12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 1/8

Home > Volume 1, Issue 1 > Article

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Find out morehere ×

An open access, peer reviewed, online­only journal dedicated to publishing research on rheumatic and musculoskeletal diseases

+

RMD Openrmdopen.bmj.comRMD Open 2015;1:e000077 doi:10.1136/rmdopen­2015­000077

Osteoarthritis

Extended report

Association between diabetes mellitus and osteoarthritis: systematic literature reviewand meta­analysis

Karine Louati1,2,3, Céline Vidal1,2, Francis Berenbaum1,2,3,4 and Jérémie Sellam1,2,3,4

Author Af f iliations

Correspondence to

Prof essor Francis Berenbaum; f rancis.berenbaum@sat.aphp.f r

AbstractObjectives To inv estigate the prev alence of osteoarthritis (OA) in patients with diabetes mellitus (DM) and prev alence of DM in patients with OA andwhether OA and DM are associated.

Design A sy stematic literature rev iew and meta­analy sis. We included cohort, case–control and cross­sectional studies assessing the number ofpatients with DM and/or OA. The mean prev alence of OA among patients with DM and DM among patients with OA was calculated. Data f rom trialsassessing an association of diabetes and OA were pooled and results are presented as unadjusted OR and 95% CI.

Results From the 299 publications, we included 49 studies in the analy sis, including 28 cross­sectional studies, 11 cohort studies and 10 case–controlstudies. In all, 21, 5 and 23 articles inv olv ed patients with OA exclusiv ely , patients with DM and the general population, respectiv ely . For 5788 patientswith DM, the mean OA prev alence was 29.5±1.2%. For 645 089 patients with OA, the prev alence of DM was 14.4±0.1%. The risk of OA was greater in theDM than non­DM population (OR=1.46 (1.08 to 1.96), p=0.01), as was DM in the OA than non­OA population (OR=1.41 (1.21 to 1.65), p<0.00 001). Amongthe 12 studies reporting an OR adjusted on at least the body mass index, 5 showed no association of DM and OA and 7 identif ied DM as an independentrisk f actor.

Conclusions This meta­analy sis highlights a high f requency of OA in patients with DM and an association between both diseases, representing a f urtherstep towards the indiv idualisation of DM­related OA within a metabolic OA phenoty pe.

Key messages

What is already known on this subject?

Metabolic sy ndrome and osteoarthritis hav e been f ound to be associated in some studies, delineating the metabolic osteoarthritis phenoty pe.Association between diabetes mellitus and osteoarthritis in epidemiological studies hav e giv en conf licting results.

What does this study add?

This is the f irst meta­analy sis showing an association between osteoarthritis and diabetes mellitus.

How might this impact on clinical practice?

Treating diabetes mellitus may be ef f ectiv e in patients with osteoarthritis.

12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 2/8

Prev ention initiativ es of osteoarthritis may be specif ically proposed to patients with diabetes mellitus.

IntroductionOsteoarthritis (OA) is the most f requent and disabling joint disease in adults. Besides its sev eral localisations, a recent hy pothesis has suggested a newclassif ication f or phenoty ping OA that includes ageing, metabolic sy ndrome (MetS) and post­traumatic ev ents and genetic­related OA. , Despite someshared pathophy siological mechanisms among these phenoty pes, each may display specif ic pathway s.

In MetS­associated OA, the mechanical impact of ov erweight or obesity on joints may easily explain knee OA. Howev er, within this phenoty pe, consideringthe epidemiological association of ov erweight or obesity and hand OA, some sy stemic f actors may participate in the pathogenic process; f or example,adipose tissue products, or ‘adipokines’, may hav e a sy stemic impact at a distance on joints. Bey ond obesity ­related OA, MetS and OA hav e beenf ound to be associated in some epidemiological studies, which suggests that the other components of MetS, such as diabetes mellitus (DM), high bloodpressure or dy slipidaemia may together or independently participate in the OA pathophy siology . , Along this line, DM and hy pergly caemia seemed to beassociated with OA in some epidemiological studies. Moreov er, the link between the two diseases may be supported by the deleterious role of glucoseexcess through the accumulation of adv anced gly cation end products (AGEs), oxidativ e stress and promotion of sy stemic inf lammation. Thissituation is well illustrated by spontaneous cartilage disruption in the rat model of streptozotocin­induced diabetes. Howev er, other publications hav equestioned the link between DM and OA. ,

To f urther address the association of OA and DM, we perf ormed a sy stematic rev iew of the literature and a meta­analy sis to inv estigate the prev alence ofOA among patients with DM and that of DM among patients with OA and to determine whether DM and OA are associated.

Methods

Systematic literature search and selection of relevant studiesWe perf ormed a sy stematic rev iew of the literature according to the Cochrane guidelines (http://handbook.cochrane.org/, 24 February 2014, date lastaccessed). Relev ant publications were selected f rom three databases (PubMed, EMBASE and the Cochrane Library ) without any limitation on time (up toJune 2013 and updated in January 2015). We also searched f or articles in the ref erences of selected publications and the main congresses ofrheumatology f or OA (American College of Rheumatology (ACR), European League Against Rheumatism (EULAR) and Osteoarthritis Research SocietyInternational (OARSI)) and congresses of endocrinology (Endocrine society 's annual meeting, European Congress of Endocrinology , American DiabetesAssociation and European Association f or the Study of Diabetes) f rom 2012 to 2014. We used the f ollowing key words f or the PubMed search: (“diabetesmellitus, ty pe 2”[MeSH] OR “diabetes mellitus, ty pe 1”[MeSH] OR “diabetes complications”[MeSH] OR “metabolic sy ndrome X”[MeSH] OR (“bloodglucose”[MeSH] OR “blood glucose”[All Fields])) AND “osteoarthritis”[MeSH] AND (“humans”[MeSH] AND (English[lang] OR French[lang])).

