Bone 59 (2014) 207210

Contents lists available at ScienceDirect

Bone
journal homepage: www.elsevier.com/locate/bone

Original Full Length Article

A United Kingdom perspective on the relationship between body mass

index (BMI) and bone health: A cross sectional analysis of data from the
Nottingham Fracture Liaison Service
Terence Ong a,, Opinder Sahota a, Wei Tan b, Lindsey Marshall c
a
Department of Healthcare for Older People, Queens Medical Centre, Nottingham NG7 2UH, UK
b
Nottingham Clinical Trials Unit, University of Nottingham, Queens Medical Centre, Nottingham NG7 2UH, UK
c
Department of Trauma and Orthopaedics, Queens Medical Centre, Nottingham NG7 2UH, UK

a r t i c l e i n f o a b s t r a c t

Article history: Objectives: This study examines the relationship between high BMI, a diagnosis of osteoporosis and low trauma
Received 30 August 2013 fractures.
Revised 29 October 2013 Method: This is a cross sectional analysis using data collected from the Nottingham Fracture Liaison Service.
Accepted 24 November 2013 A total of 4288 participants with a low trauma fracture from 1 January 2007 to 31 August 2012 were analysed.
Available online 28 November 2013
Logistic regression adjusted for potential confounders was used investigate osteoporosis and BMI. Fracture
Edited by: N. Guanabens
types were compared between those who were obese and non-obese.
Results: A total of 30% (1285) were obese. Prevalence of osteoporosis was 13.4%, 24.9%, and 40.4% in the obese,
Keywords: overweight and normal category respectively. Being obese has an odds ratio of 0.23 (95% CI 0.190.28,
Obesity p b 0.01) of having osteoporosis compared to a normal BMI category. When variable BMI cut offs were used
Body mass index (BMI 25, 30 and 35) to calculate the positive predictive value of patients not having osteoporosis, it was 80.5%,
Bone fractures 86.3% and 88.3%. Examining fracture types, obese patients when compared with the non-obese category, were
Bone fragility more likely to fracture their ankle (OR 1.48, p b 0.01) and upper arm (OR 1.48, p b 0.001), but were less likely
Bone density to fracture their wrist (OR 0.65, p b 0.001). In the elderly (N 70 years), obesity no longer inuenced ankle or
Osteoporosis
wrist fractures but there is an increased risk of upper arm fractures (OR 1.46, p = 0.005).
Conclusion: Higher BMD in obesity is not protective against fractures as there are a signicant number of fractures
in this group which may be due to body habitus, mechanism of injury and the effect of adiposity on bone. A low
trauma osteoporotic fracture will need to be redened in light of these ndings.
2013 Elsevier Inc. All rights reserved.

Introduction osteoporosis and osteoporotic fractures is expected to rise due to an

Obesity and osteoporosis are two major conditions that are highly
prevalent with signicant clinical and public health implication [14].
The World Health Organisation denes obesity as abnormal or excessive
fat accumulation that presents a risk to health. The body mass index
(BMI) is a simple way of measuring one's degree of obesity. A BMI
of more than 25 kg/m2 is considered overweight and more than
30 kg/m2 as obese [1]. A recent health report in England has shown a
rise in the prevalence of obesity in adults from 13% to 24% in men;
and 16% to 26% in women in almost 20 years [2]. It is projected that
by 2025, 47% of men and 36% of women in England will be obese costing
the National Health Service an estimated 21.5 billion in treating
disease attributed to obesity [3]. Osteoporosis is a progressive, skeletal
condition characterised by low bone mass and micro-architectural dete-
rioration in bone tissue with a consequent increase in bone fragility
and susceptibility to fracture [4]. Similar to obesity, the prevalence of
Corresponding author. Fax: +44 115 947 9947.
E-mail address: terenceong@doctors.org.uk (T. Ong).
aging population [5]. It is estimated that by 2050, there will be a 135%
and 57% rise in the numbers of hip and vertebral fractures in Europe [6].
Low BMI has been recognised as a risk factor for low BMD and fragil-
ity fracture. A recent meta-analysis described an inverse and non-linear
relationship between BMI and fracture risk. There was an increasing
fracture risk with decreasing units of BMI b25 kg/m2, but the fracture
risk reduction was less above this BMI [7]. This less straightforward re-
lationship is due to the complex interaction between the mechanical
and hormonal factors in obesity on bones. A larger BMI, as expected,
confers greater mechanical loading on bone which increases its BMD
to accommodate the heavier load [8] and correlates to higher levels of
adipose tissue that play an important role in oestrogen production. Obe-
sity is also associated with hyperinsulinaemia due to a degree of insulin
resistance and low-grade chronic inammatory state with higher levels
of proinammatory cytokines These effects mediate at the cellular level
leading to bone resorption and bone loss [912]. This complex relation-
ship is perhaps the reason that despite the higher BMD obtained in
those with higher BMI, it does not always translate into lower fracture
risk and that studies so far on obesity and bone health have yielded

