ó 2016 by the American College of Foot and Ankle Surgeons. All rights reserved. http://dx.doi.org/10.1053/j.jfas.2016.11.010
2
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
Table 1 Definitions used for medical record review Definition Radiographic union: meets all 3 of the following criteria Presence of 2 cortical continuations or bridging at fusion site Absence of hardware failure (breakage or lucent region surrounding hardware) Absence of lytic gapping of fracture site Delayed union: failure to meet criteria for union between 3 and 6 mo Nonunion: failure to meet criteria for union after 6 mo or failure to meet criteria with correctional surgery before 6 mo after fracture diagnosis Smoking history Active: smoked within previous 2 wk or returned to smoking during postoperative period Quit: stopped smoking >2 wk before surgery and did not restart during postoperative period Never: never smoked Unknown: smoking status not recorded Alcohol history: history or current documented concern of alcohol consumption
fracture. We then selected matched controls to obtain a large cohort of patients with nonunion. The present study compared the BMI of patients with long bone fractures of the foot and ankle that had progressed to nonunion and those that had healed uneventfully. To determine whether the BMI was significantly different between these 2 groups, a retrospective medical record review was performed to compare patients with radiographic nonunion with those with radiographic union. Patients and Methods The present study was a retrospective review of patients aged 18 to 90 years, selected using the International Classification of Diseases, 9th Revision, Clinical
Modification (ICD-9-CM), diagnosis codes for fractures of the ankle and metatarsals, presenting from January 1, 2006 to January 1, 2012 to Baylor Scott & White Health Memorial Hospital. We included both patients who had undergone operative and nonoperative treatment of fracture. Patients who died before 18 months after the fracture diagnosis were excluded. Pathologic fractures were excluded from the present study. The diagnosis of nonunion (ICD-9-CM code 733.82) within 18 months after the fracture diagnosis (ICD-9-CM codes 824 open or closed ankle fracture, 825.25 metatarsal fracture, closed, and 825.35 metatarsal fracture, open) were used to identify the possible patients with nonunion. A medical record review was then undertaken to confirm the diagnosis of nonunion using the criteria listed in Table 1. An example of a nonunion identified using the criteria is shown in the Fig. Patients without sufficient radiographic data to determine the presence of nonunion were excluded from the study (i.e., patients without enough follow-up data available). Laboratory data were recorded only if they were collected within 90 days of the fracture date. After identification and confirmation of all nonunion cases, matched control patients without a diagnosis of nonunion were selected using the same parameters and from the same patient population. The control patients were matched for age, sex, and fracture type in a 2:1 ratio to those with nonunion. Comparisons of the continuous variables were performed using the average values for the controls and a paired t test was used for the comparison between the cases and controls. McNemar’s tests were used to compare the clinical and radiologic outcomes for each of the controls against the respective case. A conditional logistic regression model was fit to the data to assess for possible significant factors for nonunion. The candidate predictors in the model were alcohol use, diabetes, tobacco history, renal disease, and BMI as a categorical variable. Forward, backward, and stepwise regression all selected alcohol use as the only significant predictor. The final model included alcohol use and BMI coded as a categorical variable as predictors. BMI was included in the final model because that was the aim of the present study. All analyses were completed using SAS, version 9.4 (SAS Institute, Cary, NC).
Results A total of 48 patients with nonunion were identified, and 96 control patients were matched. The mean age of the case patients with nonunion and control patients with union was 50.8 14.5 years
Fig. Radiographic evidence of nonunion. Ankle radiographs of nonunion in a 51-year-old female with a body mass index of 40 kg/m2 with a medial malleolar fracture originally fixated >8 months earlier. The parallel lines indicate the absence of cortical continuations. The bracket indicates gapping of the fracture fragments. The white arrow indicates the lucency surrounding screws.
