Limited Joint Mobility (Diabetic Cheiroarthropathy)

Hand stiffness as a complication of DM was first described in 1957 by Lundbaek in 5 adults with insulin-dependent DM and subsequently in 1974 by Rosenbloom in 3 adolescents with type 1 DM. The term cheiroarthropathy (derived from the Greek word cheiros , meaning hand) also has been used to describe this condition; however, use of the term arthropathy is inaccurate, because the pathologic process involves periarticular structures, not the joint itself.

Rosenbloom and colleagues described 3 adolescents with type 1 DM who had flexion contractures involving several fingers and thickened, tight, and waxy skin predominantly over the dorsal aspect of the metacarpophalangeal, proximal, and distal interphalangeal joints of their fingers. They also had limited range of motion of large joints, such as the wrists, elbows, knees, and ankles, and of the spine. Also, these patients had short stature, delayed sexual maturation, and retinopathy or nephropathy, the diabetic microvascular complications of DM. Typically, the skin changes of LJM begin around the metacarpophalangeal and proximal interphalangeal joints of the little finger and progress medially to involve other digits and distally to involve the distal interphalangeal joints. LJM usually is painless and only mildly disabling.

The clinical presentation of stiff hands in the adult diabetic patients described by Lundbaek differed from that in the adolescents described by Rosenbloom. The adult diabetics experienced paresthesias in their hands, followed by incapacitating pain that was exacerbated by activity. These patients also had stiff, hardened skin on their fingers and on the palms of their hands. A characteristic feature among these adults was the presence of arterial calcifications on plain radiographs. Histologic examination of skin demonstrated a paucity of elastic fibers. Rosenbloom proposed that this clinical entity is distinct from that which he observed in adolescents, and he named it the stiff hand syndrome .

The prevalence of LJM ranges between 30% and 58% among patients with type 1 DM and between 45% and 76% among those with type 2 DM, as compared with between only 4% and 20% among individuals without DM. The thickened, tightened, and waxy quality of skin on the fingers, the dorsum of the hands, and the forearms has been observed without LJM being present in 34% of diabetic patients.

LJM is diagnosed based on the presence of characteristic findings on physical examination. The inability to appose the palmar surfaces of the hands and fingers with the wrists dorsiflexed is called the prayer sign . An alternative method of testing for LJM, called the table top sign , is considered to have an abnormal result if the entire surface of the palm and fingers cannot make contact with a flat surface when patients lay their palms on a tabletop with the fingers spread apart. Limited mobility should be confirmed by demonstrating loss of passive extension of the proximal interphalangeal and metacarpophalangeal joints (<180° and 60°, respectively). Rosenbloom and colleagues proposed a classification system to stage this condition based on the type and number of joints involved ( Table 1 ).

Several cross-sectional studies have correlated the presence of LJM with age and duration of DM. Patients with type 1 DM and LJM were observed to have a higher incidence of retinopathy and nephropathy than those without LJM, even when adjusted for the confounding effect of duration of DM. Type 2 diabetics with LJM also had a higher incidence of retinopathy and nephropathy and a greater requirement for insulin therapy, but no difference in hemoglobin A _1c (HbA _1c ) levels, compared with those without LJM. However, subsequent prospective studies failed to identify LJM as a predictor of the development of diabetic microvascular complications or the presence of microvascular complications as a predictor of developing cheiroarthropathy. In contrast, the risk of developing macrovascular complications was increased among patients with type 2 DM and LJM, with 3- and 4-fold increased relative risks of developing coronary artery disease and cerebrovascular disease, respectively, even after controlling for age, duration of DM, and glycemic control.

Inadequate glycemic control influences the progression of LJM among patients with type 1 DM. Higher levels of HbA _1c were observed among patients with type 2 DM and LJM or shoulder adhesive capsulitis, suggesting that poor glycemic control might also contribute to developing these musculoskeletal conditions in type 2 DM. However, other studies have failed to demonstrate an association between glycemic control and the presence of LJM. The onset of hyperglycemia has been estimated to occur at least 4 to 7 years before type 2 DM is diagnosed; diabetic complications, such as retinopathy, already may be present when the clinical diagnosis is made. Thus, longstanding hyperglycemia before the diagnosis of DM could also predispose to developing LJM. HbA _1c determinations from the period after initiation of glucose-lowering therapy might not reflect earlier prolonged hyperglycemia, especially among patients with type 2 DM, which could account for the apparent lack of correlation between laboratory measures of glycemic control and musculoskeletal manifestations of DM.