8756-3282/$ see front matter 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.bone.2013.11.024
208 T. Ong et al. / Bone 59 (2014) 207210

mixed results. The aim of this study is to look at the relationship Table 1
between BMI, osteoporosis and low trauma fractures using data from Baseline characteristics.

the Nottingham Fracture Liaison Service (FLS). BMI categories

Underweight Ideal Overweight Obese

Method (b18.5 kg/m2) (18.524.9 kg/m2) (2529.9 kg/m2) (30 kg/m2)

Patients, n 72 1316 1615 1285

Setting and design
Queens Medical Centre is the East Midlands Trauma Centre and
where patients with an acute fracture are admitted and assessed. It
serves a population of 640,000 (Nottingham City, Broxtowe, Gedling
and Rushcliffe District), which is 1.2% of England's total population. A
total of 32.2% of Nottinghamshire's population are over the age of
50 years old which is almost similar to the total population of England
over 50 years at 34.4% [13]. Patients presenting with a fracture that
does not need inpatient treatment (most commonly non-hip and non-
vertebral fractures) are seen in the outpatient fracture clinic, run by a
dedicated orthopaedic team. The FLS identies those over the age of
50 years attending the outpatient fracture clinic with a low trauma frac-
ture, dened as a fracture after sustaining a fall from a standing height
or less. They are then referred for a dual energy X-ray absorptiometry
(DXA) scan. A diagnosis of osteoporosis is made based on the WHO
denition of a BMD of equal to or less than 2.5 standard deviation
(SD) below a young adult mean. BMD was calculated at the hip and
spine. BMI categories based on WHO denition of obesity were used
(obese: BMI of 30 kg/m2; overweight: 2529.9 kg/m2; ideal: 18.5
24.9 kg/m2; and underweight: b18.5 kg/m2). We analysed data from
the FLS database of patients presenting to the service and had a DXA
scan from 1 January 2007 to 31 August 2012. Information on missing
or incomplete data was obtained through the hospital results and
reporting system. Ethics approval was not needed for this study.
Gender, F/M 64/8 1127/189 1319/296 1056/229
Age, mean(SD) 67.1 (11) 67.0 (10) 67.1 (9) 65.4 (9)
increases the risk of developing osteoporosis and higher BMI confers a
lower risk (Table 2).
The correlation coefcient between BMD and obese and non-obese
patients was measured in the spine, femur neck and total hip. The cor-
relation was signicantly weaker in the obese than non-obese group
at all measured sites (Table 3).
Wrist fracture (52.7%) was the most prevalent fracture, followed by
upper arm (17.2%), ankle (15.5%), elbow (8.7%) and clavicle (1.9%).
When the three most common fractures were analysed, obese patients
are more likely to fracture their ankle (OR 1.48, p b 0.001) and upper
arm (OR 1.48, p b 0.001), but less likely to fracture their wrist
(OR 0.65, p N 0.001) compared to other fractures. When this was sepa-
rated by gender, the relationship only applies to women but not men
(Table 4). When the elderly (N 70 years) were looked at, obesity no
longer inuenced ankle (p = 0.42) or wrist fractures (p = 0.17) but
increased the risk of an upper arm fracture (RR 2.08, p = 0.005).
When arbitrary BMI values of 25, 30 and 35 were used, the PPV of peo-
ple not having osteoporosis was 80.5%, 86.3% and 88.3% respectively
(Table 5).
Discussion