We included observ ational studies (ie, cohort, case–control and cross­sectional studies) assessing the number of patients with OA and/or DM, or theincidence or prev alence of OA in patients with DM or DM in patients with OA, or an association between OA and DM. We excluded articles of therapeuticstudies, rev iews and case reports as well as letters; studies in which all patients had OA and DM or in which the link between OA and DM was not reported;and studies without an av ailable number of patients with each disease. Selection of articles was based on titles and abstracts than on the f ull text. Oneauthor (KL) has managed this selection.

Data extractionTwo authors (KL and CV) extracted the f ollowing data: study design and population (observ ational study , quality score, def inition of DM and OA andlocalisation of OA); exposure gly caemia (f asting blood glucose (FBG) lev el, mmol/L) or gly cosy lated haemoglobin (HbA1c; %) or number of patients withDM; outcome (number of patients with OA); body mass index (BMI, kg/m2) or number of patients with obesity as potential conf ounders; associationmeasure (relativ e risk or OR or only conclusion on an association). Then we conv ersely considered OA as an exposure f actor and DM as an outcome. Thestudy quality was assessed by the Strengthening the Reporting of Observ ational Studies in Epidemiology (STROBE) scale and results are reported as apercentage of 19 pertinent items of the 22 total items.

Statistical analysisWe perf ormed two analy ses with av ailable results f rom trials. First, we perf ormed a descriptiv e analy sis: f or cross­sectional, case–control or cohortstudies, we used the number of patients with OA or DM and total number in each population to calculate the prev alence of OA among patients with DM andthat of DM among patients with OA. To estimate this prev alence f rom cohort longitudinal prospectiv e studies, we used baseline data. Prev alence isexpressed as mean±SD. Second, we perf ormed a comparativ e analy sis using studies assessing an association between diagnosis of DM and OA in cross­sectional studies and cohort studies of the general population and case–control studies of OA or DM populations. We calculated the odds of hav ing OAamong patients with DM and of DM among patients with OA with ORs and 95% CIs. Then we used Revman V.5.2 to perf orm a meta­analy sis with a f ixed­ef f ects model. A random­ef f ects model was used with high heterogeneity among studies (>50%), ev aluated by I2, and a sensitiv ity analy sis wasperf ormed by remov ing studies with aberrant results and combining studies with the same characteristics. OR>1 and p≤0.05 was considered an increasedrisk of OA among patients with DM and/or DM among patients with OA.

Results

Literature search and characteristics of included trialsThe selection of articles is in f igure 1: f rom 299 publications, we included 49 in the analy sis. We f ound no publication bias (see online supplementary f igureS1). The articles represented 28 cross­sectional, 11 cohort and 10 case–control studies. In total, 21, 5 and 23 inv olv ed exclusiv ely patients with OA,exclusiv ely patients with DM and the general population, respectiv ely (table 1).

View this table: In this window In a new window

Table 1Description of the 49 studies selected f or analy sis

1 2

3–5

6 7

8–11

12–15

12–15

16 17

18

12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 3/8

View larger version: In a new window Download as PowerPoint Slide

Figure 1Flow chart of selection of articles. OA, osteoarthritis; DM, diabetes mellitus; RA, rheumatoid arthritis; MetS, metabolic sy ndrome; CV, cardiov ascular;ACR, American College of Rheumatology ; EULAR, The European League Against Rheumatism; OARSI, Osteoarthritis Research Society International.

The criteria of inclusion f or the general population were v ariable f or age: f rom 20 to 86 y ears f or the Iwaki Health Promotion Project, and f rom 65 to 84 y ears f or the ILSA study ; the patients had radiographs of hips f or the study of Ty ppo, or data were extracted f rom public serv ice data f or the studyof Nav arro et al. In most cases, OA was def ined by radiological and clinical criteria, usually based on the ACR criteria. DM was def ined by elev atedgly caemia, HbA1c proportion or prescription of DM treatment (table 2).

View this table: In this window In a new window

Table 2Characteristics of the 49 included studies

Only two studies specif ied the number of patients with ty pe 1 and ty pe 2 DM: , in the study of Niev es­Plaza et al, there were only 14 patients withty pe 1 DM with a similar repartition in the OA and non­OA groups, and in the study of Ray et al there were 11% patients with ty pe 1 DM with a similarrepartition in the OA and non­OA groups. The median STROBE quality score was 69% (range 33–91%; table 2). For 6 studies (4 case–control and 2 cross­sectional studies), the score was <50%, and f or 30 studies it was >60%. The country of origin of the studies was div erse (22 studies with patients f romEurope and 16 studies with patients f rom North America). For f our studies, OA was sev ere because the outcome was arthroplasty . Among the 49 studies,34 assessed the association between OA and DM, 28 the f requency of DM among patients with OA and 24 the f requency of OA among patients with DM.