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
Table 2 Demographic characteristics (N ¼ 144) Characteristic
Nonunion (n ¼ 48)
Control (n ¼ 96)
Total
p Value
Age (yr) Sex Female Male Race White Black Other BMI (kg/m2) BMI group BMI <40 kg/m2 BMI 40 kg/m2 Diabetes Peripheral neuropathy Malnutrition Vitamin D deficiency Renal disease Alcohol history Tobacco use Active Quit Never smoked
50.8 14.5
51.3 15.3
51.1 15.0
.48* .99y
30 (62.5) 18 (37.5)
58 (60.4) 38 (39.6)
88 (61.1) 56 (38.9)
36 (75.0) 4 (8.3) 8 (16.7) 32.3 7.1
70 (72.9) 10 (10.4) 16 (16.7) 29.6 7.0
106 (73.6) 14 (9.7) 24 (16.7) 31.4 7.1
40 8 15 8 2 5 8 18
84 12 21 11 2 8 8 19
124 20 36 19 4 13 16 37
.92y
(83.3) (16.7) (31.3) (16.7) (4.2) (10.4) (16.7) (37.5)
12 (26.1) 13 (28.3) 21 (45.7)
(87.5) (12.5) (21.9) (11.5) (2.1) (8.3) (8.3) (19.8)
17 (18.1) 13 (13.8) 64 (68.1)
(86.1) (13.9) (25.0) (13.2) (2.8) (9.0) (11.1) (25.7)
.23* .51y .17y .37y .49y .68y .13y .02y .06y
29 (20.7) 26 (18.6) 85 (60.7)
Abbreviation: BMI, body mass index. Data presented as n (%) for categorical variables and mean standard deviation for continuous variables. * Age and BMI continuous data were analyzed with a paired t test pairing the case against the average value of the controls. y Categorical data were analyzed using McNemar’s test.
and 51.3 15.3 years, respectively. The demographic characteristics for the cohort are listed in Table 2. The combined group of cases and controls experienced 75 ankle fractures and 69 metatarsal fractures. The effect of the BMI, measured as both categorical (<40 versus 40 kg/m2) and continuous data, was similar for the cases and matched controls. Of the 48 patients with nonunion and 96 controls with union, 8 and 12, respectively, were categorized as morbidly obese (p ¼ .23). The mean BMI of those with nonunion and those with union was 32.3 7.1 kg/m2 and 29.6 7.0 kg/m2, respectively (p ¼ .51). No initial surgical operation was needed in most cases; 57 of all 144 patients (cases and controls; 39.6%) required surgery overall. No difference was seen in the initial operations performed between the 2 groups, with 40 of the 96 controls and 17 of the 48 patients with nonunion undergoing an initial operation (p ¼ .67). No significant difference was found between the cases and controls in terms of age, sex, or race. Only a history of alcohol use was a significant factor for a patient to experience nonunion (p ¼ .02). The laboratory values for the cases and controls are listed in Table 3. No association was found between the calcium, albumin, blood urea nitrogen, or creatinine levels and nonunion. Vitamin D had too few recorded values to be analyzed, and parathyroid hormone was not analyzed owing to the small sample size. In the conditional
3
logistic regression model, no association was found for calcium, albumin, vitamin D, blood urea nitrogen, or creatinine levels and nonunion. The adjusted odds ratio estimates for alcohol history and BMI as a categorical variable are listed in Table 4. The odds ratio for patients with a history of alcohol use experiencing nonunion was 2.7 (95% confidence interval 1.2 to 6.2) and the odds ratio for patients with a BMI of 40 kg/m2 was 1.5 (95% confidence interval 0.6 to 3.9). Discussion The association of nonunion with obesity in patients has been documented previously; however, these studies were small and reported conflicting results. In a retrospective study of 344 cats with fractures, 18 developed nonunion. Cats with nonunion weighed significantly more than did the cats with union (9). Jupiter et al (7) found 16 cases of nonunion in 21 obese patients with long bone fractures of the upper and lower extremities. Decomas and Kaye (10) found that 7 of 19 patients (37%) with nonunion of the humerus were obese. However, no control group was used to compare the rates of obesity in these studies. In a retrospective study of 38 nonoperatively treated diaphyseal humeral fractures, 15 developed nonunion (11). The investigators stated that obesity was significantly associated with nonunion; however, they did not provide the number of obese patients within each group. Tucker et al (12) prospectively followed up 113 patients with femoral shaft fractures treated with intramedullary nailing. Much controversy exists regarding bone healing and vitamin D deficiencies. In a study of 35 patients with nonunion compared with 35 patients with bony union of closed fractures of the extremities, 