Statistical analysis
Baseline characteristics and fracture types are grouped into different
BMI categories. Logistic regression adjusted for age and gender was
used to investigate the potential risk factors associated with osteoporo-
sis with BMI as continuous and categorical values. Restrictive analysis of
patients using a cut off of 70 years was used to dene the elderly popu-
lation. Arbitrary BMI cut offs of 25 kg/m2, 30 kg/m2 and 35 kg/m2 were
used to calculate sensitivity, specicity, positive predictive value (PPV)
and negative predictive value (NPV) of using DXA scans to diagnosed
osteoporosis. A positive result is dened as not having osteoporosis.
Comparison of BMI categories was done between obese and non-
obese group. The non-obese group excluded those who were under-
weight. Chi-square test was used to calculate the relationship between
obese and non-obese group with different fracture sites. Relative risk
of fracture types by site was calculated in relation to other fractures.
Standard deviation (SD) described is of one standard deviation off the
mean. Statistical signicant was dene as a p value of less than 0.05.
Results
There is a high prevalence of obesity in our cohort and the preva-
lence of osteoporosis in this group is lower compared to other BMI cat-
egories. Studies have shown a relationship between BMD and BMI,
where higher BMI increases BMD [7,14,15] and that losing weight re-
duces BMD values [16]. Although there are fewer patients with a diag-
nosis of osteoporosis in the obese category, there are a signicant
number of fractures occurring in this group. This issue has been
highlighted in other studies that remarked on the signicant number
of fractures in the obese group despite BMD that is not diagnostic of
osteoporosis [1719]. Hence, despite the higher BMD in obese patients,
current literature conrms that it does not appear to be protective
against fractures.
Obesity also appears to be associated with certain fracture types. Our
study showed that when those in the obese group were compared to
the non-obese group, there was a higher risk of upper arm and ankle
fracture, but a lower risk of developing wrist fractures. Some studies
Table 2
Odds ratio, OR, of osteoporosis by BMI categories compared with an ideal BMI.
BMI category OR p 95% CI

A total of 4288 patients were included in this study. A total of 3566 Underweight 5.09 b0.001 2.888.99
Ideal
(83.2%) were women and 722 (16.8%) patients were men. The mean
Overweight 0.46 b0.001 0.390.55
(SD) age was 66 years [10]. The prevalence of obesity in our cohort Obese 0.23 b0.001 0.190.28
was 30%. Baseline characteristics of the patient demographics are
shown in Table 1.
Increasing age is a risk factor for osteoporosis (odds ratio, OR 1.06, Table 3
95% CI 1.051.07, p b 0.001). Men are less likely to develop osteoporosis Correlation coefcient, r, between BMD and BMI in obese and non-obese category.
(OR 0.69, 95% CI 0.560.85, p b 0.001). The prevalence of osteoporosis
Site measured Obese Non-obese
was 13.4%, 24.9%, 40.4% and 75% in the obese, overweight, ideal and un-
r p r p
derweight BMI category respectively. Logistic regression when adjusted
for age and gender, demonstrated that the risk of osteoporosis is less AP spine 0.05 0.08 0.26 b0.001
with increasing BMI (OR 0.88, 95% CI 0.860.89, p b 0.001). When Femur neck 0.06 0.03 0.21 b0.001
Total hip 0.16 b0.001 0.32 b0.001
BMI categories were compared with an ideal BMI, being underweight
 T. Ong et al. / Bone 59 (2014) 207210 209

