Secondary Logo

Journal Logo

OUTCOMES

Which Functional Outcomes Can be Measured in Low Back Pain Trials and Therapies?

A Prospective 2-Year Factor-, Cluster-, and Reliability-Multicenter Analysis on 42 Variables in 1049 Individuals

Niederer, Daniel PhDa; Engel, Tilman PhDb; Pfeifer, Ann-Christin PhDc; Arampatzis, Adamantios PhDd; Beck, Heidrun MDe; Wippert, Pia-Maria PhDf; Schiltenwolf, Marcus MDc; Mayer, Frank MDb

Author Information
doi: 10.1097/BRS.0000000000004028
  • Free

A multitude of outcomes, in parts measuring the same dimension, can be assessed in studies on low back pain (LBP). Numerous patient-reported outcomes, objective physical function measures, and patient-reported experience outcomes exist. To assure comparability of the effects and therapy success, a core set of outcomes, which are assessed in most trials on LBP, is available. A minimum core set is an agreed set of outcomes to be measured and reported in trials on a specific health condition. For LBP, various DELPHI surveys have revealed that such an LBP-specific core set may consists of self-report measures of physical functioning (i.e., disability), pain intensity, and health-related quality of life.1,2 Beyond these three major assessments, if patient-reported outcomes are considered, psychological functioning (depression and anxiety), pain interference/functional ability or disability, healthcare services use, self-rated health, recreation and leisure activity, temporal aspects of pain, social functioning, work ability/productivity, and sleep quality seem to be important.3

Beyond assessing patient-reported core outcomes, it is of great relevance to select outcomes best suiting the intervention and individuals of the trial/therapy. Although most studies only assess patient-reported outcomes, one in three trials measure at least one objective/functional outcome.4 More detailed, more than one third of all studies reported trunk range of motion.5,6 In agreement with the Cochrane Method Guidelines for Systematic Reviews in the Cochrane Back Review Group: “Outcomes of physical examination (range of motion, spinal flexibility, degrees of straight leg raising, or muscle strength), care-provider-centered outcomes (e.g., outcome assessor's global improvement), and other outcomes (medication use, healthcare utilization) may be included where appropriate, depending on the aim of the intervention at issue.”7,8 Interacting with psychological and social factors, neuromuscular deficits and impairments are a major contributor for both the onset and subsequent chronification of nonspecific LBP.9–11 It is, thus, beyond these above sketched meta-epidemiological aspects, of great importance (depending on the study's aim), to select outcomes reflecting the biopsychosocial paradigm of the disease.

A framework on the multidimensional assessment of chronic LBP consequently recommends that the multidimensional nature of LBP should be considered depending on what is relevant for each individual in therapies,12 and, in trials, for different study aims. More detailed, physical outcomes like muscular strength and endurance and even body composition outcomes should also be considered, were relevant.12

Consequently, assessing objective functional/physical variables beyond the core set and other self-reported outcomes are likely to be important in LBP settings. In view of the multitude of functional variables available and, in contrast, existing time constraints (together with the need not to assess the same outcome multiple time), it is of importance to know which functional variables are unique and which ones only display surrogates of another (maybe unique) outcome. Furthermore, one must know if the outcomes’ values are comparable in clusters (e.g., in pain grades), and if the functional outcomes found are reliable (standard error of measurement [SEM]). When such outcomes have been found and selected, they are usually measured multiple times to monitor a potential effect. Recommended follow-up intervals include 6, 12, and 24 months after initiating treatment, with optional follow-up at 3 months and 5 years.2

With the aim to find a statistically compiled functional test battery to assess unique outcomes in LBP, we performed a prospective 2-year factor-, cluster-, and reliability-multicenter analysis on 42 different outcomes. We hypothesize that the numerous variables are reducible to a feasible number by the factor analyses, that these variables are different between pain grade clusters, and that the variables found are reliable within each cluster.

MATERIALS AND METHODS

Design and Ethics

We adopted a multicenter, prospective cohort study. The study has been approved by the independent Ethics Committee of the University of Potsdam (committee's reference number 36/2011) and was conducted in accordance with the ethical standards of the Helsinki Declaration (Ethical Principles for Medical Research in Humanities) with its most recent version of 2013 (World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects, 2013).

Recruitment, Screening, Inclusion, and Exclusion Criteria

Volunteers were recruited during clinical LBP consultation hours, via flyers, local newspapers, and bulletins and through personal recruitment. Interested persons were screened for eligibility. Eligibility criteria were: adults aged 18 to 65 years, as well as the ability to understand the meaning of the study and to answer a questionnaire without help. Although the focus was on participants with LBP, also symptom-free participants were thus included.

Following the application of inclusion and exclusion criteria, patients were approved by a physician or approved study investigator in charge. Each participant signed informed consent before study enrolment.

Study Flow

All participants were monitored for two consecutive years, seven visits were undertaken: baseline after inclusion (M1), 1 month (M2), 3 months (M3), 6 months (M4), 9 months (M5), 1 year (M6), and 2 years (M7) after M1. During each of these seven appointments, all outcomes were assessed in a standardized order.

Outcomes and Assessments

Overall, 42 outcomes of assessment battery (see Wippert et al13 for an overview of the whole battery) were used. Questionnaires were followed by medical history and the biomechanical assessments.

Questionnaires—Self-reported Core Outcomes

All patient-reported outcomes were assessed using the validated German versions of the respective questionnaire.

Highest school and work education, job position (all closed questions with eight or nine answer possibilities), and working hours per week were asked using single-items questions; the questions were selected from the socioeconomic status questionnaire SES.14

Participants were classified into one of the five hierarchical pain and disability grades ranging from pain free (Grade 0) to severely limiting, high disability (Grade 4) according to the Chronic Pain Grade questionnaire (11-point Likert scales15). In addition, chronic characteristic pain intensity, self-reported pain-related disability, and the absence days from work were assessed by subscales of the Chronic Pain Grade questionnaire (Characteristic pain intensity [0 = “no pain” to 10 = “the worst pain imaginable”], disability [0 = “no disability” to 100 = “I was incapable of doing anything”], absence days subscale [“how many days were you absent from work in the last three months?”]).