86% of the patients had vitamin D deficiencies (defined as <30 ng/mL) in both the union and the nonunion groups. Although no significant difference was found, the study had a small sample size (13). Brinker et al (14) reviewed the data from 37 patients with nonunion of the upper and lower extremities, without obvious causes such as inadequate fixation, who had been referred to an endocrinologist from a large set of 683 consecutive patients with nonunion. Of the 37 patients, 83.8% had a newly diagnosed metabolic or endocrine abnormality, of which vitamin D deficiency was the most common, occurring in 25 patients (14). It is unknown how many of the patients without nonunion also had a vitamin D deficiency. More recently, 30 cases of unexplained tibial fracture nonunion were matched to 32 cases of tibial fractures with normal healing. These were matched by surgical treatment, age, sex, and BMI, and those with nonunion had significantly lower vitamin D levels than the levels of the controls (25.8 20.4 nmol/L versus 49.03 26.9 nmol/L) (15). In a study comparing ovariectomized and vitamin D-deficient rats (mimicking the fracture healing commonly found in postmenopausal women) against control rats, Melhus et al (16) found no difference in healing rates, mean bone mineral density of the callus, or mechanical strength
Table 3 Laboratory values stratified by group (N ¼ 144) Serum Laboratory Data
Nonunion (n ¼ 48)
Control (n ¼ 96)
Total
p Value
Calcium (mg/dL) Albumin (g/dL) 25(OH)D (ng/mL) PTH (pg/mL) BUN (mg/dL) Creatinine (mg/dL)
21; 9.1 0.6 20; 3.8 0.6 7; 39 (24 to 72) 5; 12 (3 to 73) 22; 17 (8 to 72) 24; 1.0 (0.6 to 3.3)
44; 9.0 0.5 43; 3.8 0.5 7; 33 (25 to 55) 2; 150 (107 to 193) 47;15 (6 to 84) 48; 0.9 (0.5 to 4.1)
65; 9.0 0.5 63; 3.8 0.5 14; 36 (24 to 72) 7; 38 (3 to 193) 69; 16 (6 to 84) 72; 0.9 (0.5 to 4.1)
.10* .37* NA NA .88y .87y
Abbreviations: BUN, blood urea nitrogen; 25(OH)D, 25-hydroxyvitamin D; NA, not applicable (not tested because of small sample size); PTH, parathyroid hormone. Data presented as n and mean standard deviation or n and median (range), as appropriate. * Calcium and albumin were normally distributed and analyzed using the paired t test. y BUN and creatinine were not normally distributed and the Wilcoxon sign rank test was used.
4
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
Table 4 Adjusted odds ratio estimates (N ¼ 144) Effect
Point Estimate
95% CI
Alcohol history (yes versus no) BMI (40 versus <40)
2.69 1.47
1.16 6.21 0.55 3.92
Abbreviations: BMI, body mass index; CI, confidence interval.
of bone, except for stiffness of the right femoral neck. In contrast, in a randomized, double-blind, placebo-controlled study of 30 osteoporotic women with humeral fractures, callus formation determined by regional bone mineral density at the fracture site increased in patients taking vitamin D and calcium compared with controls (17). Vitamin D deficiency and obesity appear to be linked. Up to 90% of obese individuals have been found to have vitamin D deficiencies before gastric bypass surgery (18). Also, a strong inverse correlation between vitamin D levels and BMI has been observed (19). In a comparison of Roux-en-Y gastric bypass and sleeve gastrectomy, investigators found that weight loss played a larger role in improving the vitamin D level than did the effect of malabsorption on decreasing the vitamin D level (20). Our study found no association between BMI or laboratory values and the development of nonunion. However, only 7 control and 7 nonunion cases had vitamin D data available within 90 days of the fracture. Therefore, although the median values and ranges were comparable between the groups, the rest of the statistics are unreliable. Although vitamin D deficiency has been in the spotlight of bone healing complications in recent years, one should not overlook the other factors that might be more strongly associated with nonunion. The modification of such factors might be more important for successful bone healing than vitamin D monitoring and supplementation. In an umbrella systematic review of vitamin D and health outcomes, little evidence has suggested that supplementation of vitamin D improves any of the 137 outcomes (21). Our study found that a documented current or historical concern of alcohol consumption was significantly associated with progression to nonunion (odds ratio 2.7, 95% confidence interval 1.2 to 6.2). The association