Archives of Physical Medicine and Rehabilitation
Volume 90, Issue 1 , Pages 109-117, January 2009

Is the Fear Avoidance Model Associated With the Reduced Level of Aerobic Fitness in Patients With Chronic Low Back Pain?

  • Rob J. Smeets, MD, PhD

      Affiliations

    • Rehabilitation Centre Blixembosch, Eindhoven, The Netherlands
    • School for Public Health and Primary Care, Maastricht University, Maastricht, The Netherlands
    • Rehabilitation Foundation Limburg, Hoensbroek, The Netherlands
    • Corresponding Author InformationReprint requests to Rob J. Smeets, MD, PhD, Rehabilitation Foundation Limburg, PO Box 88, 6430 AB Hoensbroek, The Netherlands
    • ,
  • Kees D. van Geel, PT

      Affiliations

    • Rehabilitation Centre Blixembosch, Eindhoven, The Netherlands
    • Fysiotherapie Beelen/Medisch trainings centrum Breukelen, Breukelen, The Netherlands
    • ,
  • Jeanine A. Verbunt, MD, PhD

      Affiliations

    • School for Public Health and Primary Care, Maastricht University, Maastricht, The Netherlands
    • Rehabilitation Foundation Limburg, Hoensbroek, The Netherlands
    • Department of Psychological Science, Maastricht University, Maastricht, The Netherlands

Article Outline

Abstract 

Smeets RJ, van Geel KD, Verbunt JA. Is the fear avoidance model associated with the reduced level of aerobic fitness in patients with chronic low back pain?

Objectives

To compare aerobic fitness of patients with chronic low back pain (CLBP) against healthy controls and to assess whether variables of the fear avoidance model are associated with loss of aerobic fitness.

Design

A case-comparison study.

Setting

Rehabilitation centers.

Participants

Patients with CLBP (n=223), and normative data from healthy subjects (n=18,082).

Interventions

Not applicable.

Main Outcome Measures

Maximal oxygen uptake (V̇o2max) was estimated on the basis of a modified submaximal Åstrand bicycle test performed by patients with CLBP (observed level of aerobic fitness) and compared with the normative data of healthy controls matched for age, sex, and level of sport activity (expected level of aerobic fitness). Pain (visual analog scale); disability (Roland Disability Questionnaire); pain-related fear (Tampa Scale for Kinesiophobia); depression (Beck Depression Inventory); catastrophizing (Pain Catastrophizing Scale); and the level of activity during sport, work/household, and leisure time (Baecke Physical Activity Questionnaire) were assessed. Multiple linear regression analysis was performed with the difference of the observed and expected level of aerobic fitness as dependent variable and putative influential factors including those of the fear avoidance model as independent variables.

Results

V̇o2max could be calculated in 175 (78%) of the patients. Both men and women with CLBP had significant lower V̇o2max than expected (10.3mL/kg lean body mass (LBM)×min−1 and 6.5mL/kg LBM×min−1, respectively; P<.001). The levels of activity during leisure time and work/household were significantly associated with this reduced level of aerobic fitness. However, the variables of the fear avoidance model were not.

Conclusions

Most patients with CLBP-associated disability have a lower level of aerobic fitness, but this is not associated with fear avoidance.

Key Words: Exercise, Low back pain, Rehabilitation

List of Abbreviations: BDI, Beck Depression Inventory, BPAQ, Baecke Physical Activity Questionnaire, CLBP, chronic low back pain, LBM, lean body mass, LBP, low back pain, V̇o2max, maximal oxygen consumption, MET, metabolic equivalent, PCS, Pain Catastrophizing Scale, RDQ, Roland Disability Questionnaire, TSK, Tampa Scale for Kinesiophobia, VIF, variance inflation factor

 

LOW BACK PAIN IS AN INCREASING medical and social problem. In 2000, the direct costs involved in LBP in The Netherlands were approximately €337 million. By including indirect costs caused by incapacity for work, costs were €1.7 billion.1 One explanation for pain becoming chronic in some patients with LBP is given by the fear avoidance model (fig 1).2

According to this model, patients may interpret their pain as threatening (catastrophizing about their pain), and this can result in pain-related fear, of which fear of movement/(re)injury is the most salient. Both this fear and expectation of adverse consequences from increasing activities may cause avoidance of physical activities. In the long run, avoidance behavior might result in disability, depression, and disuse. Disuse is defined as loss of physical activities in daily life, of which loss of aerobic fitness is a consequence.3 There is increasing evidence that LBP-related fear is more disabling than pain itself.4, 5 However, it is debated whether physical deconditioning in patients with CLBP really exists.6 In addition, evidence to confirm the hypothesized role of fear avoidance in disuse and physical deconditioning is still inconclusive.7

In an earlier study, we found that patients with CLBP, especially men, had a lower level of aerobic capacity than healthy controls matched for age, sex, and level of sport activity.8 However, in contrast with the hypothesis based on the fear avoidance model, this loss of aerobic capacity appeared not to be associated with pain-related fear, assessed by the TSK. This might be because, within the concept of fear avoidance, no distinction was made between activity avoidance and somatic focus. This distinction seems relevant where activity tolerance is more associated with physical performance than somatic focus.9 In the analysis, the influence of these 2 factors as well as 2 other important variables of the fear avoidance model (catastrophizing and depression) was not investigated. The main reason for not including these variables was the relatively low number of patients (n=84), which restricts the number of putative influential variables that can be incorporated in the multiple regression analysis.

Furthermore, based on physiologic principles that aerobic capacity is influenced by the level of physical activity,10 it was remarkable that the level of physical activity during work and household was not significantly associated with this reduced level of aerobic fitness. Again, this might be a result of the relatively low number of patients.