An observational study demonstrated a 4-fold decrease in the prevalence of LJM among children with type 1 DM, comparing the period between 1976 and 1978 with 1998. Another study assessing the prevalence of LJM among adults with type 1 DM attending an endocrinology clinic in the United Kingdom also showed a statistically significant decrease in the prevalence of LJM from 43% during the period between 1981 and 1982 to 23% in 2002. The investigators hypothesized that implementation of an intensive glycemic control regimen accounted for the decreased prevalence of LJM. Although HbA _1c levels did not correlate with the presence of LJM, there was a trend toward a lower prevalence of LJM among patients with HbA _1c level less than 7%.

Nonsteroidal anti-inflammatory drug and physical therapy remain the primary therapeutic modalities for patients with LJM. Increased range of motion and improved strength of finger flexion and extension has been reported in 3 diabetic patients with LJM who were treated with the hydantoin aldose reductase inhibitor, sorbinil ; however, clinical studies of sorbinil were stopped because of significant adverse effects, including adult respiratory distress syndrome, epidermic necrolysis, and Stevens-Johnson syndrome.

Stenosing Flexor Tenosynovitis (Trigger Finger)

Stenosing flexor tenosynovitis typically presents with locking (or “triggering”) of fingers in flexion, extension, or both, most commonly involving the thumb, middle, and ring fingers. In this condition, fibrosis and thickening of the tendon sheath, usually where it passes through a pulley or over a bony prominence, restricts movement of the flexor tendon within its tendon sheath. Tendon swelling distal to the site of constriction causes pain and difficulty flexing and extending the digit smoothly as the swollen segment of tendon passes through the narrowed space. Typically, this is followed by more severe symptoms of catching and, eventually, locking of involved fingers in flexion or extension.

The prevalence of stenosing flexor tenosynovitis ranges between 5% and 36% among patients with type 1 and 2 DM, as compared with 2% in the general population. The occurrence of trigger finger correlates significantly with duration of DM but not with glycemic control. Compared with nondiabetics, patients with DM are more likely to have multiple fingers involved simultaneously by stenosing flexor tenosynovitis.

Treatment of stenosing flexor tenosynovitis includes modification of activities to avoid triggering of digits, nonsteroidal anti-inflammatory drug therapy, splinting, corticosteroid injection into the tendon sheath, and surgical release. Corticosteroid injection is less successful among patients with type 1 or 2 DM than with nondiabetics; systemic absorption of the depot corticosteroid occasionally can perturb glycemic control. Patients with multiple digit involvement more often require surgical release to relieve symptoms.

Dupuytren’s Disease

DD is characterized by thickening of the palmar fascia, palmar or digital nodules, skin tethering, pretendinous bands, and flexion contractures of the fingers. Its presentation among patients with DM differs slightly from that in the general population. Among diabetics, DD most often involves the middle and ring fingers rather than the ring and little fingers, as in nondiabetics, and less often results in palmar contractures.

DD is more prevalent among diabetics, affecting between 16% and 42%, than in the general population. Among diabetics, DD occurs equally in both genders, whereas it occurs more often among men in the non-DM population. However, among diabetics, severe DD and development of finger contractures is seen more often in men than women. Because DD occurring in diabetic patients seldom requires surgical treatment, no significant association between DD and DM was identified in a large retrospective study of 2919 patients who had been referred for surgical treatment.

Similar to LJM, the prevalence of DD among diabetics is higher with older age and longer duration of DM. In cross-sectional studies, DD has been associated independently of age or duration of DM with the presence of retinopathy among patients with type 1 DM and of severe retinopathy among patients with type 2 DM. In a prospective study, the incidence of DD was higher among diabetic patients with hypertension and the microangiopathic complications of retinopathy, nephropathy, and neuropathy. However, when logistic regression analysis controlled for the confounding effects of age and duration of DM, the presence of DD no longer predicted the development of hypertension or of diabetic microangiopathic complications.

Intralesional corticosteroid injections and surgery followed by hand therapy have been the standard treatments for DD. Recently, injecting the thickened palmar fascia with collagenase from Clostridium histolyticum has been developed as a nonsurgical approach to treating DD. In a prospective, randomized, placebo-controlled trial of 308 patients with DD, 6.5% of whom had DM, up to 3 collagenase injections significantly reduced fixed flexion contractures and significantly improved finger-joint range of motion; there were no recurrences of DD after up to 90 days of follow-up. Only 3 serious adverse events were reported in this study, including 2 tendon ruptures and 1 case of complex regional pain syndrome. Although expensive, collagenase injection provides an alternative to surgical treatment of DD.

Adhesive Capsulitis of the Shoulder (Shoulder Periarthritis, Frozen Shoulder Syndrome)

Adhesive capsulitis usually presents as painful progressive restriction of range of shoulder motion, especially on abduction and external rotation. Its natural history can be divided into 3 phases: pain, stiffness, and recovery. The length of the recovery phase depends on the duration of the stiffness phase, with symptoms lasting for an average of 30 months. The term frozen shoulder was first used to describe this condition in 1934 by Codman. In 1945, Naviaser used the term adhesive capsulitis to describe the same entity.