between the effect of alcohol and impaired fracture healing is unclear. Previous data have shown an association between excessive alcohol consumption and osteoporosis (22), an increased fracture incidence (23), inhibition of osteoblast cell activity (24,25), impaired callus formation (26), and the development of Charcot arthropathy from alcohol-induced peripheral neuropathy (27). However, few studies have evaluated the effect of alcohol abuse in fracture healing and as a cause of nonunion. In a series of 25 cases of mandibular fracture nonunion, 18 patients were identified as having drug and alcohol abuse problems (28). In 1995, Foulk and Szabo (11) found that alcohol abuse was associated with nonunion in 38 humeral fractures; however, the number of patients with alcohol abuse in each of the case and control groups was not clear. In a study of 122 patients with femoral neck fractures, osseous complications were associated with a history of alcohol abuse (29). In that study, the osseous complications included 9 cases of nonunion, 14 cases of avascular necrosis, and 16 cases of a loss of fixation. Other investigators have suggested that because of confounding comorbidities and the methodologic shortcomings of these studies, alcohol might be seen artifactually to play an important role in bone healing (30). In the present study, we accounted for many of these comorbidities, including smoking, diabetes, age, malnutrition, and peripheral neuropathy, which were identified by both medical record review and a review of the laboratory values. We found that only alcohol affected the outcome of interest as an independent predictor (Tables 2 and 3). The present study had several limitations. First, we performed a retrospective medical record review that was limited by the amount
of data documented in the medical records. Second, we relied on radiographic union criteria for the assessment of union, because relying on medical record review of clinical assessment of union was deemed inadequate. Although many studies have used both clinical and radiographic evaluations to determine the presence of nonunion (31), it has been suggested that radiographic and clinical union data cannot be used interchangeably (32). The present study also used very specific criteria for categorization of union and nonunion to allow for a clear interpretation of groups. In conclusion, as obesity becomes more common, fractures in the obese population will become more frequent. Concerns have been raised about the alteration of bone health and healing in obese patients. Our study found no association between obesity and nonunion, but our findings were suggestive of a connection between patients with a history of an alcohol problem or a current alcohol problem might have a greater risk of nonunion. Further studies are warranted to confirm these findings. Acknowledgments We would like to thank Yang Liu, MS, for assistance in the analysis and Lori Murdoch, RN, for assistance with data collection. References 1. Centers for Disease Control and Prevention. Defining Overweight and Obesity 2012. Available at: http://www.cdc.gov/obesity/adult/defining.html. Accessed March 30 2013. 2. Edelstein SL, Barrett-Connor E. Relation between body size and bone mineral density in elderly men and women. Am J Epidemiol 138:160–169, 1993. 3. Cohen A, Dempster DW, Recker RR, Lappe JM, Zhou H, Zwahlen A, Muller R, Zhao B, Guo X, Lang T, Saeed I, Liu XS, Guo XE, Cremers S, Rosen CJ, Stein EM, Nickolas TL, McMahon DJ, Young P, Shane E. 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, 2013. 4. Xu L, Nicholson P, Wang QJ, Wang Q, Alen M, Cheng S. Fat mass accumulation compromises bone adaptation to load in Finnish women: a cross-sectional study spanning three generations. J Bone Miner Res 25:2341–2349, 2010. 5. Cao JJ. Effects of obesity on bone metabolism. J Orthop Surg Res 6:30, 2011. 6. Felson DT, Zhang Y, Hannan MT, Anderson JJ. Effects of weight and body mass index on bone mineral density in men and women: the Framingham study. J Bone Miner Res 8:567–573, 1993. 7. Jupiter JB, Ring D, Rosen H. The complications and difficulties of management of nonunion in the severely obese. J Orthop Trauma 9:363–370, 1995. 8. Lee RJ, Hsu NN, Lenz CM, Leet AI. Does obesity affect fracture healing in children? Clin Orthop Relat Res 471:1208–1213, 2013. 