Characteristics of patientsA total of 1 192 518 patients were included in the analy ses. The mean age ranged f rom 43.8±43.9 to 76.9±5.4 y ears. , The mean proportion of f emaleswas 78.92% (f rom 9.3% to 100%). , , The localisation was the knee f or 31 studies (knee only f or 13 studies), the hip f or 15 studies (hip only f or 3studies), hands f or 12 studies (hands only f or 4 studies) and the spine f or 5 studies (see online supplementary table S1). The mean FBG lev el ranged f rom3.95 (no SD av ailable) to 12.17±6.49 mmol/L and HbA1c f rom 5.1±0.1% to 7.2% (no SD av ailable). , , , The prev alence of obesity v aried greatlyf rom 9.1% to 73.2%, and BMI ranged f rom 22.3±2.7 to 33.8±5.8 g/cm2. , , , , MetS was reported in f iv e studies using dif f erent def initions inwhich hy pergly caemia was one of the items and inv olv ed 5.1–58.6% of patients. , , , , ,

Prevalence of OA among patients with DM and DM among patients with OAFor 5788 patients with DM, the mean OA prev alence was 29.5±1.2% (mean age=61.01 y ears). This prev alence was calculated by using the 5 and 12 studiesof patients with DM and the general population, respectiv ely , with av ailable data on the number of patients with OA in the DM population (see onlinesupplementary table S2). In this population, the prev alence of OA calculated with av ailable data f or each localisation was 17.2±2.0% f or the knee, , ,, , , 12.3±1.3% f or the hip , and 38.4±6.8% f or the hand. ,

For 645 089 patients with OA, the DM prev alence was 14.4±0.1%. It was calculated by using the 19 and 12 studies of patients with OA and the generalpopulation, respectiv ely , with av ailable data on the number of patients with DM in the OA population (see online supplementary table S2). Three studiesinv olv ing patients with OA were not included because they assessed semiquantitativ e or continuous v ariables, the Kellgren­Lawrence (KL) score andgly caemia but not OA or DM diagnosis. , ,

Associations between OA and DMIn total, 34 studies assessed the association of OA and DM and/or gly caemia or HbA1c proportion; 21 showed a signif icant association in their conclusionsor at least reported OR>1 in the text, , , , , , , , , , , , , , , , , , , whereas 12 studies display ed noassociation. , , , , , , , ,

Risk of OA in DM: meta­analysis and sensitivity analysesFor risk of OA in a DM v ersus non­DM population, among 32 137 patients, the ov erall OR was 1.46 (1.08 to 1.96), with high heterogeneity (I2=88%; f igure2). Af ter excluding poor­quality studies (ie, STROBE score <50%), the heterogeneity did not change (I2=88%). , Considering only studies with v alidatedcriteria f or diabetes including gly caemia or HbA1c and excluding studies with declarativ e data only , the OR was 1.58 (1.14 to 2.20), with similarheterogeneity (I2=89%). Considering all studies with the same design (ie, cross­sectional, cohort or case–control studies), the OR was signif icant f or onlycase–control studies (2.85 (1.71 to 4.73); I2=0%).

20

46 31

34

19 57 57

19

7 46

29 35 63

19 20 63 64

7 20 34 35 36

7 20 34 36 42 48

10 19 20

34 42 48 31 48 36 50

56 60 63

7 8 10 20 25 28 29 39 41 44 46 50–52 54 55 57 58 60 61 63

16 17 27 31 34 38 42 48 49

55 31

12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 4/8

View larger version: In a new window Download as PowerPoint Slide

Figure 2Forest plot f or osteoarthritis among patients with and without diabetes mellitus.

Among OA risk f actors, age and obesity hav e a strong impact on OA dev elopment. Considering only studies with patients ≥50 y ears old, the OR was 1.32(1.13 to 1.53), without heterogeneity , which suggested that the association remained ev en in patients at increased risk of OA because of their age. ,, , ,

Among the 12 studies with OR adjusted on BMI, 5 showed no association between DM and OA, , , , , but 7 identif ied DM as an independent riskf actor of OA, , , , , , , and were of higher quality as illustrated by the mean STROBE scale (69±7.4% v s 76±8.5%, respectiv ely ).Interestingly , the positiv e studies were the recent ones: six of sev en were published between 2009 and 2013. , , , , , Considering OAlocalisations: the results were signif icant f or knee OA­only and hand OA­only (OR=1.64 (1.17 to 2.29) with 9170 patients , , , , , andOR=1.31(1.07 to 1.61) with 5879 patients, , , respectiv ely ) but not f or hip (OR=0.82 (0.56 to 1.21), with 5682 patients). ,

Risk of DM in OA: meta­analysis and sensitivity analysesFor risk of DM in an OA v ersus non­OA population, among 1 040 175 patients, the ov erall OR was 1.41 (1.21 to 1.65), assessed by a random­ef f ects modelbecause of I2=95% (f igure 3).

View larger version: In a new window Download as PowerPoint Slide

Figure 3Forest Plot f or DM among patients with and without OA (OA, osteoarthritis; DM, diabetes mellitus).

We perf ormed f our sensitiv ity analy ses to strengthen the results. First, with the heterogeneity explained by two studies with aberrant results, we remov edthese two studies. , The OR remained similar: 1.42 (1.22 to 1.66), I2=96%. Second, we f ocused on sev ere OA (ie, the 3 studies with surgery as an OAoutcome corresponding to 11 805 patients); the OR was not signif icant: 1.32 (0.52 to 3.36) but with high heterogeneity (I2=85%). , , Third, weremoved the studies that did not use internationally recognised diagnosis criteria f or OA such as ACR criteria or the KL score f or OA def inition; the OR was1.32 (1.13 to 1.53) without any heterogeneity (I2=0%) with f iv e studies and 9947 patients. , , , , Fourth, we considered OA localisations: f or datainv olv ing knee OA only or hip OA only , the results were signif icant f or the knee (OR=1.51 (1.09 to 2.09) with 5 studies and 9102 patients) but not f or the hip(OR=0.71 (0.49 to 1.04) with 3 studies and 6240 patients). , , , , , , We also f ound a signif icant association f or non­weight­bearing hand OA(OR=1.31 (1.07 to 1.61)). There was no study that included only generalised OA.