Adhesive capsulitis of the shoulder occurs more commonly among diabetics. The prevalence of shoulder periarthritis among diabetics ranges from 10% to 29% and is about 5-fold that in the general population. In a cross-sectional study of 294 patients with type 1 DM and 134 patients with type 2 DM, the prevalence of shoulder adhesive capsulitis was 10% and 22%, respectively. Shoulder adhesive capsulitis was more likely to develop in older individuals with either type of DM and in those with longer duration of disease among patients with type 1 DM. The apparent lack of association between shoulder adhesive capsulitis and disease duration in type 2 DM could be explained by the years of hyperglycemia that typically precede the diagnosis of type 2 DM. After controlling for the effects of age and disease duration, shoulder adhesive capsulitis is found to be associated with the presence of autonomic neuropathy among individuals with either type of DM and with myocardial infarction among those with type 1 DM. However, the presence of shoulder adhesive capsulitis did not correlate significantly with HbA _1c levels among patients with either type of DM.

An increased incidence of LJM among diabetic patients with shoulder adhesive capsulitis has been observed in some studies but not confirmed in others. Mean HbA _1c levels may be significantly higher among those diabetic patients with both shoulder and hand involvement than among those with no upper extremity problems, which reflects the documented association between longstanding hyperglycemia and the development of LJM.

In idiopathic adhesive capsulitis, the joint capsule is thickened and adheres to the head of the humerus, thereby reducing the volume of the glenohumeral joint. Histologic and immunocytochemical examination of the joint capsule demonstrates proliferation of fibroblasts and transformation of some to myofibroblasts, which produce excess amounts of types I and III collagen, with subsequent contraction of the joint capsule that results in pain and stiffness. These findings are similar to DD of the hand. In both conditions, expression of fibrogenic growth factors, matrix metalloproteinases (MMPs), and their inhibitors was increased, but expression of the proinflammatory cytokines interleukin (IL)-1β and tumor necrosis factor (TNF)-α and of TNF-β was lower in joint capsule tissue obtained from patients with adhesive capsulitis than in DD. This correlates with the histologic observation of relatively hypocellular, dense fibrous tissue with only a few mature fibroblasts, suggesting fibrosis rather than an active inflammatory process. In adhesive capsulitis, the lower expression of MMP-14, which plays a role in remodeling, could explain slow resolution of fibrosis.

Analgesics, physical therapy, and intra-articular corticosteroid injections are first-line therapy during the initial painful phase of shoulder adhesive capsulitis. In a prospective study of a stretching-exercise program in 75 patients with idiopathic adhesive capsulitis including 8 with DM, 90% were satisfied with this treatment. However, the presence of DM was associated with an inferior outcome as evidenced by more pain, more restricted range of motion, and poorer function. During the adhesive phase, physical therapy and operative treatments are typically used. Manipulation under anesthesia may be complicated by fracture, shoulder dislocation, tendon rupture, or neurologic injury. Arthroscopic capsular release has been an effective treatment modality for refractory shoulder adhesive capsulitis among diabetic patients. In a study comparing arthroscopic capsular release in diabetic and nondiabetic patients, those with DM had poorer recovery of shoulder range-of-motion and function, but there was no significant difference in the duration of recovered range of motion or of complete pain relief. Arthroscopic capsular release is preferred to open surgical release because of the reduced length of postoperative recovery.

Pathogenesis of Limited Joint Mobility, Stenosing Flexor Tenosynovitis, Dupuytren’s Disease, and Shoulder Adhesive Capsulitis

No single mechanism has been shown to account for the development of LJM and shoulder adhesive capsulitis, of stenosing flexor tenosynovitis, and of DD among patients with DM. However, the shared cause of these conditions seems to involve connective tissue deposition around joints, in tendon sheaths, and in the palmar fascia, respectively. Although the pathogenesis of shoulder adhesive capsulitis in patients with DM has not been studied directly, its coexistence with LJM suggests a possible common mechanism. Histologic examination of skin from patients with LJM reveals thickening of the dermis and epidermis, accumulation of collagen and loss of skin appendages, increased nonenzymatic glycosylation, and cross-linkage of dermal collagen. Accumulation of advanced glycation end products (AGEs) with cross-linking of collagen and other macromolecules has been proposed as a potential pathogenetic mechanism.

