Key words frozen shoulder - intra-articular injection - NSAID
Introduction
Adhesive capsulitis is a relatively common musculoskeletal complaint in outpatient
that is due to soft tissue involvement of glenohumeral joint and is more prevalent
among women more than 50 years [1 ]
[2 ]. Pain and restricted active and passive movement of shoulder are the most common
clinical presentations [1 ].
Pain and stiffness of shoulder joint will appear within few months to one year; but
the course in some patients may progress faster [1 ]. Frozen shoulder could be idiopathic or due to some predisposing factors like diabetes,
inactivity, previous disorders of shoulder, cervical spondylitis, coronary artery
diseases, pulmonary tuberculosis, chronic obstructive pulmonary disease (COPD), bronchial
carcinoma, hyperthyroidism, hemiplegia, brain tumors, epilepsy and Parkinson disease
[2 ]
[3 ].
Diagnosis is by history of pain and reduced range of motion and in physical examination
with active and passive restricted range of motion in all directions [4 ]. Definitive diagnosis is made by artrography, that only amount of radiopaque solution
fewer than 15 ML can be injected into the joint [5 ]. Symptoms will be resolved spontaneously within 1-3 years in most of patients, but
some degree of joint movement restriction will remain [5 ].
Prevalence of adhesive capsulitis in normal population is about 2% that in diabetic
patients increases to 10–20%. The prevalence is related to the duration of diabetes.
The mean age of capsulitis in diabetic patients is lower than general population;
also duration of disease is longer and response to treatment is less. Bilateral involvement
is more seen in diabetics too [6 ].
Capsulitis ordinary happens in 4th –6th decades of life. Disease will start silently and has 3 phases:
First phase is named ‘Painful Phase’ that is accompanied with pain and restricted
range of motion. This phase lasts about 2–9 months. The second phase is known as ‘Freezing
(Adhesive) Phase’ that lasts about 3–9 months. During second phase fibrosis is formed
and pain decreases while range of motion is reduced more. Third phase named ‘Throwing
(Resolution) Phase’, which pain is subsided and also the lost motions are resolved.
This phase lasts about 12–18 months [5 ].
A number of treatments have been advocated such as rest, physiotherapy, analgesia,
acupuncture, active and passive mobilisation, oral and injected corticosteroids, capsular
distension, manipulation under anaesthesia and surgical capsular release. It is surprising
that for such a common condition there is no consensus on the most effective treatment
[3 ]
[7 ].
Because capsulitis is one of the most prevalent complains in diabetic patients and
can effect on performance and quality of life, and also because there is no definitive
treatment of it, we designed this study to compare the efficacy of 2 most popular
treatment modalities with together.
Materials and Methods
The randomized clinical trial study was conducted on diabetic patients with adhesive
shoulder capsulitis. Based on previous studies and following parameters: α=0.05, β=0.2
and d=10, about 30 cases were needed for each group.
Data were registered in a questionnaire which was included in questions about sex,
age, time of diabetes diagnosis, drug history, results of shoulder clinical examination
with goniometer and pain score based on VAS (visual analogue score). Patients were
chosen from all diabetic patients who were admitted to rheumatology clinic. Diagnosis
was confirmed by an Internal Medicine resident with clinical examination and by detection
of pain and limited range of motion in shoulder joint (Glenohumeral joint). Because
there is different response to treatment between acute and chronic form of disease,
patients with more than 6 months of disease were excluded from the study. Also patients
with other proved causes of capsulitis like degenerative diseases, infection, fractures
and stroke were excluded from the study. Other exclusion criteria’s were active peptic
ulcer, history of GI bleeding, history of coagulopathies and renal failure. After
that screened patients were examined by a rheumatologist. To rule out other causes
of disease and checking for possible confirmatory evidence of capsulitis like osteopenia,
X-ray was taken for each patient. Also for diagnosis of possible secondary causes
and as basic laboratory tests, CBC (Complete Blood Count), ESR (Erythrocytes Sedimentation
Rate), CRP (C Reactive Protein), urea, creatinine, liver enzymes and U/A (Urine Analysis)
were checked for all patients.