In this study, we collected data from sufficient patients with CLBP-associated disability to assess whether they have a lower level of aerobic fitness than healthy controls matched for age, sex, and sport activity, and whether this loss is associated with putative influential variables from the fear avoidance model. In order to get more insight in the way patients handle a straining physical activity, we also assessed whether patients who quit the submaximal test prematurely differed from successful completers on variables of the fear avoidance model.

Back to Article Outline

Methods 

Participants and Settings 

This study is part of a research project on the evaluation of rehabilitation treatment in CLBP.11 In this project, 223 patients referred by their general practitioner or medical specialist between April 2002 and December 2004 agreed to participate in a randomized controlled trial to evaluate the effectiveness of different treatments for CLBP disability. Patients were treated in 3 rehabilitation centers in the province of Noord-Brabant in The Netherlands: rehabilitation centre Breda, rehabilitation centre Leijpark in Tilburg, and rehabilitation centre Blixembosch in Eindhoven. The inclusion criteria were (1) first referral to a rehabilitation center to reduce disability caused by nonspecific LBP existing for at least 3 months, (2) age between 18 and 65 years, and (3) ability to walk at least 100m. Exclusion criteria were lumbar disk herniation with neurologic symptoms; inflammatory or neoplastic disease; fracture of the spine; spondylolisthesis or spondylolysis; lumbar spondylodesis; major psychiatric, cardiac, and/or pulmonary conditions making physical training impossible; severe addiction to drugs, narcotics, or alcohol; and pregnancy.

All data for the current study were obtained during a prerandomization testing session. Patients completed questionnaires and performed a modified submaximal Åstrand bicycle test and several performance tasks. Patients were requested not to perform heavy physical activities such as cycling to the rehabilitation center and to restrict their alcohol use to a maximum of 2 glasses over the 24 hours before testing. They were also not allowed to drink any coffee, tea, or soda 3 hours before the testing and were not allowed to use any illegal drugs such as marijuana.12 They were allowed to take their regular medication. All patients gave written informed consent to participate. The medical ethics committee of the Rehabilitation Foundation Limburg and Institute for Rehabilitation Research at Hoensbroek in The Netherlands approved the research protocol.

Testing Procedure 

Modified submaximal Åstrand bicycle test 

Although aerobic fitness is often corrected for total body weight in studies on LBP, body fat is metabolically inert, consuming essentially no oxygen, and can represent a significant portion of total weight. Hence, V̇o2max corrected for LBM provides a more refined discriminator of aerobic fitness.13 V̇o2max in mL/kg LBM×min−1 was estimated by a modified submaximal Åstrand bicycle test. In this test, especially developed for patients with CLBP by the Department of Rehabilitation of the University Hospital Groningen in The Netherlands, the workload is gradually increased based on the LBM.14 The test-retest reliability (r=.87) and the validity (r=.84) of this test are reasonable compared with a maximal exercise test.14, 15 The participants performed the test on a calibrated cycle ergometer.a, b, c The heart rate was monitored by a monitor placed on the patient's chest.d

The patients started cycling with a workload of 0.5W/kg LBM at a constant rate of 60rpm. After 2 minutes, the workload was increased to 1.5W/kg LBM. If the heart rate was still below 120 beats a minute, the workload was increased to 2.0W/kg LBM and, when necessary, to 2.5W/kg LBM. When the heart rate exceeded 120 beats a minute, the patient cycled 6 minutes with a fixed workload to reach a steady-state phase, meaning that the heart rate did not vary more than 5 beats a minute during the last minute of exercise. The average heart rate during the last minute was calculated. The V̇o2max was estimated by using the Åstrand nomogram with a correction factor for age, based on the linear association between heart rate and increase in oxygen uptake.16 If the heart rate during the last minute varied more than 5 beats a minute, no V̇o2max was estimated. The test was stopped if the patient did not reach a heart rate of at least 120 beats a minute, the heart rate exceeded the predefined rate (220 – age)×.85, or the blood pressure reached the level of 220/115mmHg (in order to prevent cardiovascular overburdening15), or if the patient showed signs of serious cardiovascular or pulmonary difficulties.

LBM of a patient was determined based on the assessment of a patient's fat percentage according to the protocol of Durnin and Womersley17 by using a skin fold caliper.e

Measures 

Demographic data and clinical characteristics 

For each patient, weight, length, age, sex, duration of complaints and disability, extent of radiation of pain to the leg (no, above knee, below knee), traumatic onset of pain, level of education, and history of back surgery were recorded.

Pain 

A 100mm-long visual analog scale with no pain on the left side and unbearable pain on the right side was used to measure pain intensity. Relevance, validity, and reliability have been sufficiently tested for patients with LBP.18, 19

Perceived disability 

The Dutch version of the RDQ measures perceived disability arising from LBP. The questionnaire consists of 24 items with yes or no answers and a total score ranging from 0 to 24. The higher the score, the more disabled a patient is. The RDQ is sufficiently valid and reliable in CLBP.20, 21

Fear of injury 

For measuring the extent to which a patient experiences fear of injury, the Dutch version of the TSK was used. The TSK measures fear of (re)injury and movement, and consists of 17 items with a total score ranging from 17 to 68. A higher score indicates more fear. The questionnaire is considered reliable and valid in CLBP.22, 23 The TSK contains 2 subscales: activity avoidance and somatic focus.22 The first subscale assesses the catastrophizing thoughts that activity might result in increasing the extent of pain or (re)injury, whereas the second subscale reflects a belief in underlying and serious medical problems. In the current study, the 2 subscales of the TSK are used separately in order to assess activity avoidance and somatic focus.