DiscussionOA is a heterogeneous disorder that can be separated in an age­related, metabolic and post­traumatic OA, representing thus the three main phenoty pes ofthe disease. Metabolic OA is wider than obesity ­related OA since MetS and OA are epidemiologically linked. , Howev er, the association between eachcomponent of the MetS and OA needs to be f urther addressed. Likewise, we aimed to assess the ov erall link between OA and DM. We perf ormed asy stematic rev iew of the literature and meta­analy sis of data f rom 49 studies inv olv ing a large sample of participants (n=1 192 518). The prev alence of OAamong patients with DM was 29.5±1.2% and that of DM among patients with OA was 14.4±0.1%. Moreov er, OA and DM were signif icantly associated: theov erall risk of OA in the DM population was 1.46 (1.08 to 1.96) and that of DM in the OA population was 1.41 (1.21 to 1.65).

In the DM population, the risk of OA was signif icant with ov erall data. All studies had approximately the same weight in the analy sis. Such a result wasconf irmed in patients older than 50 y ears: DM seems to be associated with OA, ev en when age may hav e a signif icant impact on OA, which suggests thatthe link with DM does not depend on age.

In the OA population, the risk of DM was also signif icant (OR=1.41 (1.21 to 1.65)). Data f rom two studies had an important weight on this f inding: DMprev alence in the OA group was 9.7% in the study of Rahman et al, and 9.8% in the work of Wang et al coming f rom a congress abstract, and thedef initiv e publication f or this abstract will be critical to conf irm these results. It can inf luence the f inal outcome, but the Rahman et al study was of goodquality . The association of OA and DM was not signif icant when we considered only studies of sev ere OA (ie, time to joint replacement), probably becauseof the small number of patients. , , The role of DM in progression of OA is controv ersial since Yoshimura et al hav e f ound that DM def ined asHbA1c f raction ≥5.5% was not independently associated with OA progression, whereas in a recent study ty pe 2 DM was a signif icant predictor of jointspace narrowing in males with symptomatic knee OA.

25 34

36 42 46

16 25 27 48 49

8 10 29 44 46 51 57

8 10 44 46 51 57

10 20 34 42 48 55

25 36 50 31 48

54 59

10 48 59

20 25 42 50 58

10 20 31 34 42 48 59

2 7

61 62

61

10 48 59 64

65

12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 5/8

In addition, recruitment bias at the time of joint replacement may explain the f indings because the presence of comorbidities such as DM may restrict theindication f or surgery in terms of a potential increase in subsequent perioperativ e adv erse ev ents. We f ound an especially signif icant association betweenDM and OA with the studies including hand OA only , which highlights the metabolic and sy stemic nature of hand OA, highlighted recently in the NEOcohort. , Moreov er, the impact of DM on symptoms or on structural lesions might be dif f erent. Schett et al hav e shown that symptoms of OAassessed using the Knee injury and Osteoarthritis Outcome Score (KOOS) and the Western Ontario and McMaster Univ ersities Arthritis Index (WOMAC)were more sev ere and ultrasound sy nov itis and ef f usion of knees more f requent in participants with ty pe 2 DM than those without DM. This inf lammatoryaspect in imaging corroborates with the higher release of inf lammatory mediators in OA cartilage explants f rom patients with DM than those f rom patientswithout DM. The well control of DM by antidiabetic therapies could also inf luence the prev alence of OA: HbA1c f raction, that ref lects the three lastmonths of DM control, was signif icantly higher in women knee OA f or Yoshimura et al , and f or Inoue et al. Howev er, we had no data about thehistory of the DM control during the prev ious y ears during which OA dev eloped and progressed. Moreov er, data about antidiabetic drugs were used only toidentif y patients with DM, but not as a f actor able to inf luence OA occurrence or progression. The assessment of the radiographic patterns of DM­relatedOA (ie, erosiv e OA, dif f use idiopathic skeletal hy perostosis) was not possible due to a lack of data. The prev alence of OA in patients with DM was14.4±0.1%. We measured prev alence of DM among patients with OA, emphasising the link between both diseases. This prev alence could be compared withthe prev alence of OA in the general population, but it depends on the ref erence population: 32% of OA in Iwaki Health Promotion Project (Japan) and 13%of OA in NHANES III (USA). , We lack basic data on the impact of OA on DM, but this remains to be inv estigated because of the possible sy stemicef f ects of OA.

Our meta­analy sis has some limitations. The heterogeneity was high in the f irst analy ses, probably because of population characteristics (v arious OAlocalisations or def initions, v arious DM def initions, no stratif ication on DM sev erity or treatment) or v arious ty pes and qualities of studies. Howev er, weperf ormed sev eral sensitiv ity analy ses, which allowed a decrease in the heterogeneity lev el, in particular in subgroups with a well­recognised def inition ofinclusion criteria of OA. To f urther decrease heterogeneity , we eliminated some studies: those that were of low quality or with a diagnosis of OA not basedon ACR criteria or KL grading. , , , Howev er, the heterogeneity and results did not greatly change because of the low weight of these studies inthe meta­analy sis. Another limitation is the impact of conf ounding f actors, especially age and obesity . Howev er, despite a signif icant impact of increasingage on OA, the association remained positiv e when we retained studies including patients older than 50 y ears. Moreov er, we identif ied sev en studiesshowing an association ev en af ter adjustment on BMI in their multiv ariate logistic regression. The mean score of quality was better and most were recent(results published af ter 2009). Conf ounding f actors such as joint injury , phy sical activ ity , smoking, hy pertension, dy slipidaemia might hav e af f ected ourf indings, but these f actors were taken into account in each included study . , In patients with DM, neuropathy may also af f ect OA dev elopment, but wedid not f ind this inf ormation in the selected studies.

