Annals of Physical and Rehabilitation Medicine 61 (20 18) 99- 104
`
`Available online at
`Science Direct
`www.sciencedirect.com
`
`Elsevie r Masson Fran ce
`EMiconsulte
`www.e m-co nsulte.co m
`
`ELSEVIER
`
`Review
`Hand dysfunction in type 2 diabetes mellitus: Systematic review with
`meta-analysis
`
`<l) crossMark
`
`Shub ha Gundmi a , Arun G. Mai ya a.*, Anil K. Bhat b , N. Ravishankar C,
`Manjunatha H. Hande d, K.V. Rajagopal e
`
`a Department of physiotherapy, school of allied health sciences, Manipa/ Academy of Higher Education, Udupi, Karnataka, India
`b Deparrment of orthopaedics, Kasturba medical col/ege,Manipal Academy of Higher Education. Udupi, Karnataka, India
`'Prasanna school of statistics, department of bio-statistics, Manipal Academy of Higher Education, Udupi, Kamaraka, India
`d Department of medicine, Kasturba medical college, Manipal Academy of Higher Educal'ion, Udupi, Kamataka, India
`0 Department of radiology, Kasturba medical college, Manipal Academy of Higher Education, Udupi, Karnataka, India
`
`ARTIC L E IN F O
`
`ABSTRAC T
`
`Article history:
`Received 4 August 20 17
`Acce pted 24 December 2017
`
`Keywords:
`Type 2 diabetes mellitus
`Hand
`Dysfunction
`Strength
`Dexteri ty
`
`Background: People with ty pe 2 diabetes mellitus frequently show compl ications in feet and hands.
`However, the literature has mostly foc used on foot complications. The disease can affect the strength and
`dexterity of the hands. thereby reducing function.
`Objectives: This systematic review and meta-analysis focused on identifying the existing evidence on
`how type 2 diabetes mellitus affects hand strength. dexterity and function.
`Methods: We searched MEDLlNE via PubMed, CINHAL, Scopus and Web of Science, and the Cochrane
`central register of controlled tria ls for reports of studies of grip and pinch strength as well as hand
`dexterity and function eval uated by questionnaires comparing patients with type 2 diabetes mellitus
`and healthy controls that were published between 1990 and 2017. Data are reported as standard ized
`mean difference (SMD) or mean difference (MD) and 95% confid ence intervals (C! s).
`Results: Among 2077 records retrieved. only 7 full-text articles were available for meta-analysis. For both
`the dominant and non-dominant hand, type 2 diabetes mellitus negatively affected gri p strength (SMD:
`- 1.03; 95% CI: - 2.24 to 0. 18 and -1.37, -3.0 7 to 0.33 ) and pinch strength (-1.09, -2.56 to 0.38 and
`- 1.12, -2.73 to 0.49 ), although not significantly. Dexterity of the domi nant hand did not differ between
`diabetes and co ntrol groups but was poorer for the non-dominant hand, although not significantly. Hand
`function was worse fo r diabetes tha n control groups in 2 studies (MD: -8.7; 9 5% CI: -1 6 .88 to -1.52 and
`4.69, 2.03 to 7.35 ).
`Conclusion: This systematic review with meta-analysis suggested red uced hand function, specifically
`grip and pinch strength. for people with type 2 diabetes mellitus vers us healthy controls. However, the
`sample size for all studies was low. Hence, we need studies with adeq uate sample size and randomized
`controlled trials to provide statistically significant results.
`© 2018 Elsevier Masson SAS. All rights reserved.
`
`1. Introduction
`
`Diabetes mellit us (DM ) is a group of metabolic disorders
`characterized by chronic hyperglycaemia w ith disturbed carbohy(cid:173)
`drate, fat and protein m etabolism due to absolute or relative
`deficiency in insulin secretion and/or action [ 1 ]. The pre valence of
`type 2 DM (T2DM) is increasing across the globe. According to the
`
`• Corresponding author.
`E-mail addresses: s.gmaiya@gmail.com (S. Gundmi ), arun.maiya.g@gmai l.com
`(A.G. Maiya), anil.bhat@manipal.edu (A. K. Bhat), ravi.shankar@manipal.edu
`(N . Ravishankar), manj unath.hande@manipal.edu ( M .H. Hande),
`rajagopal.kv@manipal.edu (K.V. Rajagopal).
`
`https://doi.org/ 10.1 O 16/j.rehab.2017.12.006
`1877-0657/© 2018 Elsevier Masson SAS. All rights reserved.
`
`International Diabetes Federation, 415 million adults are estimated
`to have T2DM. One in 11 adults has T2DM. T2DM is more prevalent
`in low and m iddle socio-economic countries [2 ).
