Procedure and measurements
The baseline questionnaires were distributed by post together with an accompanying letter explaining the aim and procedure of the project. It was requested that the questionnaires be sent back with the enclosed self-addressed prepaid envelopes. Reminder letters were sent after a fortnight. For the 12-month follow-up we used the same procedure as for the baseline assessment, although no reminder letters were sent. The questionnaires contained demographic questions about: sex, age, BMI, smoking habits (“No, I have never smoked”, “Yes, but I have quit”, “Yes, occasionally”, and “Yes, daily”), housing (own home or special housing) and living arrangements (alone or with someone) and marital status.
Pain-related questions were also included in the questionnaires. Chronic pain, defined as pain that had lasted longer than three months [
24] and was noted by responding YES or NO to the question “Have you been troubled by pain for the last three months or more?” Pain characteristic (intensity, duration, location) was measured using single items extracted from the brief Screening version of the Multidimensional Pain Inventory (Swedish version) [
25]. Pain intensity was measured using the item “Rate the average level of your pain during the last week” with responses on a 6-point Likert scale ranging from
No pain at all (1 point) to
Tremendous amount of pain (6 points). Duration of pain was measured in years. Primary pain localization was also identified. The respondents could choose between 7 alternatives; upper extremities, shoulder and neck, lower extremities, thorax and abdomen, back and pelvis, head and other locations (including hand and feet). The brief Swedish version of MPI has been psychometrically tested and shown to have acceptable validity and reliability among the elderly, aged 60–89 years old [
26].
Kinesiophobia (excessive fear of movement/(re) injury) related to pain was measured using an 11-item version of the
Tampa Scale of Kinesiophobia (TSK-11) [
27]. The 11 items each had four response options, all anchored with the answers
Strongly disagree, which scored one point, and
Strongly agree, which scored four points. A total summary score was then calculated and could range between 11 and 44 points. A high score indicated strong fear of movement/(re) injury, i.e. high kinesiophobia. The TSK-11 scale has been psychometrically evaluated and shown good construct validity and reliability in older adults (Cronbach’s alpha for internal consistency = 0.74–0.87) and good test-retest reliability (r = 0.747 for intraclass correlation (ICC)) [
28].
Self-efficacy was measured using the
General Self-Efficacy scale (GSE), a generic instrument that aims to measure “optimistic self-beliefs to cope with a variety of difficult demands in life”, recommended for use among adults with chronic pain [
29]. The scale consists of 10 items with alternative responses: 1 = “not at all true”, 2 = “hardly true”, 3 = “moderately true”, and 4 = “exactly true”. A sum score, ranging from 10 to 40, is then calculated. A high score indicates high self-efficacy. The scale has commonly been used in older adults as well as in pain patients [
30] and has been translated into Swedish [
31]. The GSE has been tested for its psychometric properties and has demonstrated good validity and reliability (Cronbach’s alpha for internal consistency = 0.75–0.91) and good test-retest reliability (r = 0.55–0.67) [
32].
Self-rated health was measured using an item extracted from the 12-item Short-Form Health Survey (SF-12) [
33]. The SF-12 measures health-related quality of life. The item used in the present study was: “How would you generally like to say that your health is?” The item had five alternative responses: (“Excellent health”, “Very good health”, “Good health”, “Fair health” and “Poor health”). SF-12 has been found to be valid and reliable in Swedish older adults [
34].
Physical activity recommendations from international health guidelines state that older adults should be moderately physically active (i.e. a moderate amount of effort that noticeably accelerates the heart rate) at least five days a week for a minimum of 30 min a day [
35], but a study from 2011 suggested that only 30 % in the group aged 60+ years seemed to reach these recommended levels of physical activity [
36]. Levels of physical activity were measured with
Grimby’s Activity Scale; a scale developed to evaluate self-rated physical activity in older adults [
37]. Levels of physical activity were classified using the question: “How physically active do you think you have been during the last six months?” Physical activity was classified by one of the following responses:
1.
Hardly any physical activity.
2.
Mostly sitting, sometimes a walk, light gardening or similar tasks, sometimes light household activities such as heating up food, dusting or clearing away.
3.
Light physical exercise around 2–4 h a week, such as walks, fishing, dancing, ordinary gardening etc. including walks to and from shops. Main responsibility for light domestic work such as; cooking, dusting, clearing away and making beds. Performs or takes part in weekly cleaning.
4.
Moderate exercise 1–2 h a week, such as jogging, swimming, gymnastics, heavy gardening, home-repairs or light physical activities more than four hours a week. Responsible for all domestic activities, light as well as heavy. Weekly cleaning such as doing vacuum cleaning, washing floors and window cleaning.
5.
Moderate exercise at least three hours a week such as tennis, swimming, jogging etc.
6.
Hard or very hard exercise regularly and several times a week where the physical exertion is great, such as jogging or skiing.
Response options 4–6 have previously been used to correspond to WHO’s recommended levels of physical activity [
22]. The physical activity scale has been psychometrically evaluated in older adults and has demonstrated acceptable construct validity when validated against measures of physical performance [
38,
39]. It also demonstrated acceptable construct validity when validated against various physical measures and has been found to be able to discriminate between groups, who were more active and less active, as assessed by measuring maximal oxygen uptake [
38].
Statistical analyses
Descriptive statistics are presented as the mean, standard deviation, and range, and percentiles were calculated. The Chi-squared test was used to compare categorical data; the Mann–Whitney U-test was used for ordinal data and the Student’s t-test for interval/ratio data. McNemar’s test, the Wilcoxon signed rank test, and the paired sample t-test were also used for paired group comparisons. To study the associations between the variables at baseline in those reporting chronic pain, two binary logistic regression analyses were done (by the backward stepwise likelihood ratio method) with physical activity, dichotomized as inactive (1–3) and active (4–6) [
22], as the dependent variable. One logistic regression analysis was done to find associations at baseline and one logistic regression analyses was performed in order to find possible predictors of physical activity. Psychosocial and demographic variables were included based on empiric relations to physical activity found in previous studies [
15‐
17,
20]. Demographic variables (housing and living arrangements), pain-related variables (intensity, duration), psychological variables (kinesiophobia and self-efficacy), and health-related variables (BMI, self-rated health) were entered as independent variables. To test the quality of the logistic models, the Hosmer–Lemeshow goodness-of-fit test and Nagelkerke’s R-squared were used. A value of
p > 0.05 indicates a good model fit in the Hosmer–Lemeshow goodness-of-fit test [
40]. Analyses were done using SPSS Statistics, version 18.0 (SPSS Inc., Chicago, IL).