We hav e shown an association of DM and OA, but causality is not y et clearly demonstrated. Hy pergly caemia could promote joint inf lammation andcartilage degradation through oxidativ e stress and inf lammatory mediators induction as well as through AGEs. Bey ond a chronic excess of glucose, ty pe2 DM is characterised by increased insulin resistance that may be inv olv ed in osteophy te dev elopment and subchondral bone sclerosis. , , We thusneed additional specif ic prospectiv e studies f or that purpose.

In summary , this is the f irst meta­analy sis showing an association of OA and DM, giv ing some additional clues about the delineation of the metabolic OAphenoty pe. Large prospectiv e studies are needed to address whether DM is an independent risk f actor of OA dev elopment or sev erity . If this is the case,new prev entiv e and/or curativ e modalities based on gly caemia control could be tested in OA.

AcknowledgmentsThe authors thank Laura SMALES (BioMed Editing, Toronto, Canada) f or editing the manuscript. KL, FB and JS are supported by The Foundation ArthritisNetwork Program (ROAD project).

FootnotesContributors KL, JS and FB were inv olv ed in conception and design. KL was inv olv ed in acquisition of data and statistical analy sis. KL, CV, JSand FB were inv olv ed in analy ses and interpretation of data, draf ting of the manuscript, rev ision of the manuscript and f inal approv al. KL, JS andFB had f ull access to all of the data in the study and take responsibility f or the integrity of the data and the accuracy of the data analy sis.

Competing interests None.

Provenance and peer review Not commissioned; externally peer rev iewed.

Data sharing statement No additional data are av ailable.

Receiv ed January 26, 2015.Accepted March 17, 2015.Published 2 June 2015

Published by the BMJ Publishing Group Limited. For permission to use (where not already granted under a licence) please go tohttp://group.bmj.com/group/rights­licensing/permissions

This is an Open Access article distributed in accordance with the Creativ e Commons Attribution Non Commercial (CC BY­NC 4.0) license, which permitsothers to distribute, remix, adapt, build upon this work non­commercially , and license their deriv ativ e works on dif f erent terms, prov ided the original work isproperly cited and the use is non­commercial. See: http://creativ ecommons.org/licenses/by ­nc/4.0/

References1. Yusuf E, Nelissen RG, Ioan­Facsinay A, et al. Association between weight or body mass index and handosteoarthritis: a sy stematic rev iew. Ann Rheum Dis 2010;69:761–5. doi:10.1136/ard.2008.106930

[Abstract/FREE Full text]2. Bijlsma JW, Berenbaum F, Laf eber FP. Osteoarthritis: an update with relev ance f or clinical practice. Lancet2011;377:2115–26. doi:10.1016/S0140­6736(11)60243­2 [CrossRef ] [Medline] [Web of Science]

3. Abella V, Scotece M, Conde J, et al. Adipokines, metabolic sy ndrome and rheumatic diseases. J Immunol Res2014;2014:343746. doi:10.1155/2014/343746 [Medline]

4. Chauf f ier K, Laiguillon MC, Bougault C, et al. Induction of the chemokine IL­8/Kc by the articular cartilage: possibleinf luence on osteoarthritis. Joint Bone Spine 2012;79:604–9. doi:10.1016/j.jbspin.2011.12.013 [CrossRef ]

[Medline] [Web of Science]Laiguillon MC, Houard X, Bougault C, et al. Expression and f unction of v isf atin (Nampt), an adipokine­enzyme

53 66 67

68

8 64 20

7 20

69

39 41 54 60

8 10

12

9 70 71

12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 6/8

5. Laiguillon MC, Houard X, Bougault C, et al. Expression and f unction of v isf atin (Nampt), an adipokine­enzymeinv olv ed in inf lammatory pathway s of osteoarthritis. Arthritis Res Ther 2014;16:R38. doi:10.1186/ar4467

6. Alberti KG, Zimmet P, Shaw J. The metabolic sy ndrome—a new worldwide def inition. Lancet 2005;366:1059–62.doi:10.1016/S0140­6736(05)67402­8 [CrossRef ] [Medline] [Web of Science]

7. Puenpatom RA, Victor TW. Increased prev alence of metabolic sy ndrome in indiv iduals with osteoarthritis: ananaly sis of NHANES III data. Postgrad Med 2009;121:9–20. doi:10.3810/pgm.2009.11.2073 [CrossRef ]

[Medline]8. Yoshimura N, Muraki S, Oka H, et al. Association of knee osteoarthritis with the accumulation of metabolic riskf actors such as ov erweight, hy pertension, dy slipidemia, and impaired glucose tolerance in Japanese men andwomen: the ROAD study . J Rheumatol 2011;38:921–30. doi:10.3899/jrheum.100569 [Abstract/FREE Full text]

9. Karv onen­Gutierrez CA, Sowers MR, Heeringa SG. Sex dimorphism in the association of cardiometaboliccharacteristics and osteophy tes­def ined radiographic knee osteoarthritis among obese and non­obese adults:NHANES III. Osteoarthritis Cartilage 2012;20:614–21. doi:10.1016/j.joca.2012.02.644 [CrossRef ] [Medline]

[Web of Science]10. Schett G, Kiechl S, Bonora E, et al. Vascular cell adhesion molecule 1 as a predictor of sev ere osteoarthritis of the

hip and knee joints. Arthritis Rheum 2009;60:2381–9. doi:10.1002/art.24757 [CrossRef ] [Medline][Web of Science]

11. Waine H, Nev inny D, Rosenthal J, et al. Association of osteoarthritis and diabetes mellitus. Tufts Folia Med1961;7:13–19. [Medline]

12. Berenbaum F. Diabetes­induced osteoarthritis: f rom a new paradigm to a new phenoty pe. Ann Rheum Dis2011;70:1354–6. doi:10.1136/ard.2010.146399 [Abstract/FREE Full text]

13. Mobasheri A. Glucose: an energy currency and structural precursor in articular cartilage and bone with emerging rolesas an extracellular signaling molecule and metabolic regulator. Front Endocrinol (Lausanne) 2012;3:153.

