Fracture risk in patients with type 2 diabetes mellitus and possible risk factors: a systematic review and meta-analysis

Figures & data

Figure 1 Flowchart of the literature search.

Abbreviations: IGT, impaired glucose tolerance; ISI, Institute for Scientific Information; T1DM, type 1 diabetes mellitus.

Table 1 The characteristics of the studies entered into the meta-analysis

References	Country	Follow-up period (years)a	Age at enrollment (years)a	BMI (kg/m^2)a	T2DM subjects (n)	NDM subjects (n)	T2DM vs NDM (95% CI) results
Cohort design
Heath et al^Citation24 (1980)	USA	Unknown	<103	NR	T: 986	T: 986	Adjusted RRb Vertebral T: 0.6 (0.3–0.7) Proximal humerus T: 0.6 (0.3–0.9) Distal forearm T: 0.7 (0.5–0.9) Proximal forearm T: 0.8 (0.5–0.9)
Meyer et al^Citation25 (1993)	Norway	10.9	35–49	NR	M: 180 F: 118	M: 24,333 F: 23,850	Adjusted RRc Hip M: 9.4 (2.9–30.5), F: 9.2 (3.4–24.9)
Forsen et al^Citation26 (1999)	Norway	9	≥50	28.5	M: 220 F: 274	M: 13,210 F: 13,685	Adjusted RRd Hip M: 1.2 (0.4–3.2), F: 1.8 (1.1–2.9)
Ivers et al^Citation16 (2001)	Australia	5	≥49	NR	T: 216	T: 3,438	Adjusted RRe All fractures T: 0.9 (0.7–1.2) Hip T: 0.6 (0.2–2.2) Distal forearm T: 0.7 (0.2–2.3) Proximal humerus T: 0.7 (0.2–2.3) Ankle T: 1.1 (0.6–1.9)
Nicodemus and Folsom^Citation15 (2001)	USA	11	55–69	30.5	F: 1,682	F: 30,377	Adjusted RRf Hip F: 1.7 (1.2–2.4)
Ottenbacher et al^Citation27 (2002)	USA	7	≥65	NR	T: 690	T: 2,194	Adjusted HRg Hip T: 1.5 (1.0–2.3)
de Liefde et al^Citation28 (2005)	the Netherlands	6.3±2.3	≥55	26.8±4.1	T: 792 M: 309 F: 438	T: 5,863 M: 2,382 F: 3,481	Adjusted HRh Any nonvertebral T: 1.3 (1.0–1.8), M: 1.6 (0.9–2.9), F: 1.3 (0.9–1.8) Hip T: 1.3 (0.8–2.3), M: 1.3 (0.4–4.2), F: 1.4 (0.8–2.5) Wrist T: 1.4 (0.8–2.4), M: 1.3 (0.4–10.4), F: 1.5 (0.8–2.6)
Gerdhem et al^Citation29 (2005)	Sweden	4.6	75	NR	F: 74	F: 1,058	RR Any fracture F: 0.8 (0.5–1.3)
Strotmeyer et al^Citation30 (2005)	USA	4.5	70–79		T: 566	T: 2,236	Adjusted RRi Any fracture T: 1.64 (1.1–2.5)
Bonds et al^Citation31 (2006)	USA	7	64.9±7	NR	F: 5,285	F: 88,120	Adjusted RRj All fracture F: 1.2 (1.1–1.3) Hip F: 1.5 (1.2–1.8) Wrist F: 1.02 (0.85–1.2) Ankle F: 1.1 (0.95–1.9) Foot F: 1.3 (1.07–1.6) Vertebral F: 1.3 (1.0–1.6)
Holmberg et al^Citation32 (2006)	Sweden	17	27–68	NR	M: 276 F: 166	M: 22,444 F: 10,902	Adjusted RRk Any fracture M: 2.4 (1.7–3.4), F: 1.9 (1.3–2.8) Forearm M: 2.2 (0.95–4.8), F: 1.5 (0.7–3.0) Proximal humerus M: 1.1 (0.2–7.9), F: 2.1 (0.8–6.0) Hip M: 6.4 (3.4–11.8), F: 4.0 (1.7–9.4) Ankle M: 1.4 (0.5–4.5), F: 2.4 (1.1–5.4) Vertebral M: 1.1 (0.4–3.5), F: 2.9 (1.3–6.3)
Dobnig et al^Citation33 (2006)	Australia	2	≥70	26.4±4.8	F: 583	F: 1,081	Adjusted HRl Hip F: 1.5 (1.3–1.8)
Ahmed et al^Citation34 (2006)	Norway	6	25–98	28.7	M: 175 F: 198	M: 12,639 F: 14,065	Adjusted RRm Any nonvertebral M: 1.2 (0.6–2.5), F: 1.1 (0.7–1.7) Hip M: 1.6 (0.6–4.5), F: 1.9 (1.0–3.5)
