Study design: A prospective study of psychological risk factors for first-time low back pain with repeated use of psychological questionnaires.
Objectives: To measure the reproducibility of scores from psychological questionnaires, and to compare this with changes that follow an individual's first attack of back pain. Secondly, to determine which scores predict first-time back pain.
Summary of background data: "Abnormal" psychometric scores are associated with several aspects of back pain behavior. Little is known, however, about their reproducibility or long-term stability, and there has been no definitive answer to the question: which comes first, "abnormal" scores or low back pain?
Methods: 403 volunteers with no history of "serious" low back pain (defined as pain requiring medical attention or absence from work) participated in a functional spinal assessment. At the time of initial assessment and at 6-month intervals thereafter, the volunteers completed the following questionnaires: the Health Locus of Control, which was subdivided into three sections labelled "Internal," "Powerful others," and "Chance"; the Modified Somatic Perception Questionnaire; and the Zung depression scale. Scores from the Modified Somatic Perception Questionnaire and from the Zung depression scale were added to form a measure of psychological distress. Additional questionnaires inquired about any back pain experienced in the previous 6 months. Only three volunteers had left the study at the 18-month follow-up. At that time 162 participants had reported "any" low back pain, of which 79 were "serious."
Results: Intraclass correlation coefficients for scores repeated after 6 months ranged from 0.67-0.80, and reproducibility of scores was equally high between the 0-, 6-, 12- and 18-month assessments. None of the scores were affected by "any" low back pain, and only the Modified Somatic Perception Questionnaire scores changed after "serious" back pain was reported. In a multivariate analysis, the most significant predictor of first time "serious" or "any" back pain was a history of non-"serious" back pain (P < 0.001). Of the psychological factors, the sum of Modified Somatic Perception Questionnaire scores and Zung questionnaire scores was the best predictor of "serious" back pain (P = 0.037), and the Modified Somatic Perception Questionnaire score was the best predictor of "any" back pain (P = 0.002). The 25% of participants with the highest sum of scores from the Modified Somatic Perception Questionnaire and Zung questionnaire was 2.7 times more likely to develop "serious" back pain than the 25% with the lowest sum of these scores. Nevertheless, after accounting for the affects of a history of non-"serious" back pain, psychometric scores predicted less than an additional 3% of reported back pain.
Conclusions: The scores from the Modified Somatic Perception Questionnaire and Zung questionnaire were reproducible over 18 months and were affected little by first episodes of back pain; yet these scores were significant predictors of it. "Abnormal" scores from these questionnaires precede back pain in a small number of people.