Level of activity during sports, at work, and during leisure time 

The level of activity at work, during sports, and during leisure time was measured by the BPAQ. The main goal of this 16-item self-reported questionnaire is to assess a person's daily physical activity. The BPAQ consists of 3 subscales: work or school, sports, and leisure time activities index.24 For each question, the patient is asked to score on a 5-point Likert-scale ranging from “never” to “always” or “very often.” The patient also reports his main occupation, which is then categorized as light, moderate, or heavy work according to the level of energy expenditure, so that a work intensity score can be calculated.

The work index is calculated by adding the work intensity score to the score of 7 questions regarding work-related sitting, standing, walking, heavy lifting, tiredness, sweating, and comparison with others of the same age, and dividing this total score by 8. When a patient had not worked in the preceding 3 months, the work index was set at 0. The leisure time score is calculated by summing the scores of 4 questions about watching television, walking, and cycling and dividing this score by 4. The test-retest reliability for patients with LBP is moderate to good (intraclass correlation coefficient for Work Index=.90, Sport Index=.71, and Leisure Time Index=.70).25 The validity in healthy persons is well established.26

Because household activities might be responsible for maintaining a higher level of aerobic fitness,27, 28, 29 and because the number of working hours and the hours of household activities are not assessed in the BPAQ, patients were additionally asked how many hours a week they worked and how many hours they performed household activities. In order to determine whether household and work are responsible for maintaining a higher level of aerobic capacity, we decided to use an alternative work index (BPAQ-Alternative Work Index). The calculation of the BPAQ-Alternative Work Index is in accordance with the calculation of the BPAQ-Sport Index, based on multiplying the work intensity score with the amount of hours of work and household activities a week. Based on physiologic principles, all work and household activities performed during the 3 months previous to the testing contributed to the BPAQ-Alternative Work Index.8 For example, when a person was off work for 4 weeks, he was assigned the full score for 8 weeks and no score for 4 weeks on the alternative work index.

Pain catastrophizing 

Pain catastrophizing was measured by the PCS.30 This questionnaire consists of 13 statements that describe thoughts and feelings people can experience while they are in pain. The participants are asked to reflect on past painful experiences and indicate how frequently specific thoughts occurred on a 5-point scale with the end points 0 (not at all) and 4 (all the time). The 13 items together form 1 scale for the extent of catastrophizing thoughts about pain. The psychometric properties as tested in a population of patients with CLBP appeared to be good.31, 32

Depression 

Depression was measured by the BDI.33 The BDI is a self-administered 21-item self-report scale that measures the extent of depression. The lowest score is 0, and the highest is 63. The reliability and validity of the BDI are considered to be good to high.34, 35

Controls 

In the current study, normative data on healthy subjects were derived from a Dutch database. Data on the aerobic fitness level of healthy Dutch people from all over the country were gathered between 1991 and 2003 using the submaximal Åstrand bicycle test.36 This database includes 18,082 healthy people age 17 to 70 years. They were classified into groups specified by age (range, 3–5y), sex, and the level of sport activity during the preceding 6 months. People undertaking 1 to 2 hours of sport weekly during the preceding 6 months were categorized as recreational, and those sporting less than that or not at all were categorized as untrained. In the normative dataset, mean values and SDs of V̇o2max in mL/kg LBM×min−1 are available for groups stratified by sex, age, and level of sport activity. In order to compare the patients with the normative values, the patients were classified likewise into 1 of the 3 sport activity categories on the basis of their reported level of sport activities during the preceding 6 months.

Statistical Analysis 

For all variables, the mean and SD were calculated. Because only the mean and SD of V̇o2max of the healthy controls were available per group (formed on the basis of age, sex, and level of sport activity) and not for each healthy subject individually, matching on an individual basis was not possible. Therefore, for each patient, the following formula was used:

For example, for a 32-year-old woman with CLBP performing no sports, the Zpatient could be calculated by subtracting the mean score of the group of healthy untrained women, aged 31 to 35, from her observed V̇o2max and divide this by the SD of this mean score of this normative group.

Next, all Zpatient scores were summarized (Ztotal), and SDZtotal and SEZtotal(SEZtotal) were calculated. To test the null hypothesis that patients have the same level of aerobic fitness as their healthy controls (Ztotal=0), a 1-sample Student t test with a 2-sided α of .05 was performed. Based on an estimated SD of 13.05 (50% higher than the mean SD in healthy controls), 196 patients were needed to give a power of 90% to detect a real difference of 5mL/kg LBM×min−1.

In addition to the aerobic fitness expressed in mL/kg LBM×min−1, the V̇o2max mL/kg×min−1 was calculated. Based on this V̇o2max in mL/kg×min−1, a related MET can be calculated in order to interpret the clinical relevance of the difference found; 1 MET is the amount of energy (oxygen) used by the body in rest (eg, talking on the phone or reading a book). Increasing effort will result in an increase in MET score. Most household activities score 2 METs more than standing or working at a desk. Activities based on 3 to 6 METs are considered to be of moderate intensity.16

Comparisons between 2 groups (men and women with CLBP, those who completed the test and those who did not because of fatigue or pain) were performed by using an independent-samples t test for unpaired observations with a normative distribution of continuous data. In case of nonnormative distribution, the Mann Whitney U test was used. For nominal and ordinal data, the chi-square test was used.

