Psychological factors outmatched morphological markers in predicting limitations in activities of daily living and participation in patients with lumbar stenosis

Demographic changes in the last decades led to an increased number of elderly people going along with an increase of the prevalence of degenerative disc diseases. Degenerative lumbar spinal stenosis (DLSS) is related to narrowing of the lumbar spinal canal [1] and is associated with degenerative changes of the joint complex, osteophyte formation and ligamentum flavum thickening [2]. Degeneration begins at the intervertebral disc that may cause instability of the spine segment leading to a degenerative cascade of the spine unit [2]. The prevalence of this condition, based on radiographic criteria of patients older than 60 years, is estimated to be about 50% [3]. The clinical prevalence of adults with symptoms of pain and numbness referred to the lower extremities is about 47% [4]. Patients with DLSS may have several clinical symptoms including radiating leg pain, improvement of the pain when bending forward and gait disturbances. This often results in significant reductions in activities of daily living (ADL).

There is no “gold standard” for the diagnosis of DLSS [5]. Clinical judgement is generally based on symptom severity, functional deficits, physical examination and radiographic imaging [6]. The most promising imaging test is magnetic resonance imaging (MRI) [5]. Unfortunately, older adults often demonstrate abnormal MRI findings, even in those who are asymptomatic [7].

Only a few studies have investigated how MRI-based morphological markers are associated with symptom severity and functional deficits. Zeifang et al. found that the severity of lumbar stenosis (cross-sectional area of dural sac) and walking distance in gait analysis are not associated [8]. Likewise, Sigmundsson et al. reported a weak correlation between walking distance, pain interference with ADLs, quality of life and leg and back pain levels and the cross-sectional area of the dural sac. However, they found better general health and less leg and back pain in patients with multilevel stenosis than in patients with single-level stenosis, probably because of adaptation such that patients with a longer history of pain and limitations are more able to accept and handle the consequences of disease [9]. Kuittinen et al. compared the radiological evaluations of DLSS (classification in three severity classes based on MRI inspection of the dural sac area) with clinical findings (e.g., pain interference, pain intensity, depression, walking distance). They found no clear association between the severity of DLSS and clinical findings, but reported partly milder symptoms in patients with more severe stenosis [10]. In contrast, Hong et al. found a significant, albeit small correlation between the number of affected segments (single vs. multilevel DLSS) and pain interference [11]. Taken together, the evidence regarding the association between morphological markers and functional disability reveals small but inconsistent effects. Consequently, it remains to be elucidated which other factors influence functional disability in patients with DLSS.

It is long known that low back pain is associated with psychosocial impairment, poor sleep quality and depression [12]. Patients presenting with significant psychosocial distress at surgical evaluation have a lower quality of life and more severe physical disability. After surgery, patients with moderate psychosocial distress have poorer clinical outcomes, although, in comparison to the preoperative baseline, they improved [13]. In a 10-year follow-up after decompression surgery for lumbar stenosis, Tuomainen et al. reported that patients with lumbar stenosis with elevated depressive symptoms had an increased risk of postoperative pain and disability in the long term follow-up than patients without depressive symptoms [14].

Celestin et al. reported a systematic review of variables that predict pain-related treatment outcomes in the interventional treatment of chronic back pain. They concluded that psychological factors including somatization, depression, anxiety, and poor coping, are important predictors with greater risk of poor postoperative outcome [15]. Similarly, Lee reported associations between preoperative anxiety, optimism and postoperative patient satisfaction [16], and Ekman reported that patient factors including sex, age, preoperative working and exercise status influence postoperative outcomes [17]. Methodological factors like the use of different screening questionnaires, short follow-ups and different indications for spine surgery make a comparison of studies a challenge.

Continuing demographic changes, increased reliance on surgery for DLSS, the increasing use of MRI imaging for DLSS diagnosis, the weak association between radiographic findings and clinical symptoms, and the influence of psychosocial factors on treatment outcome make it critical to improve our understanding of this disease pattern. Thus, the aim of this study was to evaluate the relative influence of sociodemographic characteristics, morphological markers, fear of movement as well as the degree of depression with respect to the prediction of ADL in patients with DLSS.

The characteristics of study participants are presented in Table 1. Study participants on average were 62.5 years old (SD = 11.7), and 50.7% were female.

The increasing number of affected patients and the growing socioeconomic consequences of DLSS underscore the need for an accurate diagnosis. Clinical judgement is generally based on symptom severity, functional deficits, physical examination and radiographic imaging [6]. However, psychological and sociodemographic factors as well as the course of disease, seem to influence the clinical presentation of DLSS (e.g. [13‐15]). To understand better which factors influence the clinical presentation of DLSS, the present study investigated to what extent sociodemographic characteristics, morphological markers as well as depression and fear of movement predict limitations in ADLs and participation in social life in a sample of mostly chronic DLSS patients without a history of surgery. Results of the stepwise regression analyses revealed that pain interference with ADLs as well as limitations in ADLs and participation could be significantly predicted with the proportion of explained variance ranging between 12% (R² = .12; PI-G-physical) and 40% (R² = .40; RehaCAT-LE; AAPI-P). Neither the sociodemographic characteristics age and gender nor disease duration nor the morphological markers number of affected segments and the presence of spondylolisthesis contributed significantly to the prediction. Among the morphological markers only stenosis severity contributed to the prediction of ADLs requiring lower extremity functioning, revealing that more severe stenosis is associated with greater limitations. The variables contributing the most to the prediction of the dependent variables were the psychological factors depression and fear of movement. Depression was the strongest predictor for limitations in participation in social life as well as for pain interference with mental and physical functioning and with the execution of recreational and household activities. Fear of movement was the strongest predictor of limitations in ADLs.