`Wit h the increase in prevalence of T2DM, complications
`associated with the disease also increase. The main reason for
`complications is poor glycaemic control and diabetes screening,
`especially in low socio-economic count ries, lack of awareness
`among people, and lack of health care facilities in rural areas
`(3 ]. T2DM affects many parts of the body, the most common
`complications being d iabetic cardiovascular problems, retinopa(cid:173)
`thy, nephropathy, and peripheral neuropathy [4 ]. Peripheral
`neuropathy with a diabetes origin affects both upper and lower
`extremities. Throughout the literature, peripheral neuropathy of
`
`Sanofi Exhibit 2178.001
`Mylan v. Sanofi
`IPR2018-01676
`
`

`

`100
`
`S. Gundmi er al. / Annals of Physical and Rehabilitation Medicine 61 (2018) 99-104
`
`foot complications with T2DM are given much attention and less is
`known about peripheral neuropathy of the hand [5].
`In T2DM, abnormal cross-linking of collagen fibres occurs due
`to accumulation of advanced glycosylation end-products, which
`leads to skin thickening and formation of nodules and contractures
`[6] . Commonly seen hand complications with T2DM are limited
`joint mobility syndrome, also known as diabetic cheiroarthropathy
`or stiff hand syndrome, Dupuytren's contracture, flexor tenosyno(cid:173)
`vitis (trigger finger) and carpal tunnel syndrome [7].
`Hand complications in patients with T2DM may affect activities
`of daily living and lead to disabilities in self-care activities. These
`result in reduced interpersonal interactions, loss of independence,
`financial burden and overall reduced quality of life [8]. However,
`we have little research pertaining to hand dysfunction in T2DM.
`With the increasing life expectancy and steep increase in number
`of people with T2DM, we need more research on hand function to
`address the standard of living and self-reliability in general and
`fine tasks.
`With the increase in prevalence of T2DM worldwide and in
`India, the accompanying complications may disturb activities of
`daily
`living and quality of life. Unlike
`the diabetic foot,
`complications of hands with T2DM are easily neglected. Only a
`few studies have assessed hand strength, dexterity and dysfunc(cid:173)
`tion in people with T2DM. The reporting of hand dysfunction in
`these patients lacks agreement among studies. Thus, considering
`the increasing rate in number of people living with T2 DM and the
`increased life expectancy, a study of hand function may help
`improve care, independence in activities of daily living and quality
`of life.
`Hence, we performed a systematic review and meta-analysis to
`provide evidence of the effect ofT2DM on hand strength, dexterity
`and function.
`
`2. Methods
`
`According to the Prisma statement, the review was performed
`for quality of reporting of a meta-analysis.
`
`2.1. Literature search
`
`We searched MEDLINE via PubMed, Scopus, Science Direct,
`Web of Science, Cochrane Central register of controlled trials, and
`CINHAL for articles published in English from June 1, 2017 to June
`15, 2017 by using the MESH and keywords "type 2 diabetes
`mellitus", "hand dysfunction", "hand function", "hand strength",
`"hand dexterity", including the Boolean operator AND/OR. Full(cid:173)
`text articles were selected for the review.
`In the meta-analysis, we included articles with the following
`3 criteria to achieve a homogenous sample for further ana lysis:
`
`• participants had T2DM;
`• age-matched controls were not diabetic or with impaired
`glucose tolerance;
`• evaluation was of hand grip strength (with the hand Jamar
`dynamometer), pinch strength (pinch meter), and dexterity
`test), with hand function assessed by
`(Purdue Pegboard
`validated questionnaires.
`
`2.2. Assessment of risk of bias
`
`The included studies were assessed for risk of bias by using the
`US National Heart, Lung and Blood Institute checklist for
`observational cohort and cross-sectional studies. In the checklist,
`6 questions were applicable to the cu rrent study. Questions 1 to
`
`5 and 11 mainly focus on methodology: characteristics of the study
`population, rate of patticipation of eligible candidates, estimation
`of the sample size and adjustment for confounding factors. The
`quality assessment was performed by 2 independent reviewers.
`The scoring was Yes, No, cannot determine/not applicable or not
`reported. The study was rated as poor quality with score < 4; fair
`with score 4 to 5, and good with all scores 2: 6. The mean score for
`the 2 reviewers was considered for each domain.
`
`2.3. Study screening and data extraction
`
`Two authors (GS and AM) independently screened all titles for
`inclusion. Abstracts of potentially eligible studies were obtained,
`then full texts. Any discrepancies between the authors were
`resolved by discussion. Data were extracted by the fi rst author (GS)
`with the help of a qualified statistician.
`
`3. Statistical analysis
`
`Because all our outcomes were continuous, we calculated mean
`difference/standardised mean difference (MD/SMD) statistics. For
`the meta-analysis, we synthesized SMDs because the study
`authors used different instruments for measuring outcomes. For
`the studies not included in meta-analyses, we calculated MDs.
