At baseline, 650 questionnaires were mailed and 502 (77.2%) people responded. Participants had a mean age of 65.2 years (SD 10.2) and 306 (61.2%) were female. Details of the baseline characteristics of the study group are summarized in Table . At baseline, the mean pain score was 65.69 (SD 19.2) and the mean disability score was 50.53 (SD 28.9). In comparing the group who had below median pain and those who had above median pain, the group with the higher pain level was made up of a higher proportion of females, and a higher proportion of people with anxiety, depression and worse general health. Those who were in the above median disability groups tended to be more likely to have depression and anxiety, reported a worse general health and catastrophized more (data not shown). Of the 502 people who responded at baseline, 436 participants consented to follow-up. Therefore, at six months, 436 questionnaires were mailed and 370 (73.7% of the original cohort of 502) participants responded. There were no significant differences between the baseline characteristics of those who did and did not respond at follow-up.
Demographics of participants split for above and below baseline median pain
Full details of unadjusted and adjusted regression coefficients at baseline are shown in Table . At baseline catastrophizing was shown to be associated with more pain (regression coefficient (β) (increase in pain score for 1 point increase in catastrophising) 5.93, 95% CI 5.13, 6.72). Once this was adjusted for age, gender, anxiety and depression, the association between catastrophizing and pain was attenuated, but still significant (β =5.15 95% CI 4.25, 6.05). Catastrophizing was also positively associated with disability both in the unadjusted (β
8.44, 95% CI 7.23, 9.65) and adjusted models (β
6.92, 95% CI 5.54, 8.31). Ignoring sensations was association with a decrease in baseline disability (adjusted β –1.72, 95% CI −3.36, –0.08), but not with pain. Self-statements and increasing behaviours were not statistically associated with baseline pain or disability.
Details of the unadjusted and adjusted regression coefficients at six months for pain and disability are shown in Table . Catastrophizing at baseline was shown to be associated with more pain and disability at 6-month follow-up, but these associations were attenuated after adjustment for age, gender, baseline pain, disability, anxiety and depression. After adjustment for potential confounders, there was a positive and statistically significant association between ignoring sensation and both pain and disability at follow-up. Using self-statements and increasing behaviour were not statistically significantly associated with pain or disability at follow-up.
Table present the associations between baseline coping strategies and pain and disability at 6-months follow-up, stratified by baseline pain or disability as appropriate. Catastrophizing, whether having below or above average pain or disability at baseline, was associated with more pain or disability at follow-up. However, once adjustments were made for confounders the association was no longer significant. In the higher baseline pain group, ignoring sensations was positively associated with pain at follow-up both in the crude (β =2.50, 95% CI 0.14, 4.86) and adjusted models (β =3.36, 95% CI 0.92, 5.80). This was not true for ignoring sensations in the lower pain group. On stratifying the sample into above and below median baseline disability, ignoring sensations, using self-statements and increasing behaviour were statistically non-significant.