Janghorbani et al^Citation35 (2006)	USA	20.4±3.4	34–59	31±6.4	F: 8,345	F: 101,343	Adjusted RRn Hip F: 2.2 (1.8–2.7)
Leslie et al^Citation36 (2007)	Canada	10	≥20	NR	T: 82,094; newly diagnosed: 42,874; short duration: 16,081; long duration: 23,139	T: 236,682	Adjusted RRo All fractures Newly diagnosed: 0.91 (0.86–0.95) Short term: 1.00 (0.93–1.07) Long term: 1.15 (1.09–1.22) Hip Newly diagnosed: 0.83 (0.75–0.92) Short term: 1.13 (1.00–1.28) Long term: 1.40 (1.28–1.53)
Lipscombe et al^Citation37 (2007)	Canada	6.1	≥66		M: 100,322 F: 97,090	M: 202,875 F: 198,525	Adjusted HRp Hip M: 1.2 (1.1–1.3), F: 1.1 (1.1–1.2)
Melton et al^Citation38 (2008)	USA	11.8	30–97		T: 1,964 M: 992 F: 972	NR	SIR All fractures T: 1.3 (1.2–1.4), M: 1.4 (1.3–1.6), F: 1.3 (1.2–1.6) Hip T: 1.1 (0.9–1.4) Wrist T: 1.2 (0.9–1.5) Vertebral T: 2.8 (2.4–3.2)
Chen et al^Citation39 (2008)	Taiwan	6	≥35		M: 227,289 F: 257,498	M: 227,303 F: 257,303	Adjusted HRq Hip M: 1.3 (1.2–1.3), F: 1.7 (1.66–1.8)
Mancini et al^Citation40 (2009)	Italy	13	44–82	31	M: 43	M: 22	Adjusted ORr Vertebral M: 6.5 (1.3–38.1)
Yamamoto et al^Citation41 (2009)	Japan	NR	≥50	23.7±3.5	M: 161 F: 137	M: 76 F: 622	OR Vertebral M: 4.73 (2.19–10.2), F: 1.9 (1.1–3.1)
Schneider et al^Citation42 (2013)	USA	20	45–64	31±6.1	T: 1,195 M: 501 F: 694	T: 13,340 M: 5,936 F: 7,404	Adjusted HRs All fractures T: 1.74 (1.4–2.2), M: 1.9 (1.5–2.5), F: 1.5 (1.1–2.2)
Chung et al^Citation43 (2013)	Korea	NR	≥50	25.1±3.5	F: 2,239	NR	Adjusted ORt Hip F: 1.03 (1.02–1.04) Vertebral F: 1.03 (1.02–1.04)
Oei et al^Citation44 (2013)	the Netherlands	12.2	≥55	26.9±4	T: 217	T: 3,715	Adjusted HRu All fractures T: 1.5 (1.1–1.9) Hip T: 1.2 (0.6–2.1) Wrist T: 1.7 (1.03–2.9)
Reyes et al^Citation45 (2014)	Spain	2.55	≥65		M: 36,865	M: 149,306	Adjusted RRv Hip M: 1.9 (1.2–3.1)
Napoli et al^Citation46 (2014)	USA	9.1	≥65	29.5±4.4	M: 801	M: 3,086	Adjusted HRw Any nonvertebral M: 1.3 (1.1–1.5)
Leslie et al^Citation47 (2014)	Canada	6	≥40	30.2±6.2	T: 6,455	T: 55,958	Adjusted HRx All fractures T: 1.3 (1.2–1.5)
Hothersall et al^Citation48 (2014)	Scotland	2	20–84		T: 180,841	T: 3,066,000	Adjusted IRRy Hip M: 0.97 (0.92–1.02), F: 1.05 (1.01–1.10)
Looker et al^Citation49 (2016)	USA	6.7	≥65	29.5±0.41	NHW: 398 NHB: 178 MA: 226	NHW: 2,554 NHB: 631 MA: 601	Adjusted HR^z Non-skull fracture NHW: 1.2 (0.93–1.6) NHB: 1.9 (1.02–3.40) MA: 2.4 (1.5–3.8)
Case–control
Kegan et al^Citation50 (2002)	USA	NR	≥45	NR	T: 1,913	T: 1,913	Adjusted ORaa Forearm T: 0.9 (0.7–1.2) Foot T: 1.4 (1.1–1.8) Proximal humerus T: 1.7 (1.2–2.3)
Vestergaard et al^Citation51 (2005)	Denmark	NR	43±27	NR	T: 3,241	T: 6,375	Adjusted ORbb Any fracture T: 1.2 (1.1–1.3) Hip T: 1.4 (1.2–1.6) Forearm T: 1.2 (1.01–1.5) Spine T: 1.3 (0.97–1.86)