9. Nolte DM, Fusco JV, Peterson ME. Incidence of and predisposing factors for nonunion of fractures involving the appendicular skeleton in cats: 18 cases (19982002). J Am Vet Med Assoc 226:77–82, 2005. 10. Decomas A, Kaye J. Risk factors associated with failure of treatment of humeral diaphyseal fractures after functional bracing. J La State Med Soc 162:33–35, 2010. 11. Foulk DA, Szabo RM. Diaphyseal humerus fractures: natural history and occurrence of nonunion. Orthopedics 18:333–335, 1995. 12. Tucker MC, Schwappach JR, Leighton RK, Coupe K, Ricci WM. Results of femoral intramedullary nailing in patients who are obese versus those who are not obese: a prospective multicenter comparison study. J Orthop Trauma 21:523–529, 2007. 13. Boszczyk AM, Zakrzewski P, Pomianowski S. Vitamin D concentration in patients with normal and impaired bone union. Pol Orthop Traumatol 78:1–3, 2013. 14. Brinker MR, O’Connor DP, Monla YT, Earthman TP. Metabolic and endocrine abnormalities in patients with nonunions. J Orthop Trauma 21:557–570, 2007. 15. Pourfeizi HH, Tabriz A, Elmi A, Aslani H. Prevalence of vitamin D deficiency and secondary hyperparathyroidism in nonunion of traumatic fractures. Acta Med Iran 51:705–710, 2013. 16. Melhus G, Solberg LB, Dimmen S, Madsen JE, Nordsletten L, Reinholt FP. Experimental osteoporosis induced by ovariectomy and vitamin D deficiency does not markedly affect fracture healing in rats. Acta Orthop 78:393–403, 2007. 17. Doetsch AM, Faber J, Lynnerup N, Watjen I, Bliddal H, Danneskiold-Samsoe B. The effect of calcium and vitamin D3 supplementation on the healing of the proximal humerus fracture: a randomized placebo-controlled study. Calcif Tissue Int 75:183–188, 2004. 18. Goldner WS, Stoner JA, Thompson J, Taylor K, Larson L, Erickson J, McBride C. Prevalence of vitamin D insufficiency and deficiency in morbidly obese patients: a comparison with non-obese controls. Obes Surg 18:145–150, 2008. 19. Lin E, Armstrong-Moore D, Liang Z, Sweeney JF, Torres WE, Ziegler TR, Tangpricha V, Gletsu-Miller N. Contribution of adipose tissue to plasma 25hydroxyvitamin D concentrations during weight loss following gastric bypass surgery. Obesity (Silver Spring) 19:588–594, 2011.
J.C. Thorud et al. / The Journal of Foot & Ankle Surgery xxx (2016) 1–5
20. Coupaye M, Breuil MC, Riviere P, Castel B, Bogard C, Dupre T, Msika S, Ledoux S. Serum vitamin D increases with weight loss in obese subjects 6 months after Roux-en-Y gastric bypass. Obes Surg 23:486–493, 2013. 21. Theodoratou E, Tzoulaki I, Zgaga L, Ioannidis JP. Vitamin D and multiple health outcomes: umbrella review of systematic reviews and meta-analyses of observational studies and randomised trials. BMJ 348:g2035, 2014. 22. Friday KE, Howard GA. Ethanol inhibits human bone cell proliferation and function in vitro. Metabolism 40:562–565, 1991. 23. Kristensson H, Lunden A, Nilsson BE. Fracture incidence and diagnostic roentgen in alcoholics. Acta Orthop Scand 51:205–207, 1980. 24. Dyer SA, Buckendahl P, Sampson HW. Alcohol consumption inhibits osteoblastic cell proliferation and activity in vivo. Alcohol 16:337–341, 1998. 25. Chakkalakal DA. Alcohol-induced bone loss and deficient bone repair. Alcohol Clin Exp Res 29:2077–2090, 2005. 26. Lauing KL, Roper PM, Nauer RK, Callaci JJ. Acute alcohol exposure impairs fracture healing and deregulates beta-catenin signaling in the fracture callus. Alcohol Clin Exp Res 36:2095–2103, 2012.
5
27. Shibuya N, La Fontaine J, Frania SJ. Alcohol-induced neuroarthropathy in the foot: a case series and review of literature. J Foot Ankle Surg 47: 118–124, 2008. 28. Mathog RH, Toma V, Clayman L, Wolf S. Nonunion of the mandible: an analysis of contributing factors. J Oral Maxillofac Surg 58:746–753, 2000. 29. Duckworth AD, Bennet SJ, Aderinto J, Keating JF. Fixation of intracapsular fractures of the femoral neck in young patients: risk factors for failure. J Bone Joint Surg Br 93:811–816, 2011. 30. Thevendran G, Younger A, Pinney S. Current concepts review: risk factors for nonunions in foot and ankle arthrodeses. Foot Ankle Int 33:1031–1040, 2012. 31. Corrales LA, Morshed S, Bhandari M, Miclau T III. Variability in the assessment of fracture-healing in orthopaedic trauma studies. J Bone Joint Surg Am 90:1862– 1868, 2008. 32. Thorud JC, Jolley T, Shibuya N, Lew E, Britt M, Butterfield T, Boike A, Hardy M, Brancheau S, Motley T, Jupiter DC. Comparison of hallux interphalangeal joint arthrodesis fixation techniques: a retrospective multicenter study. J Foot Ankle Surg 55:22–27, 2015.