14. Verzijl N, Bank RA, TeKoppele JM, et al. AGEing and osteoarthritis: a dif f erent perspectiv e. Curr Opin Rheumatol2003;15:616–22. doi:10.1097/00002281­200309000­00016 [CrossRef ] [Medline] [Web of Science]

15. Atay de SA, Yoshinari NH, Nascimento DP, et al. Experimental diabetes modulates collagen remodelling of joints inrats. Histol Histopathol 2012;27:1471–9. [Medline]

16. Sturmer T, Brenner H, Brenner RE, et al. Non­insulin dependent diabetes mellitus (NIDDM) and patterns ofosteoarthritis. The Ulm osteoarthritis study . Scand J Rheumatol 2001;30:169–71. doi:10.1080/030097401300162969

[CrossRef ] [Medline] [Web of Science]17. Frey MI, Barrett­Connor E, Sledge PA, et al. The ef f ect of noninsulin dependent diabetes mellitus on the prev alence

of clinical osteoarthritis. A population based study . J Rheumatol 1996;23:716–22. [Medline] [Web of Science]18. v on Elm E, Altman DG, Egger M, et al. [The Strengthening the Reporting of Observ ational Studies in Epidemiology

(STROBE) statement: guidelines f or reporting observ ational studies]. Rev Esp Salud Publica 2008;82:251–9.doi:10.1590/S1135­57272008000300002 [Medline] [Web of Science]

19. Ray S, Datta AK, Sinhamahapatra P, et al. Prev alence of rheumatic conditions in patients with diabetes mellitus in atertiary care hospital. J Indian Med Assoc 2011;109:74–8. [Medline]

20. Inoue R, Ishibashi Y, Tsuda E, et al. Medical problems and risk f actors of metabolic sy ndrome among radiographicknee osteoarthritis patients in the Japanese general population. J Orthop Sci 2011;16:704–9. doi:10.1007/s00776­011­0157­9 [CrossRef ] [Medline]

21. Lanas A, Tornero J, Zamorano JL. Assessment of gastrointestinal and cardiov ascular risk in patients withosteoarthritis who require NSAIDs: the LOGICA study . Ann Rheum Dis 2010;69:1453–8. doi:10.1136/ard.2009.123166

[Abstract/FREE Full text]22. Sarkar P, Pain S, Sarkar RN, et al. Rheumatological manif estations in diabetes mellitus. J Indian Med Assoc

2008;106:593–4. [Medline]

23. Reeuwijk KG, de Rooij M, v an Dijk GM, et al. Osteoarthritis of the hip or knee: which coexisting disorders aredisabling? Clin Rheumatol 2010;29:739–47. doi:10.1007/s10067­010­1392­8 [CrossRef ] [Medline]

24. Miksch A, Hermann K, Rolz A, et al. Additional impact of concomitant hy pertension and osteoarthritis on quality oflif e among patients with ty pe 2 diabetes in primary care in Germany—a cross sectional surv ey . Health Qual LifeOutcomes 2009;7:19. doi:10.1186/1477­7525­7­19 [CrossRef ] [Medline]

25. Dahaghin S, Bierma­Zeinstra SM, Koes BW, et al. Do metabolic f actors add to the ef f ect of ov erweight on handosteoarthritis? The Rotterdam Study . Ann Rheum Dis 2007;66:916–20. doi:10.1136/ard.2005.045724

[Abstract/FREE Full text]26. Adams AL, Paxton EW, Wang JQ, et al. Surgical outcomes of total knee replacement according to diabetes status

and gly cemic control, 2001 to 2009. J Bone Joint Surg Am 2013;95:481–7. doi:10.2106/JBJS.L.00109[Abstract/FREE Full text]

27. Martin K, Lethbridge­Cejku M, Muller DC, et al. Metabolic correlates of obesity and radiographic f eatures of kneeosteoarthritis: data f rom the Baltimore Longitudinal Study of Aging. J Rheumatol 1997;24:702–7. [Medline]

[Web of Science]28. Shirinsky I, Shirinsky V. Diabetes ef f ects on pain and phy sical f unction in incidence and progression subcohorts of

the osteoarthritis initiativ e: a 5­y ear longitudinal data analy sis. EULAR Congress ; 2013;Abstract OP0026.29. Hart DJ, Doy le DV, Spector TD. Association between metabolic f actors and knee osteoarthritis in women: the

Chingf ord Study . J Rheumatol 1995;22:1118–23. [Medline] [Web of Science]30. Cimmino MA, Sarzi­Puttini P, Scarpa R, et al. Clinical presentation of osteoarthritis in general practice: determinants

of pain in Italian patients in the AMICA study . Semin Arthritis Rheum 2005;35:17–23.doi:10.1016/j.semarthrit.2005.01.015 [CrossRef ] [Medline] [Web of Science]

31. Typpo T. Osteoarthritis of the hip. Radiologic f indings and etiology . Ann Chir Gynaecol Suppl 1985;201:1–38.[Medline]

32. Magnusson K, Hagen KB, Osteras N, et al. Diabetes is associated with increased hand pain in erosiv e handosteoarthritis—data f rom a population­based study . Arthritis Care Res (Hoboken) 2015;67:187–95.doi:10.1002/acr.22460

33. Conaghan PG, Peloso PM, Ev erett SV, et al. Inadequate pain relief and large f unctional loss among patients withknee osteoarthritis: ev idence f rom a prospectiv e multinational longitudinal study of osteoarthritis real­world therapies.Rheumatology (Oxford) 2015;54:270–7. doi:10.1093/rheumatology /keu332 [Abstract/FREE Full text]

34. Navarro N, Orellana C, Vasquez I, et al. High f requency of cardiov ascular disease in patients with kneeosteoarthritis in a primary care setting. EULAR Congress ; 2012;Abstract SAT0320.