Multiple linear regression analysis (stepwise regression with backward elimination of nonsignificant independent variable with the highest P value) was performed to define the contribution of independent variables to the difference in observed versus expected level of aerobic fitness in patients with CLBP (Zpatient), which is equivalent to the loss of aerobic fitness. This Zpatient score was the dependent variable. Based on the fear avoidance model, the level of pain intensity, catastrophizing, pain-related fear assessed by the activity avoidance and somatic focus scales of the TSK, disability, depression, and the level of activity during leisure time and work including household activities (alternative work index) were selected as independent variables. Because it might be hypothesized that the duration of experienced disability is associated with the level of physical inactivity as well as the level of aerobic deconditioning, duration of experienced disability was added as an independent variable.

Standardized β coefficients and significance were tested under the null hypothesis that the coefficient did not differ from 0. For performing the multiple linear regression analysis with 9 independent variables, the number of the variables times 10 (n=90) are needed, as recommended for multiple regression analysis by Dawson-Saunders and Trap.37 At each step during the regression analysis, a collinearity check was done. Collinearity is considered a problem when the VIF is above 3. All statistical analyses were performed with SPSS software, version 11.5.f

Back to Article Outline

Results 

Sample Characteristics 

Demographic data and complaint characteristics of the 223 participating patients and specified for men and women separately are shown in table 1. Men differed significantly from women regarding weight, LBM, length, duration of LBP, duration of LBP-associated disability, level of education, BPAQ work, hours of sports, and household activities. There were no further significant sex-related differences. Furthermore, about 70% of the patients were on partial sick leave because of LBP or were not working at all. Only 34% were active in any kind of sport.

Table 1. Demographic Data and Clinical Characteristics
CharacteristicsTotal (N=223)Men (n=118)Women (n=105)
Weight (kg)81±16.0(49–134)88±15.2(57–134)74±13.4(49–110)†
LBM (kg)55.6±11.2(35–85)63.7±8.2(45–85)46.9±6.3(35–65)†
Length (cm)176±9(154–200)182±6.5(170–200)169±6.1(154–184)†
Age (y)42±10.0(20–64)43±9.1(20–61)40±10.9(20–64)
Duration of LBP (mo)⁎24(12–72)31(14–84)18(11–60)†
Duration of disability (mo)⁎17(8–40)20(10–48)12(8–36)†
Trauma preceding LBP (%)18.419.517.1†
Previous back surgery (%)15.214.416.2
Radiation (%)
No13.916.910.5
Above the knee36.839.833.3
Below the knee49.343.256.2
Pain intensity (mm)49±24.7(1–99)46±25.2(1–99)52±24.0(1–99)
Perceived disability (RDQ)13.8±3.8(2–21)13.8±3.9(3–21)13.9±3.6(2–20)
Education (%)
Low62.371.052.4†
Middle36.828.046.7†
High0.910.8†
Work (%)
Full time job29.628.830.5
Partial leave23.826.321.0
Full leave38.139.836.2
No job8.55.112.4
Work (h)⁎7(0–26)13(0–36)6(0–18.5)†
Household (h)⁎0(0–10)0(0–0.50)10(0–20)†
Sports (%)343632
Level of activity
BPAQ-Work2.4(0–3.3)2.3(0–3.3)2.6(0–3.4)†
BPAQ-Leisure Time⁎2.8(2.3–3.3)2.8(2.3–3.3)2.8(2.5–3.5)
BPAQ-Sport⁎0(0–1.7)0(0–1.3)0(0–1.7)
BPAQ-AWI⁎17.0(0–31.3)15.2(0–37.2)18.2(7.6–30.4)

NOTE. Values are mean ± SD (range) unless otherwise noted.

Abbreviation: BPAQ-AWI, Baecke Physical Activity Questionnaire-Alternative Work Index.

⁎The median and the first and third quartile are presented.

†P<.05.

Measuring Aerobic Fitness 

Of all the 223 participants who started the Åstrand submaximal bicycle test, 180 completed the test according to the testing protocol. However, 5 of these 180 participants were using β-blockers and were therefore excluded from further analysis. A total of 175 participants remained. Forty-three participants did not complete the bicycle test, and the reasons for not completing are listed in table 2. The main reported reason was fatigue or pain.

Table 2. Usability of the Modified Åstrand Submaximal Test
Using β-BlockerÅstrand Test CompletedÅstrand Completed According to ProtocolUnusable Åstrand
No (n=213)Yes175
Maximum heart rate 6
Exceeded maximum blood pressure 3
RPM too low 4
Other medical reason 2
Patient stopped prematurely because of pain/fatigue 21
Other 2
Yes (n=10)Yes 5
RPM too low 2
Patient gave up 3

Abbreviation: RPM, rotations per minute.

Level of Aerobic Capacity 

As mentioned, V̇o2max could be calculated for 175 people (78%). The results are presented in table 3. The patients with CLBP had a significantly lower level of aerobic fitness compared with their matched healthy controls, with a mean loss of V̇o2max of 10.3mL/kg LBM×min−1 (21%) in men and 6.5mL/kg LBM×min−1 (13%) in women. Men with CLBP had a significantly greater difference in observed versus expected level of aerobic fitness than women with CLBP (P<.001). Furthermore, 159 patients (91%) had a lower level of aerobic fitness as predicted based on sex, age, and level of sport activity.

Table 3. Results of the Modified Åstrand Submaximal Test in LBM and z Scores
Patients With CLBPObserved Vo2max (mL/kg LBM×min−1)Expected Vo2max (mL/kg LBM×min−1)
Total (n=175)41.3±9.1(15.3 to 68.0)49.8±4.4(33.5 to 60.5)
Men (n=92)38.9±8.8(15.3 to 68.0)49.2±4.5(33.5 to 59.7)
Women (n=83)43.9±8.8(21.2 to 63.1)50.4±4.2(41.6 to 60.5)
Patients With CLBPMean z Score (95% CI)Significance
Total (n=175)−0.90(−1.01to−0.78)P<.001
Men (n=92)−1.08(−1.23to−0.93)P<.001
Women (n=83)−0.68(−0.84to−0.52)P<.001
Within patients with CLBP, difference between men and women0.40(0.18to0.62)P<.001

NOTE. Values are mean ± SD (range) for Vo2max data. Expected Vo2max is based on normative data, with each patient matched for age, sex, and sports.