`Meta-analysis was performed when at least 2 studies were
`similar in terms of the PICO process and study design providing
`relevant data. We adopted a random-effects model for the meta(cid:173)
`analysis because we anticipated considerable heterogeneity
`among the studies. To assess heterogeneity, we used the Chi2
`statistic (P < 0.1 considered statistically significant) and evaluated
`heterogeneity with the I2 statistic( > 60% considered substantial
`heterogeneity ). Meta-analysis involved use of RevMan 5.2. We
`present forest plots for all meta-analyses. When meta-analysis was
`not appropriate, the effect size is presented with 95% confidence
`intervals ( Cls ). We performed meta-analysis of the effect of gender
`and age on grip strength of the dominant hand only because of few
`studies to analyse the effect size for other outcomes.
`
`4. Results
`
`4.1 . Study selection
`
`From the electronic database search, we identified 2077 articles;
`after removing duplicates and screening for eligibility criteria,
`1579 articles were excluded. Overall, 24 full -text articles were
`eligible for review; 17 did not meet the inclusion criteria (Fig. 1 ), so
`finally, 7 articles were included in the final review and meta(cid:173)
`analysis. Records were excluded because of inappropriate title and
`study methodology; no control group; improper study design and
`outcome measure, statistical analysis, and tools used in the study;
`inappropriate data ; and publication language other than English.
`
`4.2. Study quality
`
`the review showed fair quality
`in
`The studi es included
`according to the total score on the US National Heart, Lung and
`Blood Institute checklist (Table 1 ). None of the reports stated how
`the sample size was calculated to detect the clinically significant
`effect. Various confounding factors were not taken into consid(cid:173)
`eration and could have influenced the outcome of interest.
`
`4.3. Characteristics and recruitment of participants
`
`A total of 761 participants were ana lysed: 425 in the study
`groups and 341 in the control groups. People with T2DM and
`
`Sanofi Exhibit 2178.002
`Mylan v. Sanofi
`IPR2018-01676
`
`

`

`S. Gundmi et al. / Annals of Phys ical and Rehabilirarion Medicine 61 (201 8) 99-104
`
`101
`
`Records id enti fied by
`database searching
`In= 2077)
`
`Re cord s after dupl icates
`removed
`(n=l ,6471
`
`Records scree ned
`
`(n= 1,647}
`
`-+
`
`Reco rds excluded
`(n=l579)
`
`Full t e>:t arti cles assessed
`for eligibility
`(n=24)
`
`_ . .
`
`Full text articles
`excl uded (did not meet
`inclusion crite ria n= 17)
`
`St udi es inclu ded in
`qualita tive synthesis
`(n=7)
`
`Studies incl uded in meta(cid:173)
`analysis
`ln= 7}
`
`Fig. 1. Flow of t he selection of stu dies for the meta-analysis.
`
`healthy age-matched controls were studied. Characteristics of
`participants are in the table. Most participants were recruited from
`hospital and outpatient settings.
`
`4.4. Outcome measures
`
`Each outcome of interest included in the review (Table 1 ) is
`discussed in detail below.
`
`4.4.1. Grip strength
`Among the 7 studies, 5 reported grip strength for people with
`T2DM and healthy age-matched controls [9- 11,13,14) (Figs. 2- 3 ).
`Three studies reported means and SDs for dominant and non(cid:173)
`dominant hands separately [10, 11 , 13). Meta-analysis of the
`dominant hand showed high heterogeneity among the studies
`(12 = 97%) (Fig. 2 ). We revealed a negative combined effect of grip
`strength in the dominant hand, indicating a lower mean in the
`
`Table 1
`Characteristics of the studies included in the review.
`
`study group, or diabetes group, than control group (SMD: - 1.03;
`95% CJ: - 2.24 to 0.18) and suggesting low grip strength with
`T2DM. Also, meta-analysis of grip strength for the non-dominant
`hand indicated a negative combined effect of diabetes (SMD:
`- 1.37; 95% CJ: - 3.07 to 0.33 ), suggesting low grip strength with
`T2DM (Fig. 3 ). The heterogeneity among the studies for the non(cid:173)
`dominant hand was also high (12 = 98%).
`
`4.4.2. Pinch strength
`Three studies reported pinch strength of the dominant and non(cid:173)
`dominant hand separately forT2DM and non-diabetes participants
`[11,13 ,14). We found high heterogeneity among the studies,
`12 = 96% and 97% for the dominant and non-dominant hand,
`respectively. We found a negative combined mean for both the
`dominant hand (SMD: - 1.09; 95% Cl: - 2.56 to 0.38) and non(cid:173)
`dominant hand (- 1.12, - 2.73 to 0.49), suggesting reduced pinch
`strength in study than control groups (Figs. 4-5 ).