Table 4 were used, we found that there was a very high predictive value of over
Odds ratio, OR, of obese patients sustaining a fracture compared to non-obese patients. 85% of obese women not having BMD measurements diagnostic of oste-
Fracture Female Male Combined Combined oporosis by WHO criteria. This has implication in terms of patient selec-
site (unadjusted) (adjusted by gender) tion for DXA scans in a healthcare climate of limited resources. Should
OR p OR p OR p OR p this group be started on treatment empirically as they are likely to
have higher BMD values? The decision to treat involves looking at
Ankle 1.38 0.001 1.38 0.089 1.48 b0.001 1.38 b0.001
Upper arm 1.47 b0.001 1.43 0.053 1.48 b0.001 1.46 b0.001 other risk factors for future fractures and is not based solely on BMD.
Wrist 0.63 b0.001 0.73 0.062 0.65 b0.001 0.65 b0.001 Obese women who suffer a fracture exhibit similar clinical risk factors
to non-obese women who fracture [19]. The BMD in obese women
who fracture is also lower when compared to obese women who have
have also shown a higher risk of upper arm fracture [15,20] with higher not suffered a fracture, suggesting that perhaps there is a degree of
BMI; but some have not [18]. Other studies demonstrated higher ankle fragility to these fractures [19,28]. Obese patients are less likely to be
fractures [19] in the obese group, and others have not [15]. Wrist frac- treated due to perceived protective effect of obesity on fracture due to
tures also showed mixed results with one showing a less wrist fractures higher BMD. Fracture assessment tool, such as FRAX [29] has its own
in obesity [19], and another did not show any relationship [15]. The shortcomings when predicting fracture risk in the obese patients [30].
different results could be attributed to sample sizes, different patient Obesity also increases precision errors in DXA scanning [31]. Unfortu-
demographics, differing methodologies and reporting of fractures [21]. nately, there is a lack of studies that look at treating the obese group
Despite this, there appears to be a relationship between obesity and with bone protecting agents who are not osteoporotic. Treating this
fracture site [21,22]. Our study has also shown that gender too may group at higher T-scores will increase healthcare cost.
inuence fracture site. The MrOS [23] study has showed a difference In conclusion, the denition of a low trauma osteoporotic fracture
in fracture risk in men. should be redened in light of current evidence that draws a link
There are limitations to our study. Our cohort was from a fracture between obesity and fracture risk. Threshold for treatment should be
database. Hence, we were unable to compare our cohort with obese reconsidered despite relatively higher BMD attained by obese individ-
individuals without a fracture and results may not be fully applicable uals. Plans and strategies should be put in place to tackle the projected
to the general population. Our FLS service offers all patients with a increase in obesity and fragility fractures. The use of a database and hos-
low trauma fracture a DXA scan except those whom we felt attending pital held records provides an accessible method of conducting research
for a DXA scan and complying with instructions would be difcult, and that ndings are applicable to real life settings.
e.g. those who are functionally dependent in a care home, or have ad-
vance cognitive impairment. Although this may introduce a degree of
selection bias, we feel these numbers to be relatively small and our Declaration of competing interests
practice to be best practice. Due to the data held by the Fracture Liaison
Service, we have no record of medical co-morbidities that will affect All authors declare no support from any organisation for the submit-
bone health, falls or fracture history that will have a confounding effect ted work; no nancial relationships with any organisations that might
on our results. have an interest in the submitted work in the previous three years;
Several reasons can account for the number of fractures and fracture and no other relationships or activities that could appear to have inu-
type seen in obese patients. Obesity is associated with a risk of falls due enced the submitted work.
to muscular weakness and postural instability [18,19,24,25]. There No external funding was obtained for the production of this study.
may also be impairment in the normal protective responses during a
fall (e.g. an outstretched hand) and a predisposition to fall sideways or
backwards. The fat and soft tissue padding distribution in obesity may References
explain why certain sites are more likely to fracture compared to
[1] WHO Technical Report Series. Obesity: preventing and managing the global epidemic.
other sites [19,20]. The higher impact of a fall is likely going to offset WHO; 2000.
the protective response provided by the soft tissue padding. The higher [2] The Health and Social Care Information Centre. Statistics on obesity, physical activity
BMD may just represent an adaptive response without conferring great- and diet: England; 2013.
[3] McPherson K, Marsh T, Brown M. Tackling obesities: future choicesmodelling
er protection against fractures [18]. Besides that, BMD is not on its own a future trends in obesity and the impact on health. 2nd ed. Foresight, Government