Notes:

a Data shown as mean, range, or mean ± SD.

b Adjusted for age, sex, race, residence, year, and institution.

c Adjusted for age, height, BMI, physical activity, stroke, receipt of a disability pension, marriage, and smoking.

d Adjusted for age, BMI, and daily smoking.

e Adjusted for age, sex, and BMI.

f Adjusted for age, BMI, smoking, estrogen use, and waist:hip ratio.

g Adjusted for age, BMI, smoking, and previous stroke.

h Adjusted for age, sex, BMI, smoking, serum creatinine, visual acuity, falling frequency, lower limb disability.

i Adjusted for age, sex, race, site, hip BMD, lean mass, fat mass, and abdominal visceral fat.

j Adjusted for age; ethnicity; weight; height; time-dependent history of falls; previous fracture; history of osteoporosis; trouble seeing at baseline; alcohol or tobacco use; calcium and vitamin D intake; exercise; bisphosphonate, estrogen, steroid, insulin, SERM, or thyroid hormone use.

k Adjusted for age, BMI, and smoking

l Adjusted for age, weight, and calcaneal bone mass.

m Adjusted for age, BMI, smoking, and metabolic features (mean blood pressure, HDL, and triglycerides).

n Adjusted for age, BMI, physical activity, menopausal status, estrogen use, smoking, daily intake of calcium, vitamin D, and protein.

o Adjusted for age, sex, income quintile, area of residence, and ethnicity.

p Adjusted for age-group; chronic unstable disease; prior stroke; visual impairment; neuropathy; amputation; treatment with nitrates, statins, anticonvulsants, inhaled corticosteroids, and medications that increase the risk of fall; history of BMD test; and income quintile.

q Adjusted for age as a continuous variable, geographic area, and urbanization status.

r Adjusted for age and BMI, rosiglitazone plus metformin treatment.

s Adjusted for age, sex, race/study center, BMI, sports activity tertile, alcohol consumption, cigarette smoking, glucocorticoid or antidepressant use, and thiazide diuretic use.

T Adjusted for age, BMI, duration of diabetes, DM complications, HbA1c, serum creatinine, DM medications, risk factors for osteoporosis (alcohol and smoking), and history of fragility fracture.

u Adjusted for age, sex, height, weight, and femoral neck BMD.

v Adjusted for age, BMI, smoking, alcohol consumption, use of oral corticosteroids, and comorbid conditions.

w Adjusted for age, race, clinic site, total hip BMD.

x Adjusted for WHO Fracture Risk Assessment (FRAX™) tool including femoral neck BMD.

y Adjusted for age, sex, survey, BMI, self-rated physical activity compared to others, hospital visits in past year, and smoking.

aa Adjusted for age, sex, and race/ethnicity.

bb Adjusted for the variables in the table plus prior fracture, corticosteroid use, use of antiepileptic drugs, use of diuretics (loop, thiazide, potassium-sparing, other types), use of anxiolytics and sedatives, use of neuroleptics, use of antidepressants, alcoholism, use of statins and non-statin cholesterol-lowering drugs, use of antihypertensives, myocardial infarction, stroke, number of bed days in 1999, number of contacts to GP or specialists in 1999, working or not, income, living with another person vs living alone.