Abbreviation: CI, confidence interval.

In addition, in order to calculate the differences in METs, the V̇o2max in mL/kg×min−1 was used, and the results are presented in table 4.

Table 4. Results of the Modified Åstrand Submaximal Test in Kilograms and METs
Patients with CLBPObserved Vo2max (mL/kg×min−1)Expected Vo2max (mL/kg×min−1)Difference in METs
Total (n=175)28.8±7.2(10.2–52.0)36.0±5.0(25.0–49.4)2.1
Men (n=92)29.2±7.8(10.2–52.0)37.5±5.2(25.0–49.4)2.4
Women (n=83)28.5±6.5(11.6–41.7)34.4±4.2(25.6–44.2)1.7

NOTE. Values are mean ± SD and range for Vo2max data. Expected Vo2max is based on normative data, with each patient matched for age, sex, and sports.

Patients who stopped the test prematurely because of fatigue or pain reported higher levels of pain and disability and a lower level of activity in sport and leisure time than participants who completed the test (table 5). Other factors, especially catastrophizing, activity avoidance, somatic focus, and depression, which are relevant factors in the fear avoidance model, did not differ significantly between these 2 groups.

Table 5. Comparison of Patients Who Completed Versus Patients Who Prematurely Stopped the Modified Åstrand Submaximal Test
CharacteristicsCompleted Test (n=175)Prematurely Stopped Test Because of Fatigue/Pain (n=21)P
Sex (% male)5357.692
Age (y)40.0±10.0(20–61)44.2±7.5(30–57).063
Length (cm)175.7±8.9(154–200)174±9.4(158–191).421
Weight (kg)80.1±14.8(49–134)80.9±20.0(52–133).827
Duration of pain (mo)⁎24(12–72)31(10.5–120).592
Duration of disability (mo)⁎17(9–42)24(7.5–92).632
Back surgery (%)1519.614
Trauma preceding (%)1919.967
Radiation (%) .229
No159
Above the knee3824
Below the knee4767
Pain intensity (mm)
VAS47.6±24.1(1–95)66.1±23.6(16–99).001
Perceived disability
RDQ13.5±3.8(2–21)16.1±3.0(9–20).003
Fear of injury
TSK AA18.8±4.2(9–30)18.9±3.9(11–25).905
TSK SF11.0±3.1(5–19)11.7±2.5(7–16).342
Depression
BDI⁎8(5–13)9(5.5–14).650
Pain catastrophizing
PCS⁎22(14–29)25(16–34.5).207
Level of activity
BPAQ-Work⁎2.4(0–3.3)2.4(0–3.8).974
BPAQ-Sport2.2±0.71(1–4)1.6±0.37(1–2.5)<.001
BPAQ-Leisure Time2.9±0.68(1.3–4.3)2.52±0.58(1–3.3).020
BPAQ-AWI⁎18.2(0–33.4)15.2(2.7–25.1).294

NOTE. Values are mean ± SD (range) or as otherwise noted.

Abbreviations: BPAQ-AWI, Baecke Physical Activity Questionnaire-Alternative Work Index; TSK AA, Tampa Scale for Kinesiophobia Activity Avoidance; TSK SF, Tampa Scale for Kinesiophobia Somatic Focus; VAS, visual analog scale.

⁎Values are median and first and third quartile.

Determinants of the Difference Between Observed and Expected Level of Aerobic Fitness 

Table 6 shows the results of the multiple regression analysis with the difference between observed and expected level of aerobic fitness (Zpatient score) as the dependent variable with stepwise eliminating the nonsignificant variable with the highest P value. Block 1 represents the starting model and block 8 the final model after elimination of all nonsignificant variables. Only the level of activity during leisure time and the alternative work index were significantly associated with the difference between the observed and expected level of aerobic fitness, explaining 11% of the total variance. However, the loss of V̇o2max in this population of patients with CLBP could not be explained by variables from the fear avoidance model such as catastrophizing, fear-related avoidance, and somatic focus, nor by the duration of disability. The variance inflation factors of the multiple regression analysis were within the acceptable range (VIF<3), which suggests no problems with collinearity.

Table 6. Summary of Multiple Stepwise Backward Regression Analysis For Z as Dependent Variable and Activity Avoidance, Somatic Focus, Depression, Pain Catastrophizing, Perceived Disability, Pain Intensity, Level of Activity in Sport and Leisure Time, Alternative Work Index, and Duration of Disability as Independent Variables
Model StepsIndependent VariableStandardized βR2F RatioP
Block 1(Constant) .133.007.001
Fear of injury activity avoidance (TSK AA).010 .917
Fear of injury somatic focus (TSK SF)−.062 .496
Depression (BDI).093 .326
Pain catastrophizing (PCS)−.005 .957
Perceived disability (RDQ)−.132 .172
Pain intensity (VAS).048 .589
Level of activity leisure time (BPAQ).300 <.001
Duration of disability−.043 .575
Alternative Work Index.157 .095
Block 8(Constant) .1120.000<.001
Level of activity leisure time (BPAQ).299 <.001
Alternative Work Index.189 .012

Abbreviations: TSK AA, Tampa Scale for Kinesiophobia Activity Avoidance; TSK SF, Tampa Scale for Kinesiophobia Somatic Focus; VAS, visual analog scale.