The general health quality wash assessed by a single 11-point Likert scale (0= “completely dissatisfied”, 10 = “completely satisfied”).16

History and Anamnesis

Participants’ age, sex/gender, physical activity, and training characteristics (habitual minutes of training per week, sessions per week, duration per sessions, years of experience) were asked. Heart rate, blood pressure, body height, and weight were measured using standard equipment, and the body mass index (BMI) was calculated from the latter two.

Lumbar trunk flexion and extension range of motion were assessed using simple clinical goniometers; the ratio was calculated, likewise.

Biomechanical/Functional Outcomes

In case of no contraindications, functional outcomes were assessed in a standardized order: dynamic balance—static postural control—jumping ability—maximal strength—strength endurance—jump ability fatigued. In all functional assessment assessable for both the left and right side, the side of interest was selected randomized. For standardizing purpose (setting and device familiarization) and to avoid bias, one demonstration (by the investigator) and one test run were completed before each new biomechanical outcome assessment started.

For the dynamic balance evaluation, the participants stepped one-legged (50 cm distance) without wearing shoes on a force measurement platform (sampling rate of 1 kHz; force plate: Kistler or AMTI) and had to stabilize themselves as fast as possible in a single leg stance on the platform. The first second following ground contact (threshold: ± 10% of body weight) was monitored.

For the static posturography measurement, the participants maintained in upright stance as still as possible on the force plate for 10 seconds, with their hands placed on their hips and their eyes open. In both setups, the trace length (mm) of the excursion of the center of pressure was the outcome of interest.

To assess the jumping ability, counter movement jumps (CMJs) were performed. After warmup (30 s tapping on a stepper with a 14–21 cm height) and familiarization, participants performed two bipedal CMJs, followed by two combined bipedal CMJs/reactive jumps. For the reactive jumps, participants were told to immediately add another jump after landing from the CMJ.17 Outcomes were: peak ground reaction force during jump off (N), flight time (ms), and calculated jump height (m).

Trunk strength was assessed by isokinetic testing in flexion and extension movements. A warmup with 30 repetitions at approximately 50% of maximum voluntary contractions (MVCs) was performed initially at a velocity of 60°/s. Following a 1-minute rest interval, five repetitions at MVC (60°/s) were conducted. Range of motion (isokinetic) was set to 55° (device-specifically between 40° and 45° in flexion and 10° and 15° in extension; Ferstl isomed 2000; Con-Trex MJ/TP 1000). Subsequently strength endurance was assessed by 26 repetitions at MVC using the same procedure as for the previous tests. Participants were verbally encouraged in a standardized way to elicit maximal effort during all trials. Outcomes (for maximal strength and begin/end of strength endurance) were: maximum torque (Nm) and work (J), during flexion and extension movements, and as the ratio of these two.

Statistics

Data were collected and analyzed centrally. All analyses were performed following visual and range-data plausibility control. The statistical procedures were executed after the examination of underlying assumptions (like normality distribution of the data or residuals, variance homogeneity, interval or ordinal scaled data, Bartlett test of sphericity) for the respective analyses and for parametric or rather nonparametric hypotheses testing. A two-sided a priori level of significance was set at α = 0.05 for all statistical analyses.

First, two exploratory factor analyses (using the R-syntax-based program jamovi [Version 1.0.7.0, The jamovi project 2019, Sydney, Australia]) were calculated, once for the baseline values (cross-sectional analysis) and once for the changes scores (longitudinal analysis). Factoring method was the maximum likelihood extraction method, in combination with a varimax factor rotation. Results were displayed as factor loadings and uniqueness of the single outcomes. Cumulative variance explanation was chosen as the outcome for the optimization procedure. Root mean square error of approximation along with its 90% confidence interval and the Tucker-Lewis index were selected as model fit measures. Loads above 0.3 were included. For each factor, the outcome with the strongest factor loading was selected for the upcoming statistical processes. Unique outcomes were further selected. The corresponding data were displayed using arithmetic means and standard deviations.

Second, hierarchical cluster analyses were performed (again for both, cross-sectional and gain-score data) using IBM SPSS 25 (SPSS Inc., Chicago, IL). Maximal cluster number was 4, minimal was 2. Average linkage method and Euclidean distance measure were used for the analyses. The results were plotted using dendrograms.

Third, reliability analyses using SPSS and Excel (Microsoft Excel 2013 for windows, Microsoft Corporation, Redmond, WA) were calculated. To determine the short- and long-term stability, intraclass correlation coefficients (ICCs 2.1), SEMs ( SEM = SD × √(1 − ICC), and the typical error of measurements using the coefficient of variation (= SD/mean) were calculated.

RESULTS

Participants Flow and Characteristics

From the approximately 2200 persons screened, n = 1071 participants were included into the study. The data from n = 1049 participants were analyzed, after n = 22 participants had to be excluded due to missing data. At follow-up visits, the number of participants decreased (M2: 980, equals a 7% loss since baseline; M3: 817, −22%; M4: 747, −29%; M5: 713, −32%; M6: 651, −38%; M7: 665, −37%. Most common reasons for dropout (when provided) were time issues, location-related issues (e.g., change of residence), and health complications (e.g., acute injuries, pregnancy, newly diagnosed disease). In total, 47 adverse events (AEs) occurred. The reasons for the AEs were, among others acute injuries, severe back pain, and cardiovascular complications. None of the AEs was undoubtedly affected by the studies’ outcome assessments. One serious AE was recorded, which was unrelated to the study participation (death due to a pre-existing disease).

More than half (596) of the participants were women, 450 were men, 25 provided no answer, no one was other/divers/nonbinary. At inclusion, the age ranged between 18 and 65 years, mean age was 38.7, and standard deviation 13.3 years.

Factor Analyses’ Results

The factor analyses for both the baseline/cross-sectional values are displayed in Table 1; Table 2 displays the corresponding values for the change scores.

