Differences in physical activity patterns in patients with spondylarthritis

Authors

Errata

This article is corrected by:

  1. Errata: Error in Table of Article by Haglund et al (Arthritis Care Res, December 2012) Volume 65, Issue 3, 492, Article first published online: 26 February 2013

Abstract

Objective

To study patient-reported physical activity in patients with spondylarthritis (SpA) and possible differences in physical activity patterns between the SpA subtypes and sexes.

Methods

In 2009, a questionnaire including inquiries concerning physical activity was sent to patients with a SpA diagnosis (n = 3,711). The World Health Organization (WHO) global recommendations of physical activity for health requiring 150 minutes of moderate-intensity physical activity (MI-PA) or 75 minutes of vigorous-intensity physical activity (VI-PA) per week were used as recommended levels. Standardized risk ratios (RRs) were calculated by using physical activity data from the Swedish population. The association within the SpA group between sex, age, disease-related variables, anxiety, and depression and meeting recommended levels of MI-PA and VI-PA (dependent variables) was studied with multivariate analysis.

Results

A total of 2,167 patients with SpA (48% men, mean ± SD age 55 ± 14 years) responded to the questionnaire. Sixty-eight percent of the patients met the WHO recommendations, more frequently in women than in men (70% versus 66%). The recommendations were more often met in the SpA group (RR 1.09, 95% confidence interval [95% CI] 1.04–1.15) compared with the Swedish population. There was a tendency for young women to meet the WHO recommendations less often than the Swedish population (RR 0.94, 95% CI 0.63–1.25). Different factors were found to influence whether the patients were exercising with a moderate or vigorous intensity.

Conclusion

Seven of 10 patients with SpA met the WHO recommendations of physical activity for health, but we found sex and disease subtype differences. This information can be useful in clinical practice when coaching patients to have a healthier lifestyle.

INTRODUCTION

In recent years, awareness of the value of physical activity has been raised and physiotherapy interventions focusing on physical activity and exercise regimens have become of great importance. The World Health Organization (WHO) has authored Global Recommendations of Physical Activity for Health, which has been adapted for people with inflammatory diseases (1). In the general population, regular physical activity reduces the risk of cardiovascular comorbidities, but may also prevent a number of other comorbidity disorders such as diabetes mellitus, obesity, hypertension, and osteoporosis (2–5). It is likely that compliance with the WHO recommendations of 150 minutes of moderate-intensity physical activity (MI-PA) or 75 minutes of vigorous-intensity physical activity (VI-PA) per week will promote and maintain health, and for additional health benefits adults should increase the length or the intensity of aerobic physical activity. However, many individuals with arthritis, as well as people in general, do not meet this recommended level of physical activity (6–8).

It is also well known that patients with rheumatoid arthritis (RA) and patients with spondylarthritis (SpA) have an increased risk for cardiovascular comorbidities (9, 10), which enhances the importance of a healthy lifestyle. The large body of research has been performed on patient-reported physical activity and in patients with RA (11), and we have little information concerning physical activity patterns in the SpA group, which is of great importance in coaching and when planning and providing a specific physiotherapy regimen. SpA is an inflammatory disease, which includes the following subtypes: ankylosing spondylitis (AS), psoriatic arthritis (PsA), inflammatory bowel disease (IBD)–related arthritis, and undifferentiated SpA (uSpA) with several common themes. The prevalence is reported to be between 0.3–0.45% of the adult population (12–16). SpA often causes limitations of functional capacity, impairs the ability to work, and adds a negative impact on the overall quality of life (17–20). The most successful and recommended treatment has been a combination of pharmacologic therapy and physiotherapy, including exercises for flexibility and physical fitness (21, 22).

Our aim was to study patient-reported physical activity in patients with SpA in comparison to the general population and also to study possible differences between the sexes and SpA subtypes. A second aim was to study factors associated with the different physical activity patterns (MI-PA versus VI-PA).

Significance & Innovations

  • Seven out of 10 patients with spondylarthritis (SpA) met the World Health Organization (WHO) recommendations of physical activity for health, which was slightly higher in comparison to the Swedish population.

  • Special attention may be needed for women in the age group 18–29 years who met the WHO recommendations of physical activity for health to the lowest degree.