Back to Article Outline

Discussion 

Using the Åstrand submaximal bicycle test allowed estimation of the V̇o2max in 78% of the patients with CLBP, which is in agreement with our previous study.8 All patients had CLBP-associated disability and were referred for multidisciplinary treatment. Ninety-one percent of this specific sample had a lower level of aerobic fitness compared with healthy controls matched for age, sex, and level of sport activity during the preceding 6 months. Both men and women with CLBP had a significantly lower level of aerobic fitness, but men had a significantly lower level than women. This higher loss of aerobic capacity in men than women was also reported by Nielens and Plaghki.29 Keeping in mind that most household activities cost 2 METs more than resting, the differences between observed and expected aerobic fitness for male patients of 2.4 METs and for female patients of 1.7 METs seem clinically relevant.

Patients who stopped the Åstrand submaximal bicycle test prematurely because of fatigue or pain reported higher levels of pain and disability before the test started and had a lower level of activity in sport and leisure time than those who completed the test successfully. In addition, catastrophizing and fear-related avoidance activity or somatic focus were not associated with premature test ending. The finding that a higher level of pain before testing seems to be associated with prematurely quitting the bicycle test was noted by de Gier et al.38 In their study, patients with musculoskeletal pain reporting a higher pretest pain intensity stopped earlier in an endurance test (pulling a weight down with 1 hand). This finding suggests that baseline pain determines the level of physical performance the patient is willing to perform, regardless of pain increase during testing and threat-reducing instructions. It could be that patients with more pain, consciously or unconsciously, anticipate an increase in pain and protect themselves by prematurely ending the bicycle test. The finding that patients who did not complete the bicycle test because of fatigue or pain also reported a lower level of activity in sport and leisure time and a higher level of disability might even indicate that this protective behavior is not only present during submaximal cardiovascular testing but reflects activity intolerance in daily life situations. However, this hypothesis can only be confirmed by monitoring activity in daily life using a longitudinal design.

The primary aim of the study was to assess whether putative influential factors of the fear avoidance model as well as duration of disability are associated with the loss of aerobic fitness in patients with CLBP-associated disability. It appeared that only the level of physical activity during leisure time as well as work including household were significantly associated with the difference between observed and expected level of aerobic fitness. All other putative influential variables, especially catastrophizing and pain-related fear (activity avoidance and somatic focus), failed to show a significant association. Thus there was no evidence to support the validity of the fear avoidance model and aerobic deconditioning in this sample of patients with CLBP. By including 175 patients, we ensured sufficient power to perform the linear regression analysis, and we are quite confident that our conclusions are robust. Nevertheless, this is only a cross-sectional study, and the results should be treated with caution when considering the cause-and-effect relationship between the postulated factors and the reduction of aerobic fitness.

Study Limitations 

As stated previously, this is only a cross-sectional study and the results should be treated with caution when considering the cause-and-effect relationship between the postulated factors and the reduction of aerobic fitness.

Another potential weakness of the study is the fact that patients with CLBP were tested by using a modified Åstrand submaximal test specifically developed to test patients with CLBP and proven to be reliable and valid,14, 15 whereas the healthy controls were tested using the Åstrand submaximal test.16, 39 In both tests, the patient has to reach a steady-state phase in which the heart rate does not vary more than ±5 beats a minute. The calculation of the mean heart rate during the last minute of this steady-state phase and the extrapolation of the corresponding V̇o2max are exactly the same for both tests. The only difference is that in the modified test, the final workload is reached in 2 to 6 minutes instead of the 1 to 2 minutes used in the Åstrand test. Although we did not perform a comparability study of both tests in patients with CLBP, it is not expected that the slower increase of workload leads to a higher or lower mean heart rate during the steady-state phase. Based on this, we believe that the results of the calculated V̇o2max of both tests are comparable.

Because V̇o2max scores of the healthy controls were available only per category of sport activity level, we had to match for this variable. Because V̇o2max has a moderate correlation with the sport activity level in healthy persons,40 this matching might have minimized our results. Nevertheless, and in contrast with our previous findings, the level of activity during leisure time and the alternative work index were both significantly associated with the decreased level of aerobic fitness. These results are in line with physiologic training principles, indicating that a certain intensity and duration of physical activity are necessary to sustain or increase aerobic capacity.41

A possible explanation for the significant lower level in aerobic fitness for men than women with CLBP (20% lower vs 13% lower than normative population) may be the higher level of household activities performed by women (P<.001), confirming the hypothesis of Nielens and Planghki27, 28, 29 that women are generally more active once they are at home because they are still engaged in childcare and various household duties.

By combining physical activities during work and household while calculating the Alternative Work Index, it appeared that women seemed more active than men, although this difference was not statistically significant. Because of anthropometric differences between men and women, this slightly different alternative work index score could result in a relative higher cardiovascular workload in women, which might explain their smaller loss of aerobic capacity.

It can be concluded that these patients with CLBP-associated disability are aerobically deconditioned and do have a lower level of physical activity during leisure time and work/household work. Hence the reduction of activity level seems to be an important factor in the development of this deconditioning, but it seems not to be associated with avoidance because of fear. The finding of a lower physical fitness level for patients with CLBP is in contrast with studies showing that patients with CLBP who report being severely disabled and less active still perform activities on a level comparable with the mean population.42, 43 One explanation might be that the intensity (lower speed, less power, and so forth), duration, and frequency of the activities (decline of activities) performed by patients were much lower than before the onset of LBP. Unfortunately, in the current study, no information was gathered by questionnaires regarding the quality and changes, and more specifically the decline of activities.