TABLE 1 - Factor Analyses Outcomes for the Cross-sectional Values
Real Values Model Fit: RMSEA: 0.164, 90% CI: .156-NaN; TLI: .635; BIC: −2846; χ2: 6161 (df: 553, P < 0.001)
Factor Number Uniqueness
1 2 3 4 5 6 7 8
Strength endurance: flexion torque 0.902 0.031
Strength maximal: flexion torque 0.878 0.053
Strength endurance: extension torque 0.862 0.346 0.042
Strength endurance: flexion work 0.857 −0.37 0.049
Strength endurance: extension work 0.844 0.364 0.076
Strength maximal: flexion work fatigued 0.837 0.305 −0.39 0.046
Strength maximal: extension torque 0.792 0.374 −0.35 0.085
Strength maximal: extension work fatigued 0.765 −0.38 0.132
Body height 0.718 0.621 0.005
Counter movement jump: ground reaction force 0.664 0.335 0.514 0.171
Counter movement jump postfatigue: ground reaction force 0.624 0.344 0.537 0.191
Counter movement jump: flight time fatigued 0.364 0.899 0.698
Counter movement jump: jump height fatigued 0.376 0.896 0.006
Counter movement jump: flight time 0.415 0.85 0.005
Counter movement jump: jump height 0.429 0.844 0.070
Static postural control: CoP trace length 0.074
Age −0.39 0.649
Work education 0.919
Heart rate 0.910
Chronic pain: intensity 0.912
Health quality 0.936
Dynamic balance: CoP trace length 0.972
Strength endurance: flexion-extension relation work 0.913 0.146
Strength endurance: flexion-extension relation torque 0.876 0.220
Strength endurance fatigued: flexion-extension relation torque 0.857 0.216
Maximal strength: flexion-extension relation torque 0.853 0.232
Chronic pain: disability 0.910
Body mass index 0.329 0.787 −0.43 0.005
Body weight 0.688 0.701 0.004
Blood pressure: diastolic 0.403 0.828
Blood pressure: systolic 0.397 0.829
Working hours per week 0.911
Highest education level 0.889
Job position 0.949
Range of motion: extension-flexion-relation 0.986 0.005
Range of motion: extension 0.945 0.006
Days of absenteeism from work 0.870
Training: minutes per week 0.971 0.005
Training: sessions per week 0.855 0.231
Training: duration per sessions 0.358 0.834
Training: years of experience 0.974
Range of motion: flexion 0.988 0.005
Factor leaders and unique outcomes are highlighted in bold letters.CI indicates confidence interval; CoP, center of pressure; RMSEA, root mean square error of approximation; TLI, the Tucker Lewis Index.

TABLE 2 - Factor Analyses Outcomes for the Change Score Values
Change Scores Model Fit: RMSEA: .139, 90% CI: .131−142; TLI: .715; BIC: 959; χ2: 1891 (df: 144, P < 0.001)
Factor Number Uniqueness
1 2 3 4 5 6 7 8 9
Strength endurance: flexion-extension relation work 0.872 0.200
Strength endurance fatigued: extension work 0.816 0.317
Strength endurance fatigued: flexion-extension relation torque 0.779 0.311
Strength endurance: flexion-extension relation torque 0.635 0.325 0.475
Chronic pain: pain intensity 0.976
Chronic pain: Disability 0.988
Strength endurance: extension work 0.904 0.005
Strength endurance: extension torque 0.305 −0.77 0.173
Counter movement jump: jump height 0.965 0.005
Counter movement jump: flight time 0.955 0.019
Counter movement jump: ground reaction force 0.910
Counter movement jump: flight time fatigued 0.952 0.005
Counter movement jump: jump height fatigued 0.942 0.019
Range of motion: extension-flexion-relation 0.96 0.005
 Range of motion: extension 0.958 0.015
Dynamic balance: CoP trace length 0.964
Days of absenteeism from work 0.989
Strength endurance: flexion torque 0.894 0.005
 Strength endurance: flexion work 0.34 −0.62 −0.34 0.278
Maximal strength: extension torque 0.36 0.881 0.005
Maximal strength: flexionextension relation torque 0.308 −0.62 0.358
Health quality 0.985
Strength endurance fatigued: flexion work 0.7 0.413
Strength endurance fatigued: flexion torque 0.539 −0.71 0.005
Static postural control: CoP trace length 0.959
Range of motion: flexion 0.993 0.005
CI indicates confidence interval; CoP, center of pressure; TLI, the Tucker Lewis Index.

The factor analysis for the real values revealed eight factors with a cumulative variance explanation of 61.7%. Unique were 13 of the outcomes, they could not be assigned to a factor. The corresponding change-score factor analysis revealed nine factors with a total variance explanation of 61.8%, seven outcomes were unique.

Taken together, 10 outcomes are important for both cross-sectional and change-score analyses, 11 were found to be the most valuable for cross-sectional and 4 were uniquely relevant for the change scores. These outcomes are highlighted in bold letters in the table. The model fits are acceptable.

The factor's strongest outcomes and the unique outcomes except the key outcomes disability and pain were selected for the cluster analyses. All values of the outcomes assessed more often than at baseline (outcomes except height, age, education, training history, and job position) are displayed in the Figures 1–4.

Figure 1
Figure 1:
Participants characteristics’ means and standard deviations for all visits. BMI indicates body mass index; bpm, beats per minute; h, hours.
Figure 2
Figure 2:
Patient-reported outcomes’ means and standard deviations for all visits.
Figure 3
Figure 3:
Functional outcomes’ (part 1) means and standard deviations for all visits.
Figure 4
Figure 4:
Functional outcomes’ (part 2) means and standard deviations for all visits. J indicates Joule; ms, milliseconds; Nm, Newtonmeter.

Overall, the values remain stable during the 2 years. Descriptively, a very slight total sample change over time was found in the outcomes working hours per week and health quality (increase), as well as in chronic pain intensity, chronic pain disability, range of motion flexion-extension-relation, and dynamic balance (all decrease over time). Very slight but more complex changes were found in the outcomes absence days from work (u-shape), heart rate, CMJ postfatigue flight time (both inverted u-shape), range of motion in flexion (no patterns), strength endurance flexion-extension-relation work (both first decrease, then stable), CMJ height, maximal strength extension, and strength endurance extension work (all three first increase, then stable). No descriptive tendency for a change over time was found for the other outcomes (BMI, training minutes per week, static postural control, CMJ ground reaction force, strength endurance flexion).

Clustering Analyses

The results of the cluster analyses for both the baseline values and change scores are displayed in Figure 5.

Figure 5
Figure 5:
Dendrogams of the cluster analysis for the cross-sectional/baseline-values (-A-) and the change score values (-B-).