`
`4.4.3. Hand dexterity
`Studies used the Purdue Pegboard test-to-test dexterity. Only
`2 studies reported dexterity of dominant and non-dominant hand
`separately for T2DM and non-diabetes participants [12,14]. The
`meta-analysis revealed no significant difference in hand dexterity
`in the dominant hand between the study and control groups (SOM:
`- 0.07; 95% CJ: - 0.51 to 0.36). However, we found a negative
`combined effect of dexterity for the non-dominant hand (- 0.54,
`- 1.43 to 0.35 ). The heterogeneity among the studies was J2 = 0%
`and 76% for the dominant and non-dominant hand, respectively
`(Figs. 6- 7).
`
`4.4.4. Hand fenction
`Two studies used validated questionnaires to assess hand
`function in people with T2DM versus healthy controls [12,15). Yang
`et al. used the Michigan Hand Outcome Questionnaire to test hand
`function, with a low score except for the pain dimension indicating
`poor hand function [12]. Savas et al. used the Durui:iz hand index
`(DH!) self-reporting questionnaire, with a high total score indicating
`more disability [ 15 ]. Hence we could not use these studies for meta(cid:173)
`analysis because they were not comparable, but we measured the
`MD for individual studies. The MDs between the study and control
`groups for studies conducted by Yang et al. and Savas et al. were
`- 8.7 (95% Cl - 16.88 to - 1.52) and 4.69 (2.03 to 7.35 ) [12, 15]. There(cid:173)
`fore, in these 2 studies, hand function was significantly reduced with
`T2DM as compared with healthy controls or non-d iabetics.
`
`Study
`
`Study area
`
`No of participants
`in study group
`
`No of partici pan ts
`in control group
`
`Age of pa rticipants.
`years. m ean (S D)
`
`Outcome measures used
`
`Quality assessment
`
`Savas et al., 2007
`
`Turkey
`
`44
`
`Cederlund et al., 2009
`
`Sweden
`
`23
`
`Carvalho et al.. 2014
`
`Brazil
`
`100
`
`M ohammed et al., 20 14
`
`Egypt
`
`Cetinus et al.. 2005
`
`Turkey
`
`Kaur et al.. 2016
`
`India
`
`Yang et al.. 201 7
`
`Taiwan
`
`40
`
`76
`
`50
`
`92
`
`SG: study group: CG: control group.
`
`50
`
`3 5
`
`100
`
`40
`
`47
`
`50
`
`9
`
`58.98 (9.07)
`
`75 (1.4)
`
`64.1 8 (8.10 ) (SG)
`63.82 (6.58 ) (CG)
`51 (5.59) (SG)
`48.05 (6.73 ) (CG )
`50.11 (7.6) (SG)
`46.9 (10.2) (CG )
`54.74 (2.6 1) (SG)
`55.28 (2.83) (CG )
`59.4 (9.5) (SG )
`52.9 (6.5) (CG )
`
`Grip strength
`Pinch strength
`Duruoz hand index
`questionnaire
`Grip strength
`Pinch strength
`Purdue Pegboard test
`Grip strength
`
`Grip st rength
`Pinch strength
`Grip strength
`Pinch strength
`Grip strength
`
`Perdue Pegboard test
`Michigan Hand Outcome
`questionnaire
`
`Fai r
`
`Fair
`
`Fair
`
`Fair
`
`Fair
`
`Fa ir
`
`Fair
`
`Sanofi Exhibit 2178.003
`Mylan v. Sanofi
`IPR2018-01676
`
`

`

`102
`
`S. Gundmi er al. / Annals of Physical and Rehabilitation Medicine 61 (2018) 99-104
`
`Study or Subgroup
`Carvalho F De elal 2004
`Cederlund RI et al 2009
`Centi nus E et al 2005
`Kaur P et al 2016
`Mohamed EK et al 2014
`
`Non diabetes
`Diabetes
`Mean
`SD Total Mean
`SD Total
`38.3 12.29 100
`28.3 25.93
`100
`41
`23.7
`23
`39 16.29
`35
`28.92 10.86
`76 33.22 10.53
`47
`20.76
`3.55
`7.6
`50
`32.9
`50
`24.9
`5.63
`40 42.75
`4.65
`40
`
`Weight
`20.4%
`19.9%
`20.3%
`20.0%
`19.3%
`
`Std. Mean Difference
`IV, Random, 95% Cl
`0.49 [0.21, 0.77]
`0.10 [-0.43, 0.63]
`-0.40 [-0.77, -0.03]
`·2.03 [·2.52, -1 .55]
`-3.42 [-4 .12, -2.73]
`
`Total (95% Cl)
`272
`289
`Heterogenellf. Tall"= 1.85; Chi'= 157.22, df = 4 (P < 0.00001 ); I'= 97%
`Testfor overall effect: Z= 1.67 (P = 0.10)
`
`100.0%
`
`-1.03 [-2.24, 0.18]
`
`Std. Mean Difference
`IV, Random, 95% Cl
`
`---
`
`-4
`0
`·2
`Favours [Diabetes] Favours [Non diabetes]
`
`Fig. 2. Forest plot of meta-ana lysis of grip strength of the dominant hand comparing type 2 diabetes mellitus (T2 DM) and healthy controls (negative mean difference indicates
`higher mean va lues in controls ).
