Pessimism and a lack of exercise is “a risky combination”


Anne Mannion
Anne Mannion

The biopsychosocial model has taught us that psychological factors are important in governing not only the susceptibility to back pain but also its severity and future course, write Anne Mannion and Achim Elfering.1 Psychological risk factors for the development of chronic back pain include maladaptive beliefs about the inevitability of the future as a consequence of having back trouble.2 These have been shown to be associated with pain maintenance,3 back pain-related absenteeism and presenteeism,4 and back pain-related use of the healthcare system.5 However, most of the studies have been cross-sectional in nature, making it difficult to distinguish between cause and effect (which came first, the chicken or the egg?).


We therefore decided to evaluate the longitudinal validity of beliefs, measured using the Back Beliefs Questionnaire (BBQ),6 in predicting future pain in the shoulders, neck and back. Since pessimistic back beliefs impede recovery from back pain, we hypothesised that the prognostic value of BBQ scores in predicting future pain would be greatest in those with back pain at baseline. Pessimistic back beliefs are also associated with (ineffective) passive pain management, especially when the individual’s habitual level of physical activity is already low. Thus, we hypothesised that the association between pessimistic back beliefs and future pain would be greatest in those who did not exercise at baseline. Finally, we anticipated that, in those reporting back pain at baseline, exercise would buffer the association between pessimistic beliefs and future pain, compared with those who reported back pain but were inactive.

We carried out a survey of volunteers drawn from a large group of just over 16,000 individuals who had previously participated in a population-based survey of musculoskeletal pain in Switzerland. First, we stratified them, based on whether they had reported low back pain “in the last month” in the previous survey, and then randomly selected 2,860 of these (roughly half of each, with and without back pain) for the current study. Of these, 2,507 returned a baseline questionnaire and 1,883 a one-year follow-up questionnaire.

BBQ comprises nine statements such as “back trouble will eventually stop you from working…must be rested…means you end up in a wheelchair”. Pain was assessed with four questions: “In the last four weeks, how much pain did you experience in the following body regions?” The regions included: left shoulder, right shoulder, neck and upper back, and lower back. Response options ranged from 0 (no pain) to 6 (unbearable pain).

We also asked the participants about any back pain at the time of filling in the survey (yes/no), and whether they were exercisers (yes/no) regarding participation in sport or carrying out exercise (eg. cycling to work, hiking, yoga, etc.), to allow us to analyse these as separate sub-groups.

Our hypotheses were tested using cross-lagged panel models (a type of structural equation modelling).7 This sophisticated statistical method allows us to identify both the source and direction of a hypothesised causal effect. In other words it allows us to evaluate the strength of the longitudinal directional path (or “prospective risk path”) between BBQ scores at baseline and pain at follow-up, whilst simultaneously taking into consideration (or “adjusting for”) any cross-sectional intercorrelations between BBQ and pain, at either baseline or at follow-up, and of the relationship between each variable at baseline with itself at follow-up.

For the whole group, BBQ at baseline was a statistically significant predictor of pain at follow-up, though the size of the effect (coefficient of correlation) was very small. The reverse pathway was not significant, i.e. pain at baseline did not determine beliefs at follow-up.


When the group was split into those with back pain at baseline and those without, the prospective risk path was significant only in the back pain group.

Similarly, if we split the group into those that exercised and those that did not, the prospective risk path was significant only in the group doing no exercise.

And finally, we found that if individuals had low back pain but exercised, then this diminished the risk of beliefs influencing future pain (so exercise was “protective”; bottom left in Figure 1), whereas having low back pain and doing no exercise resulted in the strongest coefficient of all. This therefore represents a high-risk group.

Overall, we were able to conclude that our findings confirmed the longitudinal validity of the BBQ. The prospective risk path over a one-year follow up was statistically significant, and there was a moderate effect size in the highest-risk group. We believe that cognitive behavioural interventions, education and exercise programmes might serve to reduce the impact of back pain in high-risk groups such as those with maladaptive back beliefs who are not regular exercisers.



Anne Mannion, Schulthess Klinik, Zürich, Switzerland, and Achim Elfering, University of Bern, Switzerland



  1. Rolli Salathé C, Kälin W, Semmer NK, et al. European Pain Journal 2013: 17; 1411–1421.
  2. Kendall NAS, Linton SJ, Main CJ. Accident Rehabilitation and Compensation Insurance Corporation of New Zealand and the National Health Committee, Wellington 2004.
  3. Elfering A, Mannion AF, Jacobshagen N, Tamcan O, Müller U. Scandinavian Journal of Work, Environment and Health 2009: 35; 437-445.
  4. Mannion AF, Horisberger B, Eisenring C, et al. Journal of Occupational and Environmental Medicine 2009: 51; 1256–1266.
  5. Mannion AF, Wieser S, Elfering A. Spine 2013: 38;1016–1025.
  6. Symonds TL, Burton AK, Tillotson KM, Main CJ. Occupational Medicine 1996: 46; 25–32.
  7. Clegg CW, Jackson PR, Wall TD. Journal of Occupational Psychology 1977: 50; 177–196.