The researcher explained about all treatment modalities for patients and written consent
was obtained from each participants. Patients were divided into 2 groups using table
of random numbers. First group got NSAID as treatment while the second group underwent
intra-articular injection of corticosteroids under sonography guide. Before treatment,
patients were evaluated for severity of pain and extent of restricted range of motion.
Pain was assessed according to VAS (Visual Analogue Score) by using Pain Assessment
Ruler (PAR). In this study pain is graded between 0–10 levels (no pain to severe pain).
Evaluation in this scaling is subjective and was done by patients and noted in questionnaire
by researcher at each visit. Range of motion also was evaluated in 3 directions of
flexion, abduction and external rotation by goniometer. Also internal rotation evaluated
by checking the ability of patients to reach dorsum of their hand to back and checking
the highest point possible on their inferior border of scapula that assessed by plus
(0–4) and noted at each visit.
Considering with different response to different types of NSAIDs, 500 mg Naproxen
twice daily was administered for all patients. Anti inflammatory dose of Naproxen
reaches by using 2 tablets daily and then its use is more convenient for the patient.
In second group patients underwent single injection of 40 mg triamcinolone at the
start of study. The 24 gauge needle was injected between medial head of humerus and
1 centimeter’ of lateral head of coracoid process and then moved directly to the posterior,
superior and lateral position. To ensure the direct injection into joint capsule,
injections were ultrasound-guided. After 1 week, all patients started home exercise
with moving shoulder in 3 directions of flexion (maximum range of 180°), abduction
(maximum range of 180°) and taking back by hand and maximum reaching digits to inferior
border of scapula. This exercise recommended 3 sessions a day and 15 times at each
session. Patients were evaluated at 2nd , 6th , 12th and 24th weeks. All registered data were transformed into SPSS-15 program and analyzed by
repeated measures, T-test and Chi-square tests.
Results
Totally 75 patients were visited. 11 patients of injection group and 7 patients of
naproxen group were excluded because discontinuing the follow-up process and finally
57 patients were included in the analysis (28 in naproxen group and 29 in injection
group), flow diagram of the study is indicated in the [Fig. 1 ]. In naproxen group, 17 (60.7%) patients were female and 11(39.3%) patients were
male while in injection group 21 (72.4%) and 8 (27.6%) were female and male respectively
(P=0.4; Chi-square). Mean age of participants were 52.78±6.72 and 55.31±7.7 years
for naproxen and injection groups respectively (P=0.19; T-test). Also there was no
significant difference between groups according to HbA1C (0.25; Chi-square) and duration
of diabetes (P=0.9; T-test) (9.3±7 years in naproxen group vs. 9.5±5.8 in injection
group).
Fig. 1 Flow diagram of the study.
Mean range of motion (flexion, abduction and external rotation) were increased in
fifth visit comparing with second visit significantly (P=0.001; paired T-test). Also
mean of internal rotation improved and pain score was decreased (P=0.001; paired T-test)
([Table 1 ]). For injection group also similar results were obtained (P=0.001; paired T-test)
([Table 2 ]).
Table 1 Mean±SD of variables at first and fifth visits and P-Values in Naproxen group (Paired
T-test).
Variable
Visit
Mean±SD
P-Value
flexion (degree)
first visit
107.6±15.7
0.001
fifth visit
167.6±22
abduction (degree)
first visit
99.2±22.6
0.001
fifth visit
170±22.9
external rotation (degree)
first visit
28.2±9.5
0.001
fifth visit
45.7±9.8
internal rotation (+ )
first visit
2.6±0.87
0.001
fifth visit
0.32±0.54
pain score
first visit
5.64±2.43
0.001
fifth visit
1.99±1.98
Table 2 Mean±SD of variables at first and fifth visits and P-Values in Triamcinolone group
(Paired T-test).