Patients with pain grades 1 to 3 (low to moderate pain and disability) show comparable real (baseline) values patterns (cluster 2, n = 949). Grade 0 (currently pain free, cluster 1, n = 110) and grade 4 (cluster 3, n = 12, severely limiting, high disability patients) are unique and cannot be cumulated with other grades. For the change-score values, grade 0 persons can be considered as a part of the grade 1 to 3 cluster. For the reliability analyses, the three clusters (grade 0, grade 1–3, grade 4) are processed separately.

Reliability Analyses

The reliability outcomes for the factors identified (without the comparators/core outcomes pain intensity and disability) are, separated for the measurement times and clusters, displayed in Table 3.

TABLE 3 - Reliability Analyses’ Properties for (- A -) Short-Term (1 mo), (- B -) Intermediate-Term (3 mo), (- C -) Long-Term 6 mo, (- D -) Long-Term 9 mo, (- E -) Follow-up 1 year, and (- F -) Follow-up 2 years
- A - ICC/Cronbachs alpha CV SEM SEM [%]
Short-term: 1 mo Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4
Patient characteristics
Body mass index 0.992 0.970 0.957 0.011 0.015 0.019 0.291 0.593 0.345 1% 2% 1%
Heart rate 0.554 0.625 0.785 0.085 0.083 0.086 6.065 6.530 5.351 9% 10% 7%
Working hours per week 0.754 0.788 N/A – n 0.151 0.171 N/A - Var 4.775 5.821 N/A - Var 14% 18% N/A - Var
Training minutes per week 0.908 0.855 0.760 0.224 0.243 0.118 116.414 102.251 46.476 25% 37% 11%
Functional outcomes
Range of motion in flexion N/A - Var 0.463 N/A – n 0.011 0.014 0.032 N/A - Var 3.450 N/A - n N/A - Var 4% N/A - n
Range of motion flexion-extension-relation 0.367 0.418 0.434 0.033 0.041 0.114 0.033 0.038 0.030 8% 9% 8%
Dynamic balance 0.841 0.797 0.728 0.124 0.135 0.152 19.868 22.850 27.134 14% 16% 16%
Static postural control 0.968 0.969 0.944 0.108 0.105 0.124 67.076 64.666 166.479 10% 10% 33%
Counter movement jump ground reaction force 0.908 0.961 0.997 0.056 0.042 0.032 117.577 72.072 20.200 8% 5% 1%
Counter movement jump height 0.967 0.974 0.187 0.065 0.057 0.107 0.013 0.011 0.013 6% 5% 8%
Maximal strength extension torque 0.951 0.950 0.986 0.097 0.110 0.169 22.026 19.941 6.785 10% 10% 5%
Strength endurance flexion torque 0.977 0.974 0.907 0.063 0.068 0.168 7.538 8.008 16.68 6% 6% 16%
strength endurance extension work 0.944 0.958 0.497 0.108 0.098 0.162 378.542 312.4 582.9 10% 9% 25%
Strength endurance flexion-extension-relation work 0.883 0.625 0.858 0.105 0.105 0.236 0.064 0.205 0.055 10% 32% 7%
Counter movement jump postfatigue flight time 0.978 0.977 0.748 0.025 0.026 0.028 10.36 10.19 8.090 2% 2% 2%
- B - ICC/Cronbachs Alpha CV SEM SEM [%]
Intermediate-term: 3 mo Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4
Patient characteristics
Body mass index 0.991 0.974 N/A - n 0.014 0.017 N/A - n 0.283 0.536 N/A - n 1% 2% N/A - n
Heart rate 0.645 0.567 N/A - n 0.084 0.088 N/A - n 6.081 7.015 N/A - n 9% 10% N/A - n
Working hours per week 0.785 0.844 N/A - n 0.227 0.172 N/A - n 6.732 5.450 N/A - n 20% 17% N/A - n
Training minutes per week 0.887 0.889 N/A - n 0.249 0.261 N/A - n 154.418 78.987 N/A - n 33% 29% N/A - n
Functional outcomes
Range of motion in flexion N/A - Var 0.458 N/A - n 0.009 0.021 N/A - n N/A - Var 5.072 N/A - n N/A - Var 6% N/A - n
Range of motion flexion-extension-relation 0.210 0.434 N/A - n 0.031 0.048 N/A - n 0.034 0.045 N/A - n 8% 10% N/A - n
Dynamic balance 0.418 0.765 N/A - n 0.160 0.144 N/A - n 70.245 25.548 N/A - n 51% 18% N/A - n
Static postural control 0.899 0.943 N/A - n 0.145 0.114 N/A - n 125.664 88.892 N/A - n 18% 14% N/A - n
Counter movement jump ground reaction force 0.890 0.939 N/A - n 0.055 0.044 N/A - n 117.486 95.296 N/A - n 8% 6% N/A - n
Counter movement jump height 0.967 0.971 N/A - n 0.066 0.061 N/A - n 0.013 0.012 N/A - n 6% 5% N/A - n
Maximal strength extension torque 0.952 0.936 N/A - n 0.101 0.124 N/A - n 21.553 22.258 N/A - n 9% 11% N/A - n
Strength endurance flexion torque 0.966 0.971 N/A - n 0.082 0.070 N/A - n 9.192 8.447 N/A - n 7% 7% N/A - n
strength endurance extension work 0.911 0.940 N/A - n 0.101 0.102 N/A - n 469.064 352.320 N/A - n 12% 10% N/A - n
Strength endurance flexion-extension-relation work 0.889 0.828 N/A - n 0.106 0.118 N/A - n 0.062 0.083 N/A - n 10% 13% N/A - n
Counter movement jump postfatigue flight time 0.963 0.973 N/A - n 0.030 0.028 N/A - n 12.851 10.789 N/A - n 3% 3% N/A - n
- C - ICC/Cronbachs Alpha CV SEM SEM [%]
Long-Term 1: 6 mo Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4
Patient characteristics
Body mass index 0.972 0.969 N/A - n 0.016 0.021 N/A – n 0.505 0.623 N/A - n 2% 3% N/A - n
Heart rate 0.456 0.477 N/A - n 0.079 0.096 N/A – n 6.270 8.130 N/A - n 10% 12% N/A - n
Working hours per week 0.951 0.768 N/A - n 0.247 0.208 N/A – n 3.539 7.004 N/A - n 10% 21% N/A - n
Training minutes per week 0.901 0.832 N/A - n 0.257 0.301 N/A - n 105.991 87.899 N/A - n 23% 32% N/A - n
Functional outcomes
Range of motion in flexion N/A - Var 0.177 N/A - n 0.019 0.023 N/A – n N/A - Var 5.763 N/A - n N/A - Var 6% N/A - n
Range of motion flexion-extension-relation 0.264 0.247 N/A - n 0.048 0.062 N/A – n 0.050 0.061 N/A - n 12% 14% N/A - n
Dynamic balance 0.825 0.755 N/A - n 0.147 0.146 N/A – n 18.779 21.972 N/A - n 14% 16% N/A - n
Static postural control 0.954 0.950 N/A - n 0.140 0.116 N/A – n 74.034 75.189 N/A - n 11% 11% N/A - n
Counter movement jump ground reaction force 0.971 0.943 N/A - n 0.043 0.050 N/A - n 60.092 85.433 N/A - n 4% 6% N/A - n
Counter movement jump height 0.962 0.968 N/A - n 0.071 0.063 N/A - n 0.014 0.012 N/A - n 6% 6% N/A - n
Maximal strength extension torque 0.938 0.926 N/A - n 0.104 0.120 N/A - n 24.713 23.069 N/A - n 11% 11% N/A - n
Strength endurance flexion torque 0.966 0.967 N/A - n 0.077 0.072 N/A - n 8.704 8.876 N/A - n 7% 7% N/A - n
Strength endurance extension work 0.946 0.953 N/A - n 0.107 0.105 N/A - n 394.756 309.105 N/A - n 10% 9% N/A - n