Another explanation might be that before developing CLBP, patients already are deconditioned compared with healthy controls, especially because 40% of the level of V̇o2max is explained by genetic factors.10 Nevertheless, a lower aerobic fitness level or physical inactivity is not a risk factor for developing CLBP.44, 45 For future research, we urge incorporation of the monitoring of activities in daily life, preferably combined with methods to measure the total energy consumption.42, 46, 47 The changes in activities can be measured only with self-report methods.

Back to Article Outline

Conclusions 

Most patients with CLBP-associated disability have a lower level of aerobic fitness than healthy controls matched for age, sex, and level of sport activity in the preceding 6 months, but this is not associated with fear avoidance. To investigate further the development and impact of loss of the level of aerobic fitness, longitudinal studies should be performed in patients with acute LBP.

Suppliers

Back to Article Outline

Acknowledgments 

The authors would like to thank the rehabilitation centers Breda, Leijpark, and Blixembosch and the many people who contributed to the successful completion of this study, particularly the participating patients, the rehabilitation center contact persons, therapists, and research assistants.

Back to Article Outline

References 

  1. Picavet HS, Schouten JS. Musculoskeletal pain in the Netherlands: prevalences, consequences and risk groups: the DMC(3)-study. Pain. 2003;102:167–178
  2. Vlaeyen JW, Linton SJ. Fear-avoidance and its consequences in chronic musculoskeletal pain: a state of the art. Pain. 2000;85:317–332
  3. Verbunt JA, Seelen HA, Vlaeyen JW, et al. Disuse and deconditioning in chronic low back pain: concepts and hypothesis on contributing mechanisms. Eur J Pain. 2003;7:9–21
  4. Crombez G, Vlaeyen JW, Heuts PH, Lysens R. Pain-related fear is more disabling than pain itself: evidence on the role of pain-related fear in chronic back pain disability. Pain. 1999;80:329–339
  5. Rainville J, Ahern DK, Phalen L, Childs LA, Sutherland R. The association of pain with physical activities in chronic low back pain. Spine. 1992;17:1060–1064
  6. Smeets RJ, Wittink H. The deconditioning paradigm for chronic low back pain unmasked?. Pain. 2007;130:201–202
  7. Verbunt JA, Seelen HA, Vlaeyen JW, van der Heijden GJ, Knottnerus JA. Fear of injury and physical deconditioning in patients with chronic low back pain. Arch Phys Med Rehabil. 2003;84:1217–1232
  8. Smeets RJ, Wittink H, Hidding A, Knottnerus JA. Do patients with chronic low back pain have a lower level of aerobic fitness than healthy controls? (Are pain, disability, fear of injury, working status, or level of leisure time activity associated with the difference in aerobic fitness level?). Spine. 2006;31:90–97
  9. Roelofs J, Goubert L, Peters ML, Vlaeyen JW, Crombez G. The Tampa Scale for Kinesiophobia: further examination of psychometric properties in patients with chronic low back pain and fibromyalgia. Eur J Pain. 2004;8:495–502
  10. McArdle WD, Katch FI, Katch VL. Exercise physiology: energy, nutrition, and human performance. In: Baltimore: Lippincott, Williams and Wilkins; 2001;p. 22–248
  11. Smeets RJ, Vlaeyen JW, Hidding A, et al. Active rehabilitation for chronic low back pain: cognitive-behavioral, physical, or both? (First direct post-treatment results from a randomized controlled trial [ISRCTN22714229]). BMC Musculoskel Disord. 2006;7:5
  12. Avakian EV, Horvath SM, Michael ED, Jacobs S. Effect of marihuana on cardiorespiratory responses to submaximal exercise. Clin Pharmacol Ther. 1979;26:777–781
  13. Milani RV, Lavie CJ, Mehra MR, Ventura HO. Understanding the basics of cardiopulmonary exercise testing. Mayo Clin Proc. 2006;81:1603–1611
  14. Schotanus J. Het meten van belastbaarheid bij patiënten met chronische lage rugklachten. Groningen: Rijksuniversiteit Groningen; 1999;
  15. Hodselmans AP, Jaegers SM, Goeken LN. Short-term outcomes of a back school program for chronic low back pain. Arch Phys Med Rehabil. 2001;82:1099–1105
  16. Astrand P, Rohdahl K. Textbook of work physiology, physiological bases of exercise. In: 3rd ed.. New York: McGraw-Hill; 1986;p. 354–391
  17. Durnin JV, Womersley J. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged 16 to 72 years. Br J Nutr. 1974;32:77–97
  18. Carlsson AM. Assessment of chronic pain, I: aspects of the reliability and validity of the visual analog scale. Pain. 1983;16:87–101
  19. Sriwatanakul K, Kelvie W, Lasagna L, Calimlim JF, Weis OF, Mehta G. Studies with different types of visual analog scales for measurement of pain. Clin Pharmacol Ther. 1983;34:234–239
  20. Roland M, Morris R. A study of the natural history of back pain, part I: development of a reliable and sensitive measure of disability in low-back pain. Spine. 1983;8:141–144
  21. Beurskens AJ, de Vet HC, Koke AJ. Responsiveness of functional status in low back pain: a comparison of different instruments. Pain. 1996;65:71–76
  22. Goubert L, Crombez G, Vlaeyen JW, Van Damme S, Van den Broeck A, van Houdenhove B. De Tampa schaal voor kinesiofobie: psychometrische karakteristieken en normering. Gedrag Gezond. 2000;28:54–62