`
`study or Subgroup
`Carvalho F De etal 2004
`Centinus Eel al 2005
`Mohamed EK et al 2014
`
`Diabetes
`Mean
`SD
`32.4 12.29
`27.54 10.51
`4.64
`19.07
`
`Non diabetes
`SD Total Weight
`Total Mean
`28.3 25.92
`100 34.1%
`100
`76 32.05
`47 33.8%
`9.3
`40 37.87
`4.65
`40 32. 1%
`
`Std. Mean Difference
`IV. Random, 95!, Cl
`0.20 (-0.08, 0.48]
`-0.45 (-0.81, -0.08]
`-4.0 1 (·4 .78,-3.24]
`
`187 100.0%
`216
`Total (95% Cl)
`Heterogeneity: Tau' = 2.20: Chi' = 101.36, df= 2 (P • 0.00001); I' = 98%
`Testfor overall effect Z = 1.58 (P = 0.12)
`
`-1.37 (-3.07, 0.33]
`
`Std. Mean Difference
`IV. Random, 95~, Cl
`
`10
`5
`0
`-5
`-10
`Favours (Study group] Favours [Control group]
`
`Fig. 3. Forest plot of meta-analysis of grip strength of the non-dominant hand comparing type 2 d iabetes mellitus (T2DM ) and healthy controls (negative mean difference
`indicates higher mean values in conlrnls).
`
`Study or Sunorou p
`Cederltmd RI et al 2009
`Centi nus E et al 2005
`Mohamed EK et al 2014
`
`Non diabetes
`Diabetes
`so Total Mean
`so Total weiaht
`Mean
`8 5.19
`23
`8.5 5.93
`35 33.3%
`47 34.0%
`8.47 2.56
`76
`9.37 1.89
`5.75 1.75
`40
`11 .3 2.11
`40 32.7%
`
`Std. Mean Difference
`IV, Ranaom, 95'¼ Ci
`-0.091-0.61, 0.44]
`-0.38 [-0.75, -0.02]
`-2.841-3.46, -2.21 ]
`
`122 100.0~.
`139
`Total (95% Cl)
`Heterogeneity: Tau'= 1.62; Chi'= 52.19, df= 2 (P < 0.00001 ); I' : 96%
`Test for overall effect Z = t .45 (P = 0.15)
`
`-1.09 [-2.56, 0.38]
`
`Std. Mean Difference
`IV, Random, 95% Cl
`
`4
`2
`0
`-2
`-4
`Favours IDiat>elesJ Favours [Non <JiabetesJ
`
`Fig. 4. Forest plot of meta-analy sis of pinch strength of the dominan t hand comparing type 2 diabetes mellitus (T2DM) and healthy controls ( negative mean difference
`indicates higher mean values in controls).
`
`Study or Subgroup
`Cederlund RI et al 2009
`Centi nus E et al 2005
`Mohamed EK et al 2014
`
`Non diabetes
`Diabetes
`so Total Mean
`SD Total Weight
`Mean
`8 5.56
`23
`8 6.74
`35
`33.3%
`8.15
`76
`8.92 1.83
`47
`34.0%
`2.5
`32.7%
`5.52 2.07
`40 10.72 1.15
`40
`
`Std. Mean Difference
`IV, Random, 95•,1, Cl
`0.00 1-0.53, 0.531
`-0.34 (-0.70, 0.03]
`-3.08 1-3. 73, -2.42]
`
`122 100.0%
`139
`Total {951, Cl)
`Heterogeneity: Tau'= 1.95; Chi"= 60.85, df= 2 (P < 0.00001); I'= 97%
`Te sl for overall effect: Z= 1.36 (P = 0.17)
`
`-1.12 [-2.73, 0.49]
`
`Std. Mean Difference
`IV, Random, 95%CI
`
`---
`
`4
`2
`0
`-2
`-4
`Favours [Diabetes] Favours [Non diabetes!
`
`Fig. S. Forest plot of meta-analysis of pinch strength of the non-dominant hand comparing T2DM and healthy controls (negative mean difference indicates higher mean
`values in controls ).