Strength endurance flexion-extension-relation work 0.895 0.816 N/A - n 0.117 0.122 N/A - n 0.060 0.081 N/A - n 9% 13% N/A - n
Counter movement jump postfatigue flight time 0.965 0.970 N/A - n 0.031 0.030 N/A - n 13.424 11.655 N/A - n 3% 3% N/A - n
- D - ICC/Cronbachs Alpha CV SEM SEM [%]
Long-Term 2: 9 mo Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4
Patient characteristics
Body mass index 0.981 0.963 N/A - n 0.020 0.024 N/A - n 0.418 0.643 N/A - n 2% 3% N/A - n
Heart rate 0.481 0.453 N/A - n 0.101 0.101 N/A - n 6.878 8.646 N/A - n 10% 13% N/A - n
Working hours per week 0.462 0.783 N/A - n 0.386 0.210 N/A - n 12.311 6.322 N/A - n 36% 19% N/A - n
Training minutes per week 0.940 0.778 N/A - n 0.253 0.318 N/A - n 84.193 112.858 N/A - n 18% 41% N/A - n
Functional outcomes
Range of motion in flexion N/A – Var 0.263 N/A - n 0.023 0.023 N/A - n N/A - Var 5.661 N/A - n N/A - Var 6% N/A - n
Range of motion flexion-extension-relation 0.355 0.227 N/A - n 0.044 0.062 N/A - n 0.045 0.062 N/A - n 10% 14% N/A - n
Dynamic balance 0.835 0.624 N/A - n 0.138 0.159 N/A - n 16.390 34.990 N/A - n 12% 25% N/A - n
Static postural control 0.945 0.938 N/A - n 0.147 0.130 N/A - n 75.420 84.112 N/A - n 11% 13% N/A - n
Counter movement jump ground reaction force 0.913 0.916 N/A - n 0.062 0.052 N/A - n 144.564 143.321 N/A - n 10% 10% N/A - n
Counter movement jump height 0.945 0.963 N/A - n 0.081 0.067 N/A - n 0.017 0.013 N/A - n 7% 6% N/A - n
Maximal strength extension torque 0.934 0.912 N/A - n 0.110 0.125 N/A - n 25.258 24.506 N/A - n 11% 12% N/A - n
Strength endurance flexion torque 0.958 0.957 N/A - n 0.080 0.081 N/A - n 9.878 10.267 N/A - n 7% 8% N/A - n
strength endurance extension work 0.932 0.940 N/A - n 0.096 0.110 N/A - n 412.440 342.576 N/A - n 11% 10% N/A - n
Strength endurance flexion-extension-relation work 0.923 0.803 N/A - n 0.105 0.120 N/A - n 0.064 0.087 N/A - n 10% 14% N/A - n
Counter movement jump postfatigue flight time 0.973 0.966 N/A - n 0.029 0.031 N/A - n 10.605 12.506 N/A - n 2% 3% N/A - n
- E - ICC/Cronbachs Alpha CV SEM SEM [%]
Follow-up 1: 1 year Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4
Patient characteristics
Body mass index 0.871 0.955 N/A - n 0.027 0.028 N/A - n 0.936 0.737 N/A - n 4% 3% N/A - n
Heart rate 0.412 0.394 N/A - n 0.113 0.099 N/A - n 7.707 7.828 N/A - n 12% 11% N/A - n
Working hours per week 0.965 0.683 N/A - n 0.114 0.241 N/A - n 3.111 8.513 N/A - n 9% 26% N/A - n
Training minutes per week 0.931 0.727 N/A - n 0.279 0.342 N/A - n 82.694 119.177 N/A - n 18% 43% N/A - n
Functional outcomes
Range of motion in flexion 0.441 0.327 N/A - n 0.013 0.022 N/A - n 3.236 5.112 N/A - n 4% 6% N/A - n
Range of motion flexion-extension-relation 0.353 0.198 N/A - n 0.041 0.064 N/A - n 0.042 0.057 N/A - n 10% 13% N/A - n
Dynamic balance 0.850 0.614 N/A - n 0.145 0.158 N/A - n 16.274 35.923 N/A - n 12% 26% N/A - n
Static postural control 0.951 0.920 N/A - n 0.135 0.132 N/A - n 68.172 97.737 N/A - n 10% 15% N/A - n
Counter movement jump ground reaction force 0.972 0.944 N/A - n 0.046 0.049 N/A - n 66.029 83.407 N/A - n 4% 6% N/A - n
Counter movement jump height 0.966 0.955 N/A - n 0.080 0.069 N/A - n 0.015 0.014 N/A - n 7% 7% N/A - n
Maximal strength extension torque 0.955 0.894 N/A - n 0.091 0.138 N/A - n 21.038 25.489 N/A - n 9% 12% N/A - n
Strength endurance flexion torque 0.964 0.948 N/A - n 0.070 0.087 N/A - n 9.026 10.213 N/A - n 7% 8% N/A - n
Strength endurance extension work 0.929 0.933 N/A - n 0.098 0.113 N/A - n 416.710 335.787 N/A - n 11% 9% N/A - n
Strength endurance flexion-extension-relation work 0.886 0.776 N/A - n 0.115 0.129 N/A - n 0.065 0.089 N/A - n 10% 14% N/A - n
Counter movement jump postfatigue flight time 0.977 0.964 N/A - n 0.028 0.033 N/A - n 11.279 12.498 N/A - n 3% 3% N/A - n
- F - ICC/Cronbachs Alpha CV SEM SEM [%]
Follow-up 2: 2 years Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4 Grade 0 Grade 1-3 Grade 4
Patient characteristics
Body mass index 0.963 0.905 N/A - n 0.022 0.032 N/A - n 0.525 1.222 N/A - n 2% 5% N/A - n
Heart rate 0.548 0.222 N/A - n 0.102 0.103 N/A - n 5.928 9.650 N/A - n 9% 14% N/A - n
Working hours per week 0.512 0.787 N/A - n 0.255 0.162 N/A - n 10.064 6.230 N/A - n 29% 19% N/A - n
Training minutes per week 0.870 0.689 N/A - n 0.273 0.346 N/A - n 107.383 139.309 N/A - n 23% 51% N/A - n
Functional outcomes
Range of motion in flexion N/A - Var 0.442 N/A - n 0.026 0.020 N/A - n N/A - Var 4.982 N/A - n N/A - Var 6% N/A - n
Range of motion flexion-extension-relation 0.186 0.238 N/A - n 0.068 0.068 N/A - n 0.059 0.059 N/A - n 13% 13% N/A - n
Dynamic balance 0.724 0.712 N/A - n 0.167 0.151 N/A - n 19.619 23.003 N/A - n 14% 16% N/A - n
Static postural control 0.957 0.954 N/A - n 0.119 0.125 N/A - n 66.613 73.218 N/A - n 10% 11% N/A - n
Counter movement jump ground reaction force 0.974 0.964 N/A - n 0.045 0.050 N/A - n 64.112 65.775 N/A - n 4% 4% N/A - n
Counter movement jump height 0.959 0.958 N/A - n 0.079 0.068 N/A - n 0.015 0.013 N/A - n 7% 6% N/A - n
Maximal strength extension torque 0.922 0.902 N/A - n 0.124 0.132 N/A - n 29.208 24.504 N/A - n 13% 12% N/A - n
Strength endurance flexion torque 0.965 0.944 N/A - n 0.070 0.087 N/A - n 8.286 11.197 N/A - n 6% 9% N/A - n
Strength endurance extension work 0.930 0.923 N/A - n 0.105 0.118 N/A - n 403.911 365.476 N/A - n 11% 10% N/A - n
Strength endurance flexion-extension-relation work 0.882 0.794 N/A - n 0.135 0.129 N/A - n 0.088 0.079 N/A - n 14% 12% N/A - n
Counter movement jump postfatigue flight time 0.960 0.947 N/A - n 0.030 0.037 N/A - n 11.999 15.284 N/A - n 3% 4% N/A - n
Each time, the intraclass correlation coefficients, the coefficient of variation, and the standard error of measurement (absolute and in percent) are displayed CV coefficient of variation.ICC indicates intraclass correlation coefficient; N/A – n; not applicable, number of participants too low; N/A – Var, not applicable, not enough variance; N/A – ICC, not applicable, ICC equals zero.