Variable
Visit
Mean±SD
P-Value
flexion (degree)
first visit
103.7.6±22.3
0.001
fifth visit
167.4±24.2
abduction (degree)
first visit
90.6±21.3
0.001
fifth visit
172.9±21.6
external rotation (degree)
first visit
30.8±11.7
0.001
fifth visit
47.4±11.4
internal rotation (+ )
first visit
2.5±1.01
0.001
fifth visit
0.24±0.43
pain score
first visit
6.18±2.17
0.001
fifth visit
2.24±2.06
There were no significant difference between groups According to flexion (P=0.51;
Repeated Measure), Abduction (P=0.76; Repeated Measure), external rotation (P=0.12;
Repeated Measure), internal rotation and also pain score (P=0.91 and P=0.90 respectively;
repeated measure) ([Table 3 ]).
Table 3 Mean±SD of variables at different visits in 2 groups (repeated measures).
Variable
Flexion
Abduction
External Rotation
Internal rotation
Pain score
first visit
naproxen
107.6±15.7
99.2±22.6
28.2±9.5
2.6±0.87
5.64±2.43
triamcinolone
103.7±22.3
90.6±21.3
30.8±11.7
2.5±1.01
6.18±2.17
second visit
naproxen
145.3±30.3
141.7±39.4
37.5±11.8
1.7±1.1
4.35±2.54
triamcinolone
156.8±29.3
141.5±39.6
41.2±11.3
1.7±1.03
3.93±2.47
third visit
naproxen
156.9±28.3
145.5±40.2
37.6±8.2
1.03±1.07
3.07±2.18
triamcinolone
159.8±27.7
155.9±37.9
42.2±11.6
1.2±0.77
2.87±2.21
fourth visit
naproxen
157.3±28.1
154.4±37.3
41.2±9.8
0.6±0.78
2.69±2.44
triamcinolone
163.1±27.4
159.4±33.2
44.8±12.3
0.65±0.72
2.77±2.47
fifth visit
naproxen
167.6±22
170±22.9
45.7±9.8
0.32±0.54
1.99±1.98
triamcinolone
167.4±24.2
172.9±21.6
47.4±11.4
0.24±0.43
2.24±2.06
P-Value
0.51
0.76
0.12
0.91
0.91
In comparison of mean flexion and abduction with maximum normal degree (180°), mean
flexion and abduction scores at the fifth visit were closer to maximum score to some
extent that had no significant relation with maximum normal range ([Table 4 ]). For naproxen group there was similar results, range of motion in abduction at
fifth visit was 170°.
Table 4 Comparison of flexion and abduction mean±SD values to maximum normal values (180°)
(one sample T-test).
Group
Variable
Mean
SD
P-Value
naproxen
flexion 3
156.9
28.3
0.001
triamcinolone
159.8
27.7
0.01
naproxen
flexion 4
157.3
28.1
0.001
triamcinolone
163.1
27.4
0.03
naproxen
flexion 5
167.6
22
0.06
triamcinolone
167.4
24.2
0.09
naproxen
abduction 3
145.5
40.2
0.001
triamcinolone
155.9
37.9
0.02
naproxen
abduction 4
154.4
37.3
0.01
triamcinolone
159.4
33.2
0.02
naproxen
abduction 5
170
22.9
0.29
triamcinolone
172.9
21.6
0.90
Discussion
Our study compared intra-articular injection of triamcinolone with NSAID (naproxen)
in diabetic frozen shoulder patients. This study was done only in diabetic patients,
and adhesive capsulitis due to any other underlying causes were excluded from the
study. Range of motion was detected precisely by goniometer. Also to guarantee the
maximum effect of treatment, injections were done under sonography guide and also
patient in NSAID group were requested to give the remnant drugs back to researcher.
Patients were followed for 6 months and evaluated within 5 visits.
After 6 months of follow-up, we did not find any significant different between 2 groups
according to flexion, abduction, external rotation, internal rotation and also pain
score. Range of motion in patients of both groups almost returned to normal range.
There is only one study which compared intra-articular corticosteroid with oral NSAID
in patients with adhesive capsulitis [8 ], but to our knowledge there is no study to compare these treatments in diabetic
patients specifically.
Arslan et al. in 2001 [8 ], studied on effect of corticosteroid, physiotherapy and NSAID on adhesive capsulitis.
10 men and 10 women allocated into 2 groups. Group A underwent 40 mg intra-articular
methylprednisolone while in group B physiotherapy and NSAID was administered. Results
showed that at the end of 12th week improvement in active and passive range of motion and pain score were similar
between groups. Sample size in this study was lower than ours and was not limited
to diabetic patients.