  • Women preferred to be physically active at a moderate intensity, while men preferred to exercise at a more vigorous intensity.

  • This information is important and can help health professionals in their work when coaching patients with SpA to achieve a healthier lifestyle.

PATIENTS AND METHODS

Study population.

The present study was based on a cross-sectional postal questionnaire survey performed in the southernmost county of Sweden between May and August 2009. In this region, all health care visits (both public and private, as well as inpatient and outpatient) are registered in the Skane Health Care Register (SHCR) and linked to a unique personal identification number. The SHCR includes information on the health care provider, date of visit, and the International Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) diagnoses as given by physicians in primary or secondary care. More details on this register are published elsewhere (13, 23). Using the SHCR, all subjects who at any time during 5 calendar years (2003–2007) had received a diagnosis of SpA were identified by their ICD-10 codes and mailed a questionnaire. Validation of the SpA diagnoses in SHCR via review of medical records indicated a high level of accuracy of the clinical diagnosis (13). All cases were age ≥18 years, and the diagnosis was required to be registered at least once by a rheumatologist or an internist or at least on 2 separate occasions by any other physician in primary or secondary care. A total of 3,711 patients with SpA were identified and included in the questionnaire survey.

The questionnaire.

The questionnaire in the survey consisted of several validated patient-reported instruments. In order to improve face and content validity, the composite questionnaire was tested in 3 focus groups, including a total of 20 patients with different SpA subtypes. This resulted only in minor corrections to improve understanding.

Details of physical activity and exercise.

Information on physical activity was collected from 3 questions in the survey. The questions were copied from recommendations on how to measure self-reported physical activity at that time (2009). This included information on intensity (moderate or vigorous)/intensive physical activity, duration (30 minutes or more), and number of days (1, 24). Patients who reported activities 5–7 days/week at moderate intensity (defined as aerobic exercise that causes mild breathlessness, e.g., walking or gardening, etc.) for at least 30 minutes a day were considered to meet the recommendations for moderate activity (MI-PArec). Those who exercised on a vigorous intensity for at least 30 minutes 2–3 times/week or more were also considered to meet the vigorous recommendations (VI-PArec). The 2 recommendations (MI-PArec and VI-PArec) were then merged and categorized into either meeting or not meeting the WHO recommendations (WHOrec) (1, 24, 25). Categorizing data this way was performed earlier in large studies in Sweden in both in the arthritis and the general population (7, 26).

Other assessments.

Patient characteristics such as age, sex, and ICD-10 diagnosis were collected from the SHCR while disease duration, smoking habits, and educational level were self-reported information. The Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) (27), consisting of 6 questions on pain and stiffness, and the Bath Ankylosing Spondylitis Functional Index (BASFI) (28), including 10 questions on physical function, were collected together with pain and fatigue as measured on numerical rating scales. In the BAS indices, as well as in the fatigue and pain scales, the total score ranged between 0–10, with 0 being best and 10 being worst. Additionally, information from the EuroQol-5 domain (EQ-5D) instrument, a generic measure assessing health-related quality of life in 5 dimensions, including mobility, self-care, usual activity, pain/discomfort, and anxiety/depression, was collected. The EQ-5D has a summary score ranging from 0 (no health) to 1 (full health) (29). Psychological status was measured with the Hospital Anxiety and Depression Scale (HADS), scoring anxiety and depression. The 2 subscales each consist of 7 items scored from 0 (no distress) to 21 (maximum distress) (30).

Data concerning individuals meeting the WHO recommendations in the Swedish population were downloaded from the National Public Health Institute (www.fhi.se). These data were gathered through a national public health survey in 2010, where 20,000 randomly selected individuals in Sweden, ages 16–84 years, received the questionnaire and answered questions about their physical activity pattern. The results were based on responses from 9,972 individuals (7).

Statistical analysis.

Standardized risk ratios (RRs) of meeting the WHOrec were calculated by dividing the observed ratios of the number of patients meeting the WHOrec in the SpA cohort by their expected ratios based on the number of subjects who met the WHOrec in the Swedish population (7). The total RRs were adjusted for age and sex. An RR >1 equals a higher proportion meeting WHOrec in the SpA cohort than in the reference population, while an RR <1 equals a lower proportion meeting WHOrec compared to the reference population. We calculated 95% confidence intervals (95% CIs) for the RRs.