DISCUSSION

The 42 initially assessed outcomes were reduced to 25. Ten outcomes were factor load leaders or unique for both cross-sectional and change-score analyses, eleven were found to be the most valuable for cross-sectional and four were uniquely relevant for the change-scores. From these, best suiting variables beyond the standard core outcomes can be selected for trials on and therapies against LBP.

Reliability

The outcomes were not all perfectly stable over time, a slight time course was found in working hours per week, health quality, chronic pain intensity, chronic pain disability, range of motion flexion-extension-relation, dynamic balance, heart rate, fatigued CMJ flight time, range of motion in flexion, strength endurance work for flexion-extension-relation, CMJ height, maximal strength in extension, and strength endurance work in extension. Most of the changes were small and likely not to be of clinical relevance. The significant changes over time may mostly be attributed to the large number of participants and is likely to display a potential small learning effect over time (e.g., in the CMJs).

For the outcomes with a potential change over time, reliability was found to be, depending on the outcome and timepoint, low (e.g., range of motion measures) to excellent (e.g., maximal strength extension torque). The other outcomes (BMI, training minutes per week, static postural control, CMJ ground reaction force, strength endurance flexion) were unchanged over time and mostly showed excellent reliability values, with ICCs over 0.9.

The three clusters are somewhat homogeneous in their outcomes’ values. They, however, display no generalizable differences in the reliability values.

Interdependence of Outcomes

Although we found selective factors and clusters, outcomes in patients with LBP are hardly ever independent. One must obey that, for example, a causal relationship between pain and disability is given: pain leads to disability,18 the interaction of pain and disability is affected by further self-report outcomes like depressive symptoms and kinesiophobia.19,20 Strength/torque is, in contrast, only slightly associated with pain or disability.21,22 The same was found for other functional outcomes and change score comparisons: mobility, trunk extension and flexion torque, and strength endurance were only slightly associated with pain and disability change scores.23 Yet, all these outcomes are trainable by exercise in LBP settings.24 These are hints that, although functional and self-report outcomes display different symptom dimensions in LBP, they both are of potential relevance as outcomes.