Abbreviations: BMD, bone mineral density; BMI, body mass index; CI, confidence interval; DM, diabetes mellitus; F, female; GP, general practitioner; HbA1c, glycated hemoglobin; HR, hazard ratio; IRR, incidence risk ratio; M, male; MA, Mexican American; NDM, nondiabetic subjects; NHB, non-Hispanic Black; NHW, non-Hispanic white; NR, not reported; OR, odds ratio; RR, relative risk; SERM, selective estrogen-receptor modulator; SIR, standardized incidence ratio; T2DM, type 2 diabetes mellitus; T, total; WHO, World Health Organization.

Figure 2 The results of meta-analysis of the association between type 2 diabetes mellitus and risk of hip fracture.

Notes: Each square shows the study specific relative risk estimate. Square sizes are proportional to the weight assigned to the study in the meta-analysis and the horizontal line shows the related 95% CI. The diamond shows the summary relative risk estimate and its width represents the corresponding 95% CI. All statistical tests were two sided. Statistical heterogeneity between studies was assessed with Cochran’s Q test. Meta-analysis was performed by using a random effects method.
Abbreviations: CI, confidence interval; ES, effect size.

Table 2 Summary relative risk estimates from case–control and cohort studies of the association between type 2 diabetes mellitus and fractures incidence using meta-analysis methods

Type of fractures	Number of studies	Summary relative risk	95% confidence interval	Between studies
				I^2 (%)	P for heterogeneity
Hip fractures	15	1.20	1.17–1.23	85.5	0.000
Wrist fractures	10	0.98	0.88–1.07	61.3	0.006
Vertebral fractures	9	1.16	1.05–1.28	95.9	0.000
Proximal humerus	5	1.09	0.86–1.31	84.0	0.000
Ankle fractures	3	1.13	0.95–1.32	0.0	0.762
Foot fractures	3	1.37	1.21–1.54	0.0	0.90
All fractures, total	27	1.17	1.15–1.20	85.5	0.000

Figure 3 The results of meta-analysis of the association between type 2 diabetes mellitus and risk of wrist fracture.

Notes: Each square shows the study specific relative risk estimate. Square sizes are proportional to the weight assigned to the study in the meta-analysis and the horizontal line shows the related 95% CI. The diamond shows the summary relative risk estimate and its width represents the corresponding 95% CI. All statistical tests were two sided. Statistical heterogeneity between studies was assessed with Cochran’s Q test. Meta-analysis was performed by using a random effects method.
Abbreviations: CI, confidence interval; ES, effect size.

Figure 4 The results of meta-analysis of association between type 2 diabetes mellitus and risk of overall fractures.

Notes: Each square shows the study specific relative risk estimate. Square sizes are proportional to the weight assigned to the study in the meta-analysis and the horizontal line shows the related 95% CI. The diamond shows the summary relative risk estimate and its width represents the corresponding 95% CI. All statistical tests were two sided. Statistical heterogeneity between studies was assessed with Cochran’s Q test. Meta-analysis was performed by using a random effects method.
Abbreviations: CI, confidence interval; ES, effect size.

Table 3 Risk factors for the association between type 2 diabetes and fracture risk

