Neuromuscular alterations in knee osteoarthritis: a systematic review and meta-analysis

Abstract

Purpose: Neuromuscular alterations are commonly reported in knee osteoarthritis (OA) patients and neuromuscular training is being suggested as a first-line treatment. It is crucial to completely understand these alterations, including neural, structural and functional changes, to optimize interventions for knee OA patients. The aim of this systematic review was to systematically evaluate and determine the current evidence on the neuromuscular alterations of periarticular muscles in patients with knee OA when compared to age- and sex-matched healthy controls. Methods: We searched five databases (Pubmed, Embase, Web of Science, Scopus, Cochrane Central Register of Controlled Trials) for studies comparing neuromuscular alterations of periarticular muscles (i.e. quadriceps, hamstring, gastrocnemius, soleus, popliteus) in a knee OA population to an age- and sex-matched asymptomatic control group. We did not accept contralateral knee as the control group, as there is evidence of bilateral changes in unilateral disease presence. We only included studies reporting structural severity of the OA patients. Outcomes included normalized isokinetic strength, voluntary activation, cortical and spinal-reflex excitability, muscle morphology, and electromyography. Only moderate- and high-quality studies are included assessed by a modified Downs and Black checklist. We performed meta-analyses with random-effects model, grouping on radiographic OA severity, for outcome measures reported by two or more studies. Level of evidence is decided based on van Tulder et al. (2003). We also created an evidence gap-map to show which outcomes need more research. Results: We identified and included 7 high- and 37 medium-quality studies. One prospective cohort and 43 case-control studies reported data for 1643 OA patients and 1532 healthy controls. OA definitions hugely differed among studies in terms of structural severity. There was a high risk of selection bias in the included studies and blinding was not performed in most (n=40/44). Most studies reported quadriceps or hamstring muscle strength (n=24/44), while neural or structural alterations have been of less interest. Studies generally measured quadriceps (n=43/44) or hamstring muscles (n=23/44), while gastrocnemius (n=10/44), soleus (n=1/44) or popliteus (n=0/44) have been of less interest. Our findings indicated deficits in quadriceps (-0.72 [-0.95, -0.50], I2=72%, moderate evidence) and hamstring strength (-0.51 [-0.69, -0.33], I2=0%, moderate evidence) in knee OA patients. We also found lower quadriceps voluntary activation (-0.58 [-0.92, -0.23], I2=36%, moderate evidence) in knee OA group. No change was seen in quadriceps torque variability (0.43 [-0.60, 1.46], I2=84, limited evidence). Structural changes within the muscle focused on muscle volume and intermuscular fat for quadriceps. Results indicated no change in quadriceps muscle volume (0.10 [-0.03, 0.23], I2=0, strong evidence) but increased quadriceps intermuscular fat volume (0.39 [0.26, 0.52], I2=0, strong evidence) in knee OA population. For muscle activation, patients with knee OA showed increased muscle co-contraction for vastus medialis/medial gastrocnemius (0.88 [0.45, 1.30], I2=0%, moderate evidence) but not for vastus medialis/medial hamstring (0.23 [-0.18, 0.64], I2=0%, moderate evidence) during functional tasks. Conclusions: Our findings indicated that deficits in quadriceps and hamstring muscle strength in concurrence with quadriceps voluntary activation deficits are evident in knee OA patients. The lack of deficits of the quadriceps volume indicates that quadriceps weakness may be the result of neural inhibition more than muscle atrophy. This may support the use of neuromuscular training exercises in knee OA populations to increase neural drive. We also found increased co-contraction levels for vastus medialis and medial gastrocnemius muscles but not for vastus medialis and medial hamstring muscles. This interesting finding indicates that the interaction of quadriceps-gastrocnemius muscles may be more relevant than quadriceps-hamstring muscles in knee OA patients. Co-contractions differences may alter the loading of the knee joint; therefore, future studies optimizing the interventions for these co-contraction alterations in knee OA patients may be beneficial to avoid further degeneration. Despite huge number of studies being published in knee OA research, there are still very low evidence levels and poorly reported outcomes in studies measuring neuromuscular manifestations of the disease. High-quality studies with matched healthy control groups are still needed to understand neuromuscular alterations associated with knee OA. Better definition of OA following current guidelines and clearer participant selection processes are needed in future studies to increase evidence level.