`
`Study or Subgroup
`Cederlund RI et al 2009
`Yang CJ etal 2017
`
`Diabetes
`Mean
`SO Total
`10 4.44
`23
`12.1
`2.1
`26
`
`Non diabetes
`SO Total Weight
`Mean
`10 7.41
`35 67.6%
`9 32.4%
`14.2
`18
`
`Std. Mean Difference
`IV, Random, 95% Cl
`0.00 [-0.53, 0.53]
`-0 .23 [-0.99, 0.53]
`
`Total (95% Cl)
`49
`Heterogeneity: Tau•: o.oo: Chi'= 0.23, df: 1 (P = 0.63); I'= 0%
`Testfor overall etrect: Z= 0.33 (P = 0.74)
`
`44 100.0%
`
`.0.07 [.0.51, 0.36]
`
`Std. Mean Difference
`IV, Random, 95% Cl
`
`2
`1
`0
`-1
`-2
`Favours [Diabetes] Favours [Non diabetes)
`
`Fig. 6. Forest plot of meta-analysis of hand dexterity of the dominant hand comparing type 2 diabetes mellitus (T2DM ) and healthy controls (poor negative mean difference
`favours neither group).
`
`Study or Subgroup
`Cederlund RI et al 2009
`Yang CJ etal 2017
`
`Diabetes
`Mean
`SD Total
`10 5.19
`23
`11 .5
`2.3
`92
`
`Non diabetes
`Mean
`SD Total
`11 10.37
`35
`9
`13.9
`2.5
`
`Std. Mean Difference
`IV, Random, 95% Cl
`-0.11 1-0.64, o.411
`-1.03 [-1.73,-0.33)
`
`Weioht
`53.3%
`46.7%
`
`Total (95% Cl)
`115
`Heterogeneity: Tau•= 0.32; Chi'= 4.19, df= 1 (P = 0.04); I'= 76%
`Test for overall effect: Z : 1.18 (P = 0. 24)
`
`44 100.0%
`
`.0.54 [-1.43, 0.35]
`
`Std. Mean Difference
`
`4
`2
`0
`-2
`-4
`Favours [Diabetes! Favours [Non diabetes)
`
`Fig. 7. Forest plot of meta-ana lysis of hand dexterity of the non-dominant hand compari ng type 2 diabetes mellitus (T2DM ) and healthy controls (negative mean difference
`indicates h igher mean val ues in controls).
`
`Sanofi Exhibit 2178.004
`Mylan v. Sanofi
`IPR2018-01676
`
`

`

`S. Gundmi et al./ Annals of Physical and Rehabilirarion Medicine 61 (2018) 99-104
`
`103
`
`0
`
`0
`50
`
`0
`
`55
`
`60
`
`age
`
`65
`
`70
`
`75
`
`Fig. 8. Meta-regressi on of effect of mean age on grip strength of the dominant hand.
`
`4.5. Meta-analysis of grip strength in the dominant hand
`
`We performed a meta-regression of grip strength in the
`dominant hand with effect estimates as the dependent variable
`and mean age as a covariate. Mean age had no effect on the effect
`estimates (coefficient= 0.09, P = 0.98, 12 residuals= 0%) (Fig. 8 ),
`which suggests no effect of age on grip strength measurement in
`studies. Sub-group analysis of the effect of gender on grip strength
`of the dominant hand revealed minimal effect of gender (SMD:
`- 1.03; 95% Cl: - 2.24 to 0.18 ) (Fig. 9 ).
`
`5. Discussion
`
`To summarise the result of relevant studies from different geo(cid:173)
`graphical areas, our meta-analysis included 7 studies reporting grip
`and pinch strength as well as hand dexterity and function for people
`with T2DM versus healthy controls (9- 15). Our results suggest
`inconsistency and variability among studies in results. We found weak
`agreement among the studies in rep01ting our outcomes of interest.
`To the best of our knowledge, this is the first meta -analysis
`reporting on hand function with T2DM. We analysed 3 domains -
`grip and pinch strength and hand dexterity - which ultimately
`affect hand function, quality of life, and activities of daily living in
`people with T2DM.
`Our meta-analysis revealed a weak negative mean difference in
`grip and pinch strength between people with T2DM and healthy
`controls. Several factors affect hand strength measurements:
`gender, level of physical activity, age, hand dominance, type of
`
`occupation, body composition and position of the extremity
`[ 11 ]. Previous studies reported that stiff subcutaneous tissue might
`affect strength measurement (14). However, all the confounding
`variables were not considered by the studies when assessing
`strength, which might have in fluenced the outcome measures and
`led to varied results.
`A number of investigations have involved peripheral neuropathy
`in lower extremities and its effect on strength. However, few studies
`have investigated the relation between neuropathy of upper
`extremities and its effect on strength. Cetinus et al. suggested that
`low hand strength with T2DM might be due to the severity of the
`neuropathy, although the authors did not assess this in their study
`[14, 16]. Studies suggested that stiffness of the subcutaneous tissue
`in the diabetic hand might lead to decreased strength (16).