  23. Vlaeyen JW, Kole-Snijders AM, Boeren RG, van Eek H. Fear of movement/(re)injury in chronic low back pain and its relation to behavioral performance. Pain. 1995;62:363–372
  24. Baecke JA, Burema J, Frijters JE. A short questionnaire for the measurement of habitual physical activity in epidemiological studies. Am J Clin Nutr. 1982;36:936–942
  25. Jacob T, Baras M, Zeev A, Epstein L. Low back pain: reliability of a set of pain measurement tools. Arch Phys Med Rehabil. 2001;82:735–742
  26. Philippaerts RM, Westerterp KR, Lefevre J. Doubly labelled water validation of three physical activity questionnaires. Int J Sports Med. 1999;20:284–289
  27. Nielens H, Plaghki L. Evaluation of physical adaptation to exercise of chronic pain patients by step-test procedure. Pain Clin 199;4:21-5.
  28. Nielens H, Plaghki L. Perception of pain and exertion during exercise on a cycle ergometer in chronic pain patients. Clin J Pain. 1994;10:204–209
  29. Nielens H, Plaghki L. Cardiorespiratory fitness, physical activity level, and chronic pain: are men more affected than women?. Clin J Pain. 2001;17:129–137
  30. Sullivan MJ, Bishop SR, Pivik J. The Pain Catastrophizing Scale: development and validation. Psychol Assess. 1995;7:524–532
  31. Van Damme S, Crombez G, Vlaeyen JW, Goubert L, Van den Broeck A, Van Houdenhove B. De Pain Catastrophizing Scale: Psychometrische karakteristieken en normering/The Pain Catastrophizing Scale: psychometric characteristics and norms. Gedragstherapie. 2000;33:209–220
  32. Osman A, Barrios FX, Gutierrez PM, Kopper BA, Merrifield T, Grittmann L. The Pain Catastrophizing Scale: further psychometric evaluation with adult samples. J Behav Med. 2000;23:351–365
  33. Beck AT, Rush AJ, Shaw BF, Emmery G. Cognitive therapy of depression. New York: Wiley and Sons; 1979;
  34. Beck AT, Steer RA, Garbin MG. Psychometric properties of the Beck Depression Inventory: twenty-five years of evaluation. Clin Psychol Rev. 1988;8:77–100
  35. Zitman FG, Griez EJ, Hooijer C. Standaardisering depressievragenlijsten. Tijdschr Psychiatr. 1989;31:114–135
  36. Vos JA. Ergometrie en trainingsbegeleiding. Amersfoort: Nederlands Paramedisch Instituut; 2004;
  37. Dawson-Saunders B, Trapp RG. Basic and clinical biostatistics. In: Norwalk: Appleton & Lange; 1998;p. 207–229
  38. de Gier M, Peters ML, Vlaeyen JW. Fear of pain, physical performance, and attentional processes in patients with fibromyalgia. Pain. 2003;104:121–130
  39. Keller A, Hellesnes J, Brox JI. Reliability of the isokinetic trunk extensor test, Biering-Sorensen test, and Astrand bicycle test: assessment of intraclass correlation coefficient and critical difference in patients with chronic low back pain and healthy individuals. Spine. 2001;26:771–777
  40. Jacobs DR, Ainsworth BE, Hartman TJ, Leon AS. A simultaneous evaluation of 10 commonly used physical activity questionnaires. Med Sci Sports Exerc. 1993;25:81–91
  41. American College of Sports Medicine. Position stand on the recommended quantity and quality of exercise for developing and maintaining cardiorespiratory and muscular fitness, and flexibility in adults. Med Sci Sports Exerc. 1998;30:975–991
  42. Verbunt JA, Westerterp KR, van der Heijden GJ, Seelen HA, Vlaeyen JW, Knottnerus JA. Physical activity in daily life in patients with chronic low back pain. Arch Phys Med Rehabil. 2001;82:726–730
  43. van den Berg Emons RJ, Schasfoort FC, de Vos LA, Bussmann JB, Stam HJ. Impact of chronic pain on everyday physical activity. Eur J Pain. 2007;11:587–593
  44. Battié MC, Bigos SJ, Fisher LD, et al. A prospective study of the role of cardiovascular risk factors and fitness in industrial back pain complaints. Spine. 1989;14:141–147
  45. Picavet HS, Schuit AJ. Physical inactivity: a risk factor for low back pain in the general population?. J Epidemiol Community Health. 2003;57:517–518
  46. Westerterp KR. Assessment of physical activity level in relation to obesity: current evidence and research issues. Med Sci Sports Exerc. 1999;31:S522–S525
  47. Philippaerts RM, Westerterp KR, Lefevre J. Comparison of two questionnaires with a tri-axial accelerometer to assess physical activity patterns. Int J Sports Med. 2001;22:34–39
  • a Tunturi E3; Accell Group NV, PO Box 435, 8440 AK, Heerenveen, The Netherlands.
  • b Jaeger ER 800; Ergoline GmbH, Lindenstrasse 5, 72475 Bitz, Germany.
  • c Lode Concord 1.0; Lode BV, Groningen, Zernikepark 16, 9747 AN, Groningen, The Netherlands.
  • d Polar Favor; Polar Electro Oy, Professorintie 5, FIN-90440 Kempele, Finland.
  • e Servier; Servier Nederland Pharma BV, Einsteinweg 5-B, 2333 CC Leiden, The Netherlands.
  • f SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

 Supported by Zorgonderzoek Nederland/Medische Wetenschappen (grant no. 014-32-007).

 No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit on the authors or on any organization with which the authors are associated.

PII: S0003-9993(08)01506-2

doi:10.1016/j.apmr.2008.07.009

Archives of Physical Medicine and Rehabilitation
Volume 90, Issue 1 , Pages 109-117, January 2009

Article Tools

* Image Usage & Resolution