marker of bone quality and strength. Bone quality is determined by its Ofce for Science; 2007.
structure, composition of bone matrix, microarchitecture, and the way [4] Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A. Incidence
and economic burden of osteoporosis-related fractures in the United States,
it repairs microdamage [26,27]. Research has shown that fat or adiposity 20052025. J Bone Miner Res 2007;22:46575.
has hormonal inuences on bone mechanism [9,11]. Our study has shown [5] WHO Technical Report Series. Assessment of fracture risk and its application to
a stronger correlation between BMD with BMI up until 30 kg/m2 at all screening for postmenopausal osteoporosis. WHO; 1994.
[6] Compston JE, Papapoulos SE, Blanchard F. Report on osteoporosis in the European
sites measured. The degree of adiposity and its impact on bone may ex-
community: current status and recommendations for the future. Osteoporos Int
plain the weaker BMD correlation in the obese group, although further 1998;8:5314.
studies will be needed. [7] De Laet C, Kanis JA, Odn A, Johanson H, Johnell O, Delmas P, et al. Body mass index
An important issue to consider in light of this is the management and as a predictor of fracture risk: a meta-analysis. Osteoporos Int 2005;16:13308.
[8] Ribot C, Tremollieres F, Pouilles J. The effect of obesity on post-menopausal bone loss
treatment of bone health in the obese group. When arbitrary BMI values and the risk of osteoporosis. Adv Nutr Res 1994;9:25771.
[9] Reid IR. Relationship between fat and bone. Osteoporos Int 2008;19:595606.
[10] Zhao L, Liu Y, Liu P, Hamilton J, Recker R, Deng H. Relationship of obesity with oste-
oporosis. J Clin Endocrinol Metab 2007;92:16406.
Table 5 [11] Cao JJ. Effects of obesity on bone metabolism. J Orthop Surg Res 2011;6:30.
BMI cut offs (25 kg/m2, 30 kg/m2 and 35 kg/m2) using DXA scans to exclude a diagnosis of [12] Migliaccio S, Greco EA, Fornari R, Donini LM, Lenzi A. Is obesity in women protective
osteoporosis. against osteoporosis? Diabetes Metab Syndr Obes 2011;4:27382.
[13] Census. First release July November 2012. Nottinghamshire County Council (2013).
BMI 25 kg/m2 BMI 30 kg/m2 BMI 35 kg/m2 http://www.nottinghamshire.gov.uk/living/business/economicdata/census2011/
Sensitivity 73.5% 34.5% 11.5% rstrelease/; 2011. [Accessed 1 April 2013].
Specicity 52.0% 85.3% 95.9% [14] Barrera G, Bunout D, Gatts V, Maza M, Leiva L, Hirsch S. A high body mass index
protects against femoral neck osteoporosis in healthy elderly subjects. Nutrition
PPV 80.5% 86.3% 88.3%
2004;20:76971.
NPV 42.1% 32.6% 28.7%
[15] Gnudi S, Sitta E, Lisi L. Relationship of body mass index with main limb fragility
PPV positive predictive value; NPV negative predictive value. fractures in postmenopausal women. J Bone Miner Metab 2009;27:47984.
210 T. Ong et al. / Bone 59 (2014) 207210
[16] Ensrud KE, Fullman RL, Barrett-Connor E, Cauley JA, Stefanick ML, Fink HA, et al. Vol-
untary weight reduction in older men increases hip bone loss: The Osteoporotic
Fractures in Men Study. J Clin Endocrinol Metab 2005;90:19982004.
[17] Siris ES, Chen Y, Abbott TA, Barrett-Connor E, Miller PD, Wehren LE, Berger ML. Bone
mineral density thresholds for pharmacological intervention to prevent fractures.
Arch Intern Med 2004;164:110812.
[18] Premaor MO, Pilbrow L, Tonkin C, Parker RA, Compston J. Obesity and fractures in
postmenopausal women. J Bone Miner Res 2010;25:2927.
[19] Compston JE, Watts ND, Chapurlat R. Obesity is not protective against fracture in
postmenopausal women: GLOW. Am J Med 2011;124:104350.
[20] Prieto-Alhambra D, Premaor MO, Avils FF, Hermosilla E, Martinez-Laguna D,
Carbonell-Abella C, et al. The association between fracture and obesity is site-
dependent: a population-based study in postmenopausal women. J Bone Miner
Res 2012;27:294300.
[21] Compston J. Obesity and bone. Curr Osteoporos Rep 2013;11:305.
[22] Nielson CM, Srikanth P, Orwoll ES. Obesity and fracture in men and women: an
epidemiologic perspective. J Bone Miner Res 2012;27:110.
[23] Nielson CM, Marshall LM, Adams AL, LeBlanc ES, Cawthon PM, Ensrud K, et al. BMI
and fracture risk in older men: The Osteoporotic Fractures in Men Study (MrOS). J
Bone Miner Res 2011;26:496502.
[24] Singh D, Park W, Levy MS, Jung ES. The effects of obesity and standing time on
postural sway during prolonged quiet standing. Ergonomics 2009;52:97786.
[25] Himes CL, Reynolds SL. Effect of obesity on falls, injury and disability. J Am Geriatr
Soc 2012;60:1249.
[26] Compston J. Bone quality: what is it and how is it measured. Arq Bras Endocrinol
Metabol 2006;50:57985.
[27] Seeman E, Delmas PD. Bone qualitythe material and structural basis of bone
strength and fragility. N Engl J Med 2006;354:225061.
[28] Premaor MO, Ensrud K, Lui L, Parker RA, Cauley J, Hillier TA, et al. Risk factors for
nonvertebral fracture in obese older women. J Clin Endocrinol Metab
2011;96:241421.
[29] Kanis JA, Johnell O, Oden A, Johansson H, McCloskey E. FRAX and the assess-
ment of fracture probability in men and women from the UK. Osteoporos Int
2008;19:38597.
[30] Premaor M, Parker RA, Cummings S, Ensrud K, Cauley JA, Lui L, et al. Predictive
value of FRAX for fracture in obese older women. J Bone Miner Res
2013;28:18895.
[31] Knapp K, Welsman J, Hopkins S, Fogelman I, Blake G. Obesity increases preci-
sion error in dual-energy X-ray absorptiometry measurements. J Clin Dent
2012;15:3159.