`Dexterity assessed by the Purdue Pegboard test evaluates the
`gross movement of the fingers, hands, and fine fingertip dexterity
`necessary in assembling the task. Our meta-analysis of 2 studies of
`dexterity assessed by the Purdue Pegboard test showed no
`significant difference in the combined mean for the dominant hand
`[12,14], which suggests that people with T2DM have good coping
`strategies to compensate for altered gross motor activity of fingers
`and hand and fine motor activities of the fingers in the dominant
`hand. Meta-analysis of the non-dominant hand revealed a negative
`combined mean, which suggests a poor compensatory mechanism of
`the non-dominant hand. However, both the studies did not consider
`confounding factors such as visual co-ordination and involvement of
`basal ganglia, which might have influenced the dexterity.
`Hand function was assessed in 2 studies with 2 validated
`questionnaires [ 12,15 ]. The MD suggested significantly reduced
`hand function with T2DM. People with T2DM show various
`disorders of the hand and fingers, which may affect activities of
`daily living. A study evaluating hand functioning in patients with
`diabetes and its impact on quality of life in the physical and mental
`dimensions suggested that impaired hand function leads to lower
`acceptance of the disease, depression and red uced quality of life [ 1 7 ].
`We performed a meta- regression of age and gender on grip
`strength of the dominant hand to analyse the heterogeneity among
`the studies, which could have influenced the outcome of interest.
`Mean age and gender had no influence on grip strength. However,
`meta-regression of duration of diabetes could not be performed
`because 2 studies did not report this information.
`The limitation of the current analysis was that none of the studies
`reported a sample size calculation, which might have affected the
`
`Non diabetes
`Diabetes
`Moan
`SD Total Mean
`SD Total Wolght
`Stud or Sub roup
`Both male and female
`100 20.4%
`100 28.3 25.93
`Carvalho F Do otal 2004
`47
`20.3%
`76 33.22 10.53
`Cenlinus Eel at 2005
`50
`20.0%
`50
`32.9
`7,6
`Kaur Pct al 2016
`Subtotal (95% Ct)
`197
`60.8%
`226
`Heterogeneity: Tau'= 1.36; Chi'= 78.78, df = 2 (P < 0.00001 ); I'= 97%
`Test for overall effect: Z = 0.93 (P = 0.35)
`
`38.3 12.29
`28.92 10.86
`20.76 3.55
`
`Std. Mean Difference
`IV, Random, 95% Cl
`
`0.49 (0.21, 0.77]
`-0.40 [-0.77, -0.03]
`-2.03 [-2 .52, -1.55]
`-0.63 [-1.97, D.70 I
`
`Std. Mean Difference
`IV, Random, 95% Cl
`
`...
`
`----
`
`Male only
`Cede~und RI el al 2009
`Subtotal (95% Cl)
`Heterogeneity: Not applicable
`Test for overall effect: Z = 0.38 (P = 0.71)
`
`41
`
`23.7
`
`23
`23
`
`39 16.29
`
`35
`35
`
`19.9%
`19.9%
`
`0.1 o [-0.43, o.63]
`0.10 [-0.43, o.63I
`
`Female only
`Mohamed EK el al 2014
`Subtotal (95% Cl)
`Helerogeneily: Not applicable
`Tesl for overall etfecl: Z = 9.61 (P < 0.00001)
`
`24.9 5.63
`
`40 42.75
`40
`
`4.65
`
`40
`40
`
`19.3%
`19.3%
`
`-3.42 [-4.12, -2.73]
`-3.42 [-4.12, -2.73)
`
`----~
`
`272
`289
`Total (95% Cl)
`100.0%
`Heterogeneity: Tau'= 1.85; Chi'= 157.22,df =4 (P < 0.00001); I'= 97%
`Tesl for overall effecl: Z = 1.67 (P = 0.10)
`Test for suborouo differences: Chi'= 62.94. df = 2 (P < 0.00001l. I'= 96.8%
`
`-1.03 [-2.24, 0,181
`
`4
`2
`0
`4
`-4
`Favours [Diabetes] Favours [Non diabetes]
`
`Fig. 9. Meta-analysis of effect of gender on grip strength of the dominant hand.
`
`Sanofi Exhibit 2178.005
`Mylan v. Sanofi
`IPR2018-01676
`
`

`

`104
`
`S. Gundmi er al. / Annals of Physical and Rehabilitation Medicine 61 (2018) 99- 104
`
`difference in hand function between the study and control groups. As
`well, confounding factors (including level of physical activity, hand
`dominance, type of occupation, body composition, position of the
`extremity), which could have influenced the outcome of interest,
`were not considered in the studies.
`
`6. Conclusion
`
`This systematic review and meta-analysis revealed a negative
`mean differe nce in grip and pinch strength as well as hand function
`and dexterity between people with T2DM and healthy controls.