Subgroup	Studies (n)	Summary relative risk	95% confidence interval	Between studies		Between subgroups
				I^2 (%)	P for heterogeneity	I^2 (%)	P for heterogeneity
Age (years)
50–59	5	1.17	1.15–1.21	94.7	0.000	85.5	<0.001
60–69	12	1.20	1.10–1.30	81.2	0.000
≥70	12	1.30	1.21–1.40	45.7	0.005
Sex
Female	7	1.44	1.18–1.70	91.3	0.000	91.3	0.043
Male	3	1.90	1.3–2.58	0.00	0.000
BMI (kg/m^2)
<30	11	1.44	1.24–1.65	29.5	0.193	92.6	<0.001
≥30	11	1.30	1.22–1.37	72.5	0.000
Geographic area
Europe	11	1.10	1.03–1.13	69.8	0.000	90.6	0.29
North America	17	1.18	1.15–1.20	85.5	0.000
Asia	3	1.24	1.14–1.40	90.6	0.000
Australia	2	1.18	1.0–1.36	91.7	0.000
Follow-up period, years
<10	21	1.19	1.16–122	83.2	0.000	89.9	0.004
≥10	8	1.13	1.08–1.18	89.9	0.000
Duration of diabetes, years
<10	6	1.00	0.93–1.06	78.2	0.003	93.2	<0.001
≥10	6	1.19	1.13–1.25	93.2	0.000
Physical activity	3	0.75	0.65–0.85	92.4	0.000	NR	NR
Smoking status	3	1.29	0.92–1.88	92.0	0.000	NR	NR
Users of systemic corticosteroids	3	1.51	1.29–1.72	38.6	0.196	NR	NR
Insulin therapy	11	1.52	1.42–1.61	4.8	0.393	NR	NR
Treated with thiazolidinediones	3	0.75	0.60–0.91	0.0	0.513	NR	NR

Abbreviations: BMI, body mass index; NR, not reported.

Figure 5 Begg’s funnel plot for publication bias in the risk difference analysis.

Abbreviation: RR, relative risk.

Figure 6 The meta-regression analysis of the relationship between the risk of factors in diabetic patients and the year of study.

Abbreviations: CI, confidence interval; Coef, coefficient of variation; REML, restricted (or residual) maximum likelihood estimation; RR, relative risk.

HeathH3rdMeltonLJ3rdChuCPDiabetes mellitus and risk of skeletal fractureN Engl J Med1980303105675707402221

 PubMed Web of Science ®Google Scholar

MeyerHETverdalAFalchJARisk factors for hip fracture in middle-aged Norwegian women and menAm J Epidemiol199313711120312118322761

 PubMed Web of Science ®Google Scholar

ForsenLMeyerHEMidthjellKEdnaTHDiabetes mellitus and the incidence of hip fracture: results from the Nord-Trøndelag Health SurveyDiabetologia199942892092510491750

 PubMed Web of Science ®Google Scholar

IversRQCummingRGMitchellPPedutoAJBlue Mountains Eye StudyDiabetes and risk of fracture: the Blue Mountains Eye StudyDiabetes Care20012471198120311423502

 PubMed Web of Science ®Google Scholar

NicodemusKKFolsomARIowa Women’s Health StudyType 1 and type 2 diabetes and incident hip fractures in postmenopausal womenDiabetes Care20012471192119711423501

 PubMed Web of Science ®Google Scholar

OttenbacherKJOstirGVPeekMKGoodwinJSMarkidesKSDiabetes mellitus as a risk factor for hip fracture in Mexican-American older adultsJ Gerontol A Biol Sci Med Sci20025710M648M65312242318

 PubMed Web of Science ®Google Scholar

de LiefdeIIvan der KliftMde LaetCEvan DaelePLHofmanAPolsHABone mineral density and fracture risk in type-2 diabetes mellitus: the Rotterdam StudyOsteoporos Int200516121713172015940395

 PubMed Web of Science ®Google Scholar

GerdhemPIsakssonAAkessonKObrantKJIncreased bone density and decreased bone turnover, but no evident alteration of fracture susceptibility in elderly women with diabetes mellitusOsteoporos Int200516121506151215824889

 PubMed Web of Science ®Google Scholar

StrotmeyerESCauleyJASchwartzAVNevittMCResnickHEBauerDCNontraumatic fracture risk with diabetes mellitus and impaired fasting glucose in older white and black adults: the health, aging, and body composition studyArch Intern Med2005165141612161716043679

 PubMedGoogle Scholar

BondsDELarsonJCSchwartzAVRisk of fracture in women with type 2 diabetes: the Women’s Health Initiative Observational StudyJ Clin Endocrinol Metab20069193404341016804043

 PubMed Web of Science ®Google Scholar

HolmbergAHJohnellONilssonPMNilssonJBerglundGAkessonKRisk factors for fragility fracture in middle age. A prospective population-based study of 33,000 men and womenOsteoporos Int20061771065107716758143