In a study by Buchbinder et al., effect of oral corticosteroids was compared to placebo.
They pointed out that a 3 week course of prednisolone 30 mg daily in patients with
adhesive capsulitis is better than placebo to improve pain, function, and range of
motion [9 ].
In 2007, Russel et al. compared prednisolone and triamcinolone in painful shoulder.
After 2 weeks of follow-up, improvement in pain and range of motion was seen in 92%
of patients got prednisolone and 50% of patients got triamcinolone. Patients under
treatment of prednisolone had faster recovery [10 ]. In these study patients had painful shoulder with any reason and duration of follow-up
was only 2 weeks. In 2008 Isar Ahmad [11 ], compared these 2 drugs in adhesive capsulitis, but did not showed any difference
between 2 groups. Also this study suggested that triamcinolone had better results
in diabetic patients. This study compared diabetics and non diabetics, while our study
has focused on diabetic patients only.
In 2011, Roh et al. examined the efficacy of corticosteroid injections for the treatment
of adhesive capsulitis in patients with diabetes mellitus. A group of patients were
undergone injection and home exercise and another group only did home exercise. In
conclusion authors resulted that a corticosteroid injection in diabetic patients decreases
the pain perception and accelerates the functional recovery in the early post-injection
period [12 ]. This article did not compare the effect of corticosteroids with any oral drugs.
Smith et al. in 2005, compared intra articular injection of triamcinolone under fluoroscopy
guide following with 12 sessions of physiotherapy with triamcinolone only in adhesive
capsulitis patients. Results showed that corticosteroid with physiotherapy have better
results than using only corticosteroids [13 ].
Widiastuti-Samekto et al. in 2010 compared injection of corticosteroid with oral corticosteroid
in 26 patients with adhesive capsulitis. Based on results cure rate of injection group
was 5.8 times more than oral group and after a week 62% of patients in injection group
remitted while only 14% in oral group had remission. Compared to our study, this study
had smaller sample size and follow-up period was shorter than our study too [14 ].
Sakeni et al. in Turkey surveyed on effect of corticosteroid injection accompanied
with exercise at home. Based on this study intra-articular corticosteroid has additive
effect to exercise in acceleration of remission especially during first week. There
was no comparison in this study with other treatment modality [15 ].
A systematic review was done by Bruce Arroll [16 ] to determine improvement of symptoms of intra-articular and subacromial injections
of corticosteroid for rotator cuff tendonitis and frozen shoulder. 7 articles that
compared corticosteroid vs. placebo and 3 articles compared corticosteroid vs. NSAID
were included in the study. The results indicated that sub acromial injection of corticosteroid
is suitable for improvement of tendonitis and possibly is more appropriated than NSAID.
But there is lack of evidence for determination of intra-articular injection of corticosteroids
in adhesive capsulitis.
A valuable Meta analysis in 2012 by Maund et al., resulted that there may be benefit
from adding a single intra-articular steroid injection to home exercise in patients
with frozen shoulder of less than 6 months duration. This study also reported contradictory
results from some other studies in this context and concluded that there is limited
clinical evidence on the effectiveness of treatments for primary frozen shoulder [17 ].
In our study we didn’t find any difference between intra-articular injection and NSAID.
Because diabetic patients had simultaneous complications such as nephropathy or hypertension
and considering with potential side effects of NSAIDs like gastrointestinal bleeding
and also the use of aspirin at same time by many diabetic patients that can intensify
the side effects of NSAIDs, it seems that administration of 1 injection of triamcinolone
had equal treatment effects with less side effects and can be suggested as the method
of choice in diabetic patients.
Conclusion
Based on our study, intra articular corticosteroid and NSAID are effective in treatment
of adhesive capsulitis both and there is no significant difference between these 2
treatment modalities in diabetic patients. Based on our knowledge our study was the
first study on comparison of corticosteroid and NSAID in frozen shoulder of diabetic
patients. Because diabetic patients have other condition such as hypertension or nephropathy,
1 intra-articular injection may be more appropriated in comparison to receiving NSAID
for 1 month.