Zullig LL, Bosworth HB, Jef f rey s AS, et al. The association of comorbid conditions with patient­reported outcomes in

12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 7/8

35. Zullig LL, Bosworth HB, Jef f rey s AS, et al. The association of comorbid conditions with patient­reported outcomes inVeterans with hip and knee osteoarthritis. Clin Rheumatol 2014. Published Online First.

36. Orellana C, Nav arro N, Calv et J, et al. Higher f requency of metabolic sy ndrome in patients with hand osteoarthritisis more pronounced in obese patients. EULAR Congress ; 2012;Abstract FRI0305.

37. Bija MD, Luma HN, Temf ack E, et al. Patterns of knee osteoarthritis in a hospital setting in sub­Saharan Af rica. ClinRheumatol 2014. Published Online First.

38. Anderson JJ, Felson DT. Factors associated with osteoarthritis of the knee in the f irst national Health and NutritionExamination Surv ey (HANES I). Ev idence f or an association with ov erweight, race, and phy sical demands of work.Am J Epidemiol 1988;128:179–89. [Abstract/FREE Full text]

39. Silv eri F, Brecciaroli D, Argentati F, et al. Serum lev els of insulin in ov erweight patients with osteoarthritis of theknee. J Rheumatol 1994;21:1899–902. [Medline] [Web of Science]

40. Perruccio AV, Kandel RA, Dav is AM. Cardiov ascular disease in osteoarthritis: hip v ersus knee and the inf luence ofmultiple symptomatic joint inv olv ement. ACR Congress ; 2013;Abstract 266.

41. Wang Y, Peng R, Ma R. Epidemiological inv estigation of osteoarthritis in middle­aged Mongolian and senior residentsof the inner Mongolia autonomous region. Iran Red Crescent Med J 2013;15:e8303. doi:10.5812/ircmj.8303

42. Shin D. Association between metabolic sy ndrome, radiographic knee osteoarthritis, and intensity of knee pain:results of a national surv ey . J Clin Endocrinol Metab 2014;99:3177–83. doi:10.1210/jc.2014­1043 [CrossRef ]

[Medline]43. Jamsen E, Peltola M, Eskelinen A, et al. Comorbid diseases as predictors of surv iv al of primary total hip and knee

replacements: a nationwide register­based study of 96 754 operations on patients with primary osteoarthritis. AnnRheum Dis 2012;72:1975–82. doi:10.1136/annrheumdis­2012­202064

44. Dav ies­Tuck ML, Wang Y, Wluka AE, et al. Increased f asting serum glucose concentration is associated withadv erse knee structural changes in adults with no knee symptoms and diabetes. Maturitas 2012;72:373–8.doi:10.1016/j.maturitas.2012.05.013

45. Baker C, Aileen L, Lav alley MP, et al. Automated telephone­linked communication: a nov el approach to enhancelong­term adherence to resistance training exercice among people with knee osteoarthritis. ACR Congress ;2013;Abstract 1818.

46. Siv iero P, Tonin P, Maggi S. Functional limitations of upper limbs in older diabetic indiv iduals. The Italian LongitudinalStudy on Aging. Aging Clin Exp Res 2009;21:458–62. doi:10.1007/BF03327449 [Medline]

47. Peniston JH, Gold MS, Wieman MS, et al. Long­term tolerability of topical diclof enac sodium 1% gel f or osteoarthritisin seniors and patients with comorbidities. Clin Interv Aging 2012;7:517–23. doi:10.2147/CIA.S35416 [Medline]

48. Engstrom G, Gerhardsson de Verdier M, Rollof J, et al. C­reactiv e protein, metabolic sy ndrome and incidence ofsev ere hip and knee osteoarthritis. A population­based cohort study . Osteoarthritis Cartilage 2009;17:168–73.doi:10.1016/j.joca.2008.07.003 [CrossRef ] [Medline] [Web of Science]

49. Bagge E, Bjelle A, Eden S, et al. Factors associated with radiographic osteoarthritis: results f rom the populationstudy 70­y ear­old people in Goteborg. J Rheumatol 1991;18:1218–22. [Medline] [Web of Science]

50. Haugen IK, Ramachandran VS, Misra D, et al. Hand osteoarthritis in relation to mortality and incidence ofcardiov ascular disease: data f rom the Framingham Heart Study . Ann Rheum Dis 2013;7474–81.

51. Sowers M, Karv onen­Gutierrez CA, Palmieri­Smith R, et al. Knee osteoarthritis in obese women with cardiometabolicclustering. Arthritis Rheum 2009;61:1328–36. doi:10.1002/art.24739 [CrossRef ] [Medline]

52. Visser W, den Heijer M, Spoelman W, et al. Glucose and insulin concentration in association with hand osteoarthritis:the NEO study . EULAR Congress ; 2013;Abstract SAT0350.