The minor contribution of morphological markers in predicting ADL limitations is consistent with published findings. Most studies investigating the association between the severity of lumbar stenosis and clinical findings (e.g., pain interference, pain intensity, depression, walking distance, gait) did not report any strong correlations [8‐10]. When significant correlations were found, they were generally of small size (e.g. [11], our study). Interestingly, the direction of the associations is inconsistent. In our study more severe stenosis was associated with greater limitations in ADLs requiring lower extremity functioning. Likewise, Hong et al. reported a significant, but small correlation between the number of affected segments (single vs. multilevel DLSS) and pain interference [11]. In contrast, Sigmundsson et al. reported better general health and less leg and back pain in patients with multilevel stenosis as compared to patients with single-level stenosis [9]. This finding is in agreement with our study. We found a significant, but small negative correlation between the number of stenotic segments and pain interference and a significant positive correlation with participation in daily activities. This finding suggests that under special circumstances patients with more affected segments, probably because of adaption, report less pain interference and more participation than patients with only a single affected segment. The results of Kuittinen et al.’s study point in a similar direction. They found a complex association between severity of DLSS and clinical findings, with partly milder symptoms in patients with more severe stenosis; they concluded that DLSS might not solely be an anatomical disorder, but that it has other underlying pathobiological mechanisms [10].

The literature spotlights psychological factors, as for example depression, as potential risk factors for poor clinical presentation (e.g., [13‐15]). In our study a total of 41.5% of the patients had a clinically noticeable depression score. Furthermore, depression and fear of movement were the most powerful predictors of functional disability. The finding that fear of movement is an important predictor for limitations in ADLs and participation is in accordance with Lotzke et al.’s revised fear avoidance model (FAM) [30]. This model was developed to explain the progression of chronic musculoskeletal pain and disability [31]. According to this model, patients with a high pain-related fear, fear of movement and reduced self-efficacy are at risk of developing disabilities, depressed mood and low levels of physical activity [30]. Fear of movement as a protective reaction may be adaptive in the short term but may worsen the problem in the long term [31]. In accordance with Vlaeyen & Linton [31], Lotzke et al. suggested identifying and targeting fear of movement preoperatively to achieve a better postoperative functional outcome [30]. Evidence in support of FAM also comes from [32] who predicted short term pain and disability following lumbar disc surgery.

Interindividual differences in reacting to pain experiences might be responsible for the inconsistent findings regarding the association between morphological markers and clinical findings. Although being indispensable for the diagnosis, morphology seems to play a minor role in explaining limitations in ADLs. Here, the way patients are able to handle the experience of pain seems to be more important as proposed by the FAM [31]. A surgeon should be aware of this influence of psychological factors when choosing the optimal intervention with regard to prognosis on quality of life after treatment.

DLSS is one of the most frequent indications for spinal surgery [33] and in literature surgical therapy, particularly recommended for severe stenosis, seems to be superior to conservative treatments [34]. Many studies show that operative treatment is associated with greater improvement of pain and function in the first years [34, 35], even so the benefits diminish over time [35]. But some patients do not benefit from surgery. Fritsch et al. conducted a meta-analysis about the clinical course of pain and disability following surgery for spinal stenosis. They report, that in general patients experienced extensive reductions in pain and disability in the first 3-month post-surgery though pain level and disability persisted at 5 years follow up [36]. Further, surgery for spinal stenosis seems to affect depression 1 year later. However, persistence of depression after surgery correlated with a worse clinical outcome and a higher rate of unmet expectations [37, 38]. In a Cochrane database analysis, Zaina et al. found 10 to 24% complications in surgical treatment and no complications in non-operative treatment [39]. Repeated operations may result from recurrent spinal stenosis or an increasing spondylolisthesis [40]. In a metanalysis Machado et al. reported a reoperation rate from 3 to 28% [41].

Readers should note several study limitations. While the sample size allowed us to detect small effects, it was not sufficiently large to cross validate the results. We recruited patients during in- and outpatient therapy from two clinics; results may not generalize to other clinics and regions. Because patients were free to decline study participation, self-selection bias may limit generalizability. Nonetheless, findings are consistent with prior studies and support the need to screen for and treat psychological distress before surgery.

Altogether, it is difficult but extremely important for the surgeon to identify patient factors that contribute to the understanding of who is most likely to benefit from surgery. To do so, clinical information, morphological markers and psychological variables, in particular depression and fear of movement, must be taken into account. Future studies should investigate whether psychological screening or a preoperative psychological consultation helps avoid operations or achieves better postsurgical outcomes.

Psychological factors may be a risk factor for a worse clinical presentation and a poorer postoperative outcome. Early psychological screening and treatment, when needed, is recommended and should be implemented routinely. This protocol may help avoid operations with a low likelihood of functional improvement and, direct treatment to patients with psychological disease who may achieve better postoperative outcomes with treatment. Further studies must show to what extent psychological screening and treatment enables better patient outcomes.