If known pathways and interactions can be considered and if a biopsychosocial problem can early be identified and addressed, it can be targeted better.13,25,26 Consequently, the focus of the (individualized) therapy may be set more accurately.27,28 LBP is a complex, dynamic, and multidimensional health issue29 and, as confirmed by our change scores’ analyses, accompanied by continually changing surrogate values. It is thus relevant to know reliability values of different timepoints and outcomes, and to individually select outcome dimensions relevant for gain score or cross-sectional measurements.

What Dimension should I Measure?

Beyond the core set of outcomes, a selection of the outcomes to be assessed should be undertaken based on the trial and/or patient characteristics. After the risk factors and therapy goals assessment, outcome measures that are individually relevant, that can be easily assessed, and that capture the most important domains of the biopsychosocial model, not only but including functional/physical measures can be selected from the multitude of available outcomes.12 This individualization is valid for, on the one hand, individual patients’ preferences, but also for the individualization of trials’ outcome selection.

As compelling evidence on (a small but relevant) effect of exercise on LBP,30,31 even with hints on a dose-response relationship32 exists, selecting functional outcomes may be, not only but in particular, relevant in measuring the effects of exercise trials and therapies.

Cross-validated Tools to Assess the Same Dimension

Usually, multiple instruments measuring the same health construct in the same patient population are available. Numerous patient-reported outcomes for measuring back-specific functional status in patients with LBP were adopted. Most often, the Oswestry Disability Index or the Roland Morris Disability Questionnaire are used. Against the background of different timepoints (the chronic pain disability assessments, e.g., assess the last 3 months retrospectively), the intervention and patients’ characteristics, it is quite challenging to select the best outcome.

The same is valid for the functional outcomes: Trunk Extension Endurance and Isolated Lumbar Extension Strength, for example, were found to be only slightly correlated in patients LBP.33 A solution could be to select an instrument best matching with the target dimension (= validity), to rate if the selected outcome is feasible in the present setting of the core set (= costs, patient and responder burden, equipment needs and other practical aspects3) and if it fits with the present specialities of LBP.12,34

Although the outcomes should fit the participants and interventions, the core outcomes should not be neglected. Any inconsistency between studies limits the comparison of findings among trials, which may strongly affect the conclusions of systematic reviews.35,36

Methodological Considerations and Limitations

The conducted test battery in the current study lasted for one single measurement timepoint over 2 hours. That may lead already to a considerable time burden and stress for the participants and was the reason why we have not assessed further outcomes. As they were found to, likewise, provide a partially unique value,37 it would, nevertheless have been of relevance to assess, for example, kinaesthesia/proprioception, and movement behavior beyond simple range of motion.3,35,36

This must be considered as a potential limitation. It is, unfortunately, further impossible to include binary variables (like analgesic use, sex/gender—although measured) into factor analyses as the underlying assumptions for factor analyses would be violated. Only a minor share of our participants was classified as grade 4 (Cluster 3). That limits the applicability of the reliability findings of the Cluster 3 participants.

CONCLUSION

To mirror the multidimensional components of LBP, it is recommended that outcomes beyond core variables be tailored to the patient's and trial characteristics.

We found 25 potentially meaningful factors in the context of self-reported outcomes (working hours per week, health quality of life, chronic pain intensity, chronic pain disability, absence days from work, and training minutes per week), objective functional measurements (range of motion flexion-extension-relation, dynamic balance, CMJ postfatigue flight time, range of motion in flexion, strength endurance flexion-extension-relation work, CMJ height, maximal strength extension, strength endurance extension work, static postural control, CMJ ground reaction force, strength endurance flexion), and body characteristics (BMI, heart rate) derived by the factoring analyses. Dimensions of relevance for cross-sectional and gain-score analyses were mostly congruent. The present framework may help to select appropriate functional outcomes to complement testing batteries on core set's outcomes like pain intensity, disability, and health quality when planning and conducting trials and therapies on LBP. For all potentially relevant timepoints, the SEMs may help to interpret the gain scores of individual studies on (cluster 1) participants without current pain, (cluster 2) low intensity and disability back pain patients, and (cluster 3) clinical pain patients.

Key Points

  • The aim of this study was to provide evidence on a set of potential relevant outcomes by investigating their uniqueness and usefulness.
  • We found 25 potentially meaningful functional outcomes in the context of objective functional measurements (such as trunk range of motion, dynamic and static balance, and strength, and muscle fatigue resistance) and body characteristics.
  • The present framework may help to select appropriate functional outcomes to complement testing batteries on core set's outcomes when planning and conducting trials and therapies on LBP.