53. Visser AW, de Mutsert R, le Cessie S, et al. The relativ e contribution of mechanical stress and sy stemic processesin dif f erent ty pes of osteoarthritis: the NEO study . Ann Rheum Dis 2014. Published Online First

54. Philbin EF, Ries MD, Grof f GD, et al. Osteoarthritis as a determinant of an adv erse coronary heart disease riskprof ile. J Cardiovasc Risk 1996;3:529–33. doi:10.1097/00043798­199612000­00008 [CrossRef ] [Medline]

55. Ladjimi A, Youssef S, Chamakhi S, et al. [Rheumatologic manif estations in diabetes]. Tunis Med 1985;63:213–19.[Medline]

56. Denko CW, Boja B, Moskowitz RW. Growth promoting peptides in osteoarthritis and dif f use idiopathic skeletalhy perostosis—insulin, insulin­like growth f actor­I, growth hormone. J Rheumatol 1994;21:1725–30. [Medline]

[Web of Science]

57. Niev es­Plaza M, Castro­Santana LE, Font YM, et al. Association of hand or knee osteoarthritis with diabetes mellitusin a population of Hispanics f rom Puerto Rico. J Clin Rheumatol 2013;19:1–6. doi:10.1097/RHU.0b013e31827cd578

[Medline]58. Cimmino MA, Cutolo M. Plasma glucose concentration in symptomatic osteoarthritis: a clinical and epidemiological

surv ey . Clin Exp Rheumatol 1990;8:251–7. [Medline] [Web of Science]59. Lindberg H, Nilsson BE. Coinciding morbidity in patients with coxarthrosis. An epidemiological study of roentgen

examinations. Arch Orthop Trauma Surg 1985;104:82–4. doi:10.1007/BF00454242 [Medline]60. Dequeker J, Burssens A, Bouillon R. Dy namics of growth hormone secretion in patients with osteoporosis and in

patients with osteoarthrosis. Horm Res 1982;16:353–6. doi:10.1159/000179525 [Medline]

61. Rahman MM, Kopec JA, Cibere J, et al. The relationship between osteoarthritis and cardiov ascular disease in apopulation health surv ey : a cross­sectional study . BMJ Open 2013;3:pii: e002624 doi:10.1136/bmjopen­2013­002624

62. Wang S, Ganguli AX, Macaulay D, et al. The economic burden of osteoarthritis in Americans: analy sis f rom apriv ately insured population. ACR Congress ; 2013.

63. Horn CA, Bradley JD, Brandt KD, et al. Impairment of osteophy te f ormation in hy pergly cemic patients with ty pe IIdiabetes mellitus and knee osteoarthritis. Arthritis Rheum 1992;35:336–42. doi:10.1002/art.1780350313 [CrossRef ]

[Medline] [Web of Science]64. Yoshimura N, Muraki S, Oka H, et al. Accumulation of metabolic risk f actors such as ov erweight, hy pertension,

dy slipidaemia, and impaired glucose tolerance raises the risk of occurrence and progression of knee osteoarthritis: a3­y ear f ollow­up of the ROAD study . Osteoarthritis Cartilage 2012;20:1217–26. doi:10.1016/j.joca.2012.06.006

[CrossRef ] [Medline] [Web of Science]

65. Eymard F, Parsons C, Edwards MH, et al. Diabetes is a risk f actor f or knee osteoarthritis progression. OsteoarthritisCartilage 2015. Published Online First: 3 February 2015. doi:10.1016/j.joca.2015.01.013

66. Sellam J, Berenbaum F. Is osteoarthritis a metabolic disease? Joint Bone Spine 2013;80:568–73.doi:10.1016/j.jbspin.2013.09.007 [CrossRef ] [Medline] [Web of Science]

Schett G, Kley er A, Perricone C, et al. Diabetes is an independent predictor f or sev ere osteoarthritis: results f rom a

12/13/2015 Association between diabetes mellitus and osteoarthritis: systematic literature review and meta­analysis ­­ Louati et al. 1 (1) ­­ RMD Open

http://rmdopen.bmj.com/content/1/1/e000077.full 8/8

67. Schett G, Kley er A, Perricone C, et al. Diabetes is an independent predictor f or sev ere osteoarthritis: results f rom alongitudinal cohort study . Diabetes Care 2013;36:403–9. doi:10.2337/dc12­0924 [Abstract/FREE Full text]

68. Laiguillon MC, Courties A, Houard X, et al. Characterization of diabetic osteoarthritic cartilage and role of highglucose env ironment on chondrocy te activ ation: toward pathophy siological delineation of diabetes mellitus­relatedosteoarthritis. Osteoarthritis Cartilage Published Online First: 15 May 2015. doi:10.1016/j.joca.2015.04.026

69. Berenbaum F, Eymard F, Houard X. Osteoarthritis, inf lammation and obesity . Curr Opin Rheumatol 2013;25:114–18.doi:10.1097/BOR.0b013e32835a9414 [CrossRef ] [Medline]

70. Okazaki K, Jingushi S, Ikenoue T, et al. Expression of insulin­like growth f actor I messenger ribonucleic acid indev eloping osteophy tes in murine experimental osteoarthritis and in rats inoculated with growth hormone­secretingtumor. Endocrinology 1999;140:4821–30. doi:10.1210/endo.140.10.7053 [CrossRef ] [Medline]

[Web of Science]71. Massicotte F, Fernandes JC, Martel­Pelletier J, et al. Modulation of insulin­like growth f actor 1 lev els in human

osteoarthritic subchondral bone osteoblasts. Bone 2006;38:333–41. doi:10.1016/j.bone.2005.09.007