`However, because of low sample size and heterogeneity among the
`studies, our study revealed no significant effect for the outcome
`measures. Hence, we see a strong need for studies of larger sample
`size, which m ight provide better ev idence. Studies should control
`for confounding variables that might affect hand activity also. As
`for foot evaluation in T2DM, hand evaluation should be incorpo(cid:173)
`rated in healthcare providers' routine assessment. Future studies
`should measure the effect of vaiious treatments for better hand
`function in people with T2DM.
`
`Author contributions
`
`Conception and design: G.S., A.M.; search strategy and data
`analysis: R.N. ; data analysis and interpretation of data: G.S., A.M.,
`A.B.K., M.H.H., R.K.V.; editing: A.B.K.; statistical interpretation: G.S.,
`A.M.,R.N.
`
`Disclosure of interest
`
`The authors declare that they have no competing interest.
`
`Acknowledgments
`
`We thank the World Diabetes Foundation WDF 15-941 for
`supporting the study and Mani pal University for their approval.
`
`References
`
`I accessed
`
`I 11 Seshaih V. Hand book on diabetes mell itus. 6th ed. All Ind ia publishers and
`distributors: New Delhi. 2013
`121 l.D.F .. Diabetes atlas. 7th edition. 2015 (accessed 15 Dec 2017 lhttps:/fwww.
`idf.org/e-1 i brary / epidemiology-research/ diabetes-atlas/
`13-diabetes-atlas-seventh-edition.html.
`131 Global report on diabetes. World Health Organization. http://www.apps.who.
`int/iris /bitstream/10665/204871 /1 /9789241565257 _eng.
`pdf?ua=l&utm_source=blog&utm_campaign=rc_blogpost. 201 6
`15 Dec 2017).
`14] Soumya D, Srilatha B. Late stage complications of diabetes and insulin resis(cid:173)
`tance. J Diabetes Metab 2011:2:167.
`IS] Marchettini P. Lacerenza M, Mauri E. Marangoni C. Painful peripheral neurop(cid:173)
`athies. Curr neuropharmacol 2006;4:175- 81.
`161 Singh VP, Bali A. Singh N. Jaggi AS. Advanced glycation end products and
`diabetic compl ications. Korean J Physiol Pharmacol 2014; 18: 1- 14.
`17] Pandey A, Usman K Reddy H. Gutch M, Qidwai S. Prevalence of hand disorders
`in type 2 diabetes mellitus and its correlation with microvascular complica(cid:173)
`tions .. Ann Med Health Sci Res 2013;3 :349-54.
`18] Casanova JE, Casanova JS, Young MJ. Hand function in patients with diabetes
`mellitus. South MedJ 1991;84:llll - 3.
`19] Kaur P, Gill S, Sandhu R. Arora AK_ Hand grip strength in type 2 diabetics and
`non-diabetics. Pak J Physiol 2015: 11 :32-4.
`11 OJ De Caivalho F. Jakimiu FO, Skare TL. Diabetic hands: a study on strength and
`function. Diabetes Metab Syndr 2014;8:162-5.
`I 11 J Ce tin us E. Buyukbese MA, Uzel M. Ekerbicer H. Karaoguz A. Hand grip strength
`in patients with type 2 diabetes mell itus. Diabetes Res Cli Pract 2005;70:
`278-86.
`11 2] Yang CJ. Hsu HY, Lu CH, Chao YL. Chiu HY, Kuo LC. Do we underestimate
`influences of diabetic mononeuropathy or polyneuropathy on hand functional
`performance and life quality? J Diabetes fnvestig 2017:1-7_
`I 13 J Mohamed EK Mana! I. Al-rashidi. Effect of longstanding diabetes mellitus type
`II on handgrip strength and pinch power of females in the city of Hail-KSA.
`IOSR J Nurs Heal Sci 2014;3 :41 - 4.
`I 14] Cederlund RI. Thomsen N. Thrainsdottir S. Eriksson KF, SundkvistG, Dahlin LB.
`Hand disorders, hand Function. and activities of daily living in elderly men
`with type 2 diabetes. J Diabetes Complications 2009;23:32- 9_
`I 15] Sava~ S. Kiiroglu BK Koyuncuoglu HR, Uzar E, Celik H, Tamer NM. The effects of
`the diabetes related soft tissue hand lesions and the reduced hand strength on
`functional disability of hand in type 2 diabetic patients. Diabetes Res Clin Pract
`2007:77:77- 83.
`11 6] Lundbaek K. Stiff hands in long-term diabetes. Acta Med Scand 1957;158:
`447-51.
`I 17] Lewko J. Kochanowicz J, Zarzycki W, Mariak Z, Gorska M, Krajewska-Kulak E.
`Poor hand function in diabetics. Its causes and effects on quality of life. Saudi
`Med J 2012:33:429-35.
`
`Sanofi Exhibit 2178.006
`Mylan v. Sanofi
`IPR2018-01676
`
`