References

1. Dworkin RH, Turk DC, Farrar JT, et al. Core outcome measures for chronic pain clinical trials: IMMPACT recommendations. Pain 2005; 113:9–19.
2. Clement RC, Welander A, Stowell C, et al. A proposed set of metrics for standardized outcome reporting in the management of low back pain. Acta orthopaedica 2015; 86:523–533.
3. Chiarotto A, Terwee CB, Ostelo RW. Choosing the right outcome measurement instruments for patients with low back pain. Best Pract Res Clin Rheumatol 2016; 30:1003–1020.
4. Froud R, Patel S, Rajendran D, et al. A systematic review of outcome measures use, analytical approaches, reporting methods, and publication volume by year in low back pain trials published between 1980 and 2012: Respice, adspice, et prospice. PLoS One 2016; 11:e0164573.
5. Gianola S, Frigerio P, Agostini M, et al. Completeness of outcomes description reported in low back pain rehabilitation interventions: a survey of 185 randomized trials. Physiother Can 2016; 68:267–274.
6. Liberati A. Need to realign patient-oriented and commercial and academic research. Lancet 2011; 378:1777–1778.
7. Furlan AD, Pennick V, Bombardier C, Van Tulder M. Editorial Board, Cochrane Back Review Group. 2009 updated method guidelines for systematic reviews in the Cochrane Back Review Group. Spine (Phila Pa 1976) 2009; 34:1929–1941.
8. Furlan AD, Malmivaara A, Chou R, et al. 2015 Updated Method Guideline for Systematic Reviews in the Cochrane Back and Neck Group. Spine (Phila Pa 1976) 2015; 40:1660–1673.
9. Borghuis J, Hof AL, Lemmink KA. The importance of sensory-motor control in providing core stability: implications for measurement and training. Sports Med 2008; 38:893–916.
10. Brown SHM, McGill SM. The intrinsic stiffness of the in vivo lumbar spine in response to quick releases: implications for reflexive requirements. J Electromyogr Kinesiol 2009; 19:727–736.
11. Hartvigsen J, Lings S, Leboeuf-Yde C, Bakketeig L. Psychosocial factors at work in relation to low back pain and consequences of low back pain; a systematic, critical review of prospective cohort studies. Occup Environ Med 2004; 61:e2.
12. Tagliaferri SD, Miller CT, Owen PJ, et al. Domains of chronic low back pain and assessing treatment effectiveness: a clinical perspective. Pain Pract 2020; 20:211–225.
13. Wippert P-M, Puschmann A-K, Drießlein D, et al. Development of a risk stratification and prevention index for stratified care in chronic low back pain. Focus: yellow flags (MiSpEx network). Pain Rep 2017; 2:e623.
14. Lampert T, Kroll L, Müters S, et al. Messung des sozioökonomischen Status in der Studie zur Gesundheit Erwachsener in Deutschland (DEGS1). Bundesgesundheitsbl 2013; 56:631–636.
15. Von Korff M, Ormel J, Keefe FJ, Dworkin SF. Grading the severity of chronic pain. Pain 1992; 50:133–149.
16. Isaksson R-M, Holmgren L, Lundblad D, Brulin C, Eliasson M. Time trends in symptoms and prehospital delay time in women vs. men with myocardial infarction over a 15-year period. The Northern Sweden MONICA Study. Eur J Cardiovasc Nurs 2008; 7:152–158.
17. Niederer D, Vogt L, Wippert P-M, et al. Medicine in spine exercise (MiSpEx) for nonspecific low back pain patients: study protocol for a multicentre, single-blind randomized controlled trial. Trials 2016; 17:507.
18. Lee H, Mansell G, McAuley JH, et al. Causal mechanisms in the clinical course and treatment of back pain. Best Pract Res Clin Rheumatol 2016; 30:1074–1083.
19. Lee H, Hübscher M, Moseley GL, et al. How does pain lead to disability? A systematic review and meta-analysis of mediation studies in people with back and neck pain. Pain 2015; 156:988–997.
20. Mühlhauser Y, Vogt L, Niederer D. How and how fast does pain lead to disability? A multilevel mediation analysis on structural, temporal and biopsychosocial pathways in patients with chronic nonspecific low back pain. Musculoskelet Sci Pract 2020; 49:102199.
21. Verbrugghe J, Agten A, Stevens S, et al. Disability, kinesiophobia, perceived stress, and pain are not associated with trunk muscle strength or aerobic capacity in chronic nonspecific low back pain. Phys Ther Sport 2020; 43:77–83.
22. Hu H, Zheng Y, Wang X, et al. Correlations between lumbar neuromuscular function and pain, lumbar disability in patients with nonspecific low back pain: a cross-sectional study. Medicine (Baltimore) 2017; 96:e7991.
23. Steiger F, Wirth B, De Bruin ED, Mannion AF. Is a positive clinical outcome after exercise therapy for chronic non-specific low back pain contingent upon a corresponding improvement in the targeted aspect(s) of performance? A systematic review. Eur Spine J 2012; 21:575–598.
24. Verbrugghe J, Agten A, Stevens S, et al. Exercise intensity matters in chronic nonspecific low back pain rehabilitation. Med Sci Sports Exerc 2019; 51:2434–2442.
25. Wippert P-M, Puschmann A-K, Arampatzis A, Schiltenwolf M, Mayer F. Diagnosis of psychosocial risk factors in prevention of low back pain in athletes (MiSpEx). BMJ Open Sport Exerc Med 2017; 3:e000295.
26. Linton SJ, Flink IK, Vlaeyen JWS. Understanding the etiology of chronic pain from a psychological perspective. Phys Ther 2018; 98:315–324.
27. Wippert P-M, Puschmann A-K, Drießlein D, et al. Personalized treatment suggestions: the validity and applicability of the risk-prevention-index social in low back pain exercise treatments. J Clin Med 2020; 9:1197.
28. Pincus T, Kent P, Bronfort G, Loisel P, Pransky G, Hartvigsen J. Twenty-five years with the biopsychosocial model of low back pain-is it time to celebrate? A report from the twelfth international forum for primary care research on low back pain. Spine (Phila Pa 1976) 2013; 38:2118–2123.
29. Snelgrove S, Liossi C. Living with chronic low back pain: a metasynthesis of qualitative research. Chronic Illn 2013; 9:283–301.
30. Owen PJ, Miller CT, Mundell NL, et al. Which specific modes of exercise training are most effective for treating low back pain? Network meta-analysis. Br J Sports Med 2019; 54:1279–1287.
31. Niederer D, Engel T, Vogt L, et al. Motor control stabilisation exercise for patients with non-specific low back pain: a prospective meta-analysis with multilevel meta-regressions on intervention effects. J Clin Med 2020; 9:3058.
32. Mueller J, Niederer D. Dose-response-relationship of stabilisation exercises in patients with chronic non-specific low back pain: a systematic review with meta-regression. Sci Rep 2020; 10:16921.
33. Conway R, Behennah J, Fisher J, Osborne N, Steele J. Associations between trunk extension endurance and isolated lumbar extension strength in both asymptomatic participants and those with chronic low back pain. Healthcare (Basel, Switzerland) 2016; 4:
34. Boers M, Kirwan F R, Wells G, et al. Developing core outcome measurement sets for clinical trials: OMERACT filter 2.0. J Clin Epidemiol 2014; 67:745–753.
35. Chiarotto A, Deyo RA, Terwee CB, et al. Erratum to: Core outcome domains for clinical trials in non-specific low back pain. Eur Spine J 2015; 24:2097.
36. Chiarotto A, Deyo RA, Terwee CB, et al. Core outcome domains for clinical trials in non-specific low back pain. Eur Spine J 2015; 24:1127–1142.
37. Tong MH, Mousavi SJ, Kiers H, Ferreira P, Refshauge K, Van Dieën J. Is there a relationship between lumbar proprioception and low back pain? A systematic review with meta-analysis. Arch Phys Med Rehabil 2017; 98:120.e2–136.e2.
Keywords:

biomechanics; CNLBP; lumbalgia; measurement properties; outcome assessments

Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.