 PubMed Web of Science ®Google Scholar

DobnigHPiswanger-SolknerJCRothMType 2 diabetes mellitus in nursing home patients: effects on bone turnover, bone mass, and fracture riskJ Clin Endocrinol Metab20069193355336316735485

 PubMed Web of Science ®Google Scholar

AhmedLAJoakimsenRMBerntsenGKFonneboVSchirmerHDiabetes mellitus and the risk of non-vertebral fractures: the Tromsø StudyOsteoporos Int200617449550016283065

 PubMed Web of Science ®Google Scholar

JanghorbaniMFeskanichDWillettWCHuFProspective study of diabetes mellitus and risk of hip fracture: the Nurses’ Health StudyDiabetes Care20062971573157816801581

 PubMed Web of Science ®Google Scholar

LeslieWLixLPriorHDerksenShMetgeCO’NeilJBiphasic fracture risk in diabetes: a population-based studyBone20074061595160117392047

 PubMed Web of Science ®Google Scholar

LipscombeLJamalSBoothGKawkerGThe risk of hip fractures in older individuals with diabetesDiabetes Care200730483584117392544

 PubMed Web of Science ®Google Scholar

MeltonLJLeibsonCLAchenbachSJTherneauTMKhoslaSFracture risk in type 2 diabetes: update of a population-based studyJ Bone Miner Res20082381334134218348689

 PubMed Web of Science ®Google Scholar

ChenHFHoCALiCYIncreased risks of hip fracture in diabetic patients of TaiwanDiabetes Care2008311758017947342

 PubMed Web of Science ®Google Scholar

ManciniTMazziottiGDogaMVertebral fractures in males with type 2 diabetes treated with rosiglitazoneBone200945478478819527806

 PubMed Web of Science ®Google Scholar

YamamotoMYamaguchiTYamauchiMKajiHSugimotoTDiabetic patients have an increased risk of vertebral fractures independent of BMD or diabetic complicationsJ Bone Miner Res200924470270919049338

 PubMed Web of Science ®Google Scholar

SchneiderAWilliamsEBrancatiFBleckerSCoreshJSelvinEDiabetes and risk of fracture-related hospitalization: the atherosclerosis risk in communities studyDiabetes Care20133651153115823248194

 PubMed Web of Science ®Google Scholar

ChungDJChoiHJChungYSLimSKYangSOShinCSThe prevalence and risk factors of vertebral fractures in Korean patients with type 2 diabetesJ Bone Miner Metab201331216116823076296

 PubMed Web of Science ®Google Scholar

OeiLZillikensMDehghanAHigh bone mineral density and fracture risk in type 2 diabetes as skeletal complications of inadequate glucose control: the Rotterdam StudyDiabetes Care20133661619162823315602

 PubMed Web of Science ®Google Scholar

ReyesCEstradaPNoguésXThe impact of common co-morbidities (as measured using the Charlson index) on hip fracture risk in elderly men: a population-based cohort studyOsteoporos Int20142561751175824676845

 PubMed Web of Science ®Google Scholar

NapoliNStrotmeyerESEnsrudKEFracture risk in diabetic elderly men: the MrOS studyDiabetologia201457102057206524908567

 PubMed Web of Science ®Google Scholar

LeslieWDMorinSNLixLMMajumdarSRDoes diabetes modify the effect of FRAX risk factors for predicting major osteoporotic and hip fracture?Osteoporos Int201425122817282425092059

 PubMed Web of Science ®Google Scholar

HothersallEJLivingstoneSJLookerHCContemporary risk of hip fracture in type 1 and type 2 diabetes: a national registry study from ScotlandJ Bone Miner Res20142951054106024155126

 PubMed Web of Science ®Google Scholar

LookerACEberhardtMSSaydahSHDiabetes and fracture risk in older U.S. adultsBone20168291525576672

 PubMed Web of Science ®Google Scholar

KeganTHKelseyJLSidneySQuesenberryCPFoot problems as risk factors of fracturesAm J Epidemiol20021551092693111994232

 PubMed Web of Science ®Google Scholar

VestergaardPRejnmarkLMosekildeLRelative fracture risk in patients with diabetes mellitus, and the impact of insulin and oral antidiabetic medication on relative fracture riskDiabetologia20054871292129915909154

 PubMed Web of Science ®Google Scholar