AGEs are a heterogeneous group of compounds formed by a nonenzymatic reaction between the ketone group on reducing sugars and free amino groups on proteins, called the Maillard reaction. Initially, reversible Schiff bases and subsequently, more stable, covalently bound Amadori products (such as HbA _1c ) are formed. These intermediate structures then undergo complex rearrangements to form the irreversibly cross-linked AGEs. Lipids and nucleic acids exposed to reducing sugars also can form irreversible adducts with AGEs. Originally, all AGEs were believed to exhibit yellow-brown fluorescence and to cross-link macromolecules; however, several of the more recently identified AGEs, such as N-carboxymethyllysine and pyrraline, do not share these properties. Interaction of AGEs with the receptor for AGEs induces oxidative stress and inflammation and may promote thrombosis.

The extent of AGE formation depends on the concentration of reducing sugars and the duration of time over which macromolecules are exposed to these sugars. Thus, long-lived macromolecules, such as extracellular matrix components in vascular endothelium, nerves, and connective tissue, are more likely to become modified with AGEs when exposed to high concentrations of glucose for prolonged periods of time. AGEs have been implicated in the development of both micro- and macrovascular complications of DM. The accumulation of AGEs in skin correlates with the presence of retinopathy, nephropathy, and LJM among patients with type 1 DM. Because the Amadori product HbA _1c is an intermediate in the process of AGE formation, HbA _1c levels do not necessarily correlate with the extent of AGE modification of extracellular matrix macromolecules. Thus, despite the lack of correlation between HbA _1c levels and upper extremity problems in some studies, it is likely that poorer glycemic control over time with resulting AGE formation influences the development of hand and shoulder problems among patients with DM.

Calcific Shoulder Periarthritis (Tendinitis)

Calcific tendinitis is a painful condition most commonly affecting the shoulder in which calcium hydroxyapatite crystals deposit predominantly in periarticular areas. In the shoulder, these crystals may also deposit within the tendons of the rotator cuff. The incidence of calcific shoulder periarthritis is increased among patients with DM. In a case-control study, 824 hospitalized patients with type 2 diabetes and 320 age- and gender-matched hospitalized nondiabetic patients were evaluated with plain radiographs of the shoulder. Shoulder calcifications were evident in 31.8% of the patients with DM but in only 10% of controls without DM. Only 32.4% of those diabetics in whom calcific shoulder periarthritis had been diagnosed on plain radiographs experienced pain or limited range of motion. Calcific tendonitis may coexist with adhesive capsulitis in the shoulder.

Carpal Tunnel Syndrome

Carpal tunnel syndrome (CTS) is an entrapment neuropathy caused by compression of the median nerve within the carpal tunnel. CTS presents with pain and paresthesias of the thumb, index, and middle fingers and of the radial aspect of the ring finger, which may be reproduced on physical examination by percussion of the median nerve at the wrist (Tinel’s test) or on wrist dorsiflexion (Phalen’s test). In diabetics, bilateral involvement of the median and ulnar nerves has been described, with resulting atrophy of the thenar, hypothenar, and intrinsic muscles of the hand and nerve conduction abnormalities.

DM was identified as a risk factor of CTS in a large case-control study of 3391 patients who were identified as having CTS in the UK General Practice Research Database. In another case-control study of 791 patients with CTS who were referred for electrodiagnostic studies, DM, female gender, obesity, and age between 41 and 60 years were observed significantly more frequently among patients with CTS. However, when these patients were stratified by body mass index (BMI), there no longer was a statistically significant association between DM and CTS, suggesting that obesity might have been a confounding factor. Although not a risk factor of unilateral CTS, DM is a risk factor of the development of bilateral CTS. In a study of patients with type 1 or type 2 DM compared with control subjects without DM, Cagliero and colleagues found that the prevalence of CTS among patients with DM was not significantly increased above that in nondiabetic controls.

The observation of simultaneous median and ulnar nerve involvement among patients with DM and the doubled prevalence of CTS in the setting of diabetic polyneuropathy raise the possibility that median neuropathy in diabetics may be caused by factors other than mechanical compression. The impaired median nerve regeneration and recovery observed after carpal tunnel release among some patients with DM suggests that intrinsic nerve pathology, in addition to external compression, might contribute to the pathogenesis of CTS in DM. Proposed pathogenetic factors include hyperglycemia reducing myoinositol transport into nerves and increasing levels of endoneural free sugars, which may result in AGE modification of structural nerve proteins, nerve ischemia causing endoneural hypoxia and demyelination, and deficiencies of neurotrophic growth factors, such as nerve growth factor (NGF) and insulin-like growth factor (IGF)-1. However, 2 subsequent prospective case-control studies found no difference in the rate of electrophysiologic recovery after carpal tunnel release when comparing patients with DM to nondiabetic controls or diabetic patients with polyneuropathy to those without.