The Mann-Whitney U test or the chi-square test was used to analyze differences between groups. Multivariate logistic regression analysis was used to study the associations with MI-PArec or VI-PArec as the dependent variables, with sex, SpA subtypes, age, smoking habits, education level, anxiety, depression, self-reported physical function, self-reported disease activity, and health-related quality of life as independent variables. The BASFI had a correlation of rs = >0.3 with BASDAI and EQ-5D, which is why each of these variables was included separately in the multiple logistic regression analysis, together with the other independent variables when analyzing MI-PArec and VI-PArec. All analyses were performed using SPSS software, version 17, for Windows. The Regional Ethical Review Board at Lund University, Sweden, approved the study. Informed consent was obtained from all patients.

RESULTS

Out of 3,711 patients with SpA who received the questionnaire, 2,851 (76%) responded to the study; of these, 684 patients (18%) declined participation and 2,167 (58%, 1,045 men [mean ± SD age 56 ± 13 years] and 1,122 women [mean ± SD age 55 ± 14 years]) returned the questionnaire and were thus included in the analysis.

Analysis of nonresponders showed that patients with AS were most likely to respond to the questionnaire and that patients with IBD-related arthritis were the least likely, while sex did not play an important role. The analysis of nonresponders versus responders also showed that higher age significantly predicted a higher response in men. Women's tendency to respond increased slightly with age as well, except for the AS subtype, where the likelihood declined with age.

Patient characteristics.

Disease activity (mean ± SD BASDAI 4.4 ± 2.2) and function (mean ± SD BASFI 3.2 ± 2.5) was impaired in patients with SpA. Women reported worse disease activity (BASDAI), function (BASFI), health-related quality of life (EQ-5D), depression (HADS), fatigue, and pain scores compared to men (Table 1).

Table 1. Clinical characteristics of the patients in each subtype*
SpA subtypeAllWomenMenP
  • *

    Values are the mean ± SD unless indicated otherwise. SpA = spondylarthritis; AS = ankylosing spondylitis; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; BASFI = Bath Ankylosing Spondylitis Functional Index; EQ-5D = EuroQol 5-domain; HADS = Hospital Anxiety and Depression Scale; PsA = psoriatic arthritis; IBD-related arthritis = inflammatory arthritis associated with inflammatory bowel disease.

AS, no.571197374 
 Age, years55 ± 1452 ± 1456 ± 140.01
 Disease duration, years20 ± 1417 ± 1222 ± 14< 0.001
 BASDAI (range 0–10)3.9 ± 2.24.2 ± 2.33.7 ± 2.20.03
 BASFI (range 0–10)3.3 ± 2.63.5 ± 2.53.3 ± 2.60.29
 Fatigue (range 0–10)4.4 ± 2.75.0 ± 2.84.1 ± 2.6< 0.001
 Pain (range 0–10)3.7 ± 2.54.2 ± 2.63.4 ± 2.40.01
 EQ-5D (range 0–1)0.75 ± 0.170.74 ± 0.160.75 ± 0.170.22
 HADS anxiety (range 0–21)9.6 ± 1.99.4 ± 1.99.6 ± 1.90.56
 HADS depression (range 0–21)9.0 ± 1.88.9 ± 1.69.0 ± 1.90.55
PsA, no.1,185681504 
 Age, years58 ± 1357 ± 1458 ± 130.50
 Disease duration, years13 ± 1112 ± 1113 ± 110.36
 BASDAI (range 0–10)4.8 ± 2.15.1 ± 2.04.3 ± 2.2< 0.001
 BASFI (range 0–10)3.2 ± 2.63.8 ± 2.72.5 ± 2.3< 0.001
 Fatigue (range 0–10)4.8 ± 2.75.3 ± 2.64.1 ± 2.7< 0.001
 Pain (range 0–10)4.2 ± 2.54.6 ± 2.43.6 ± 2.5< 0.001
 EQ-5D (range 0–1)0.74 ± 0.170.72 ± 0.170.77 ± 0.16< 0.001
 HADS anxiety (range 0–21)9.3 ± 2.09.3 ± 2.09.2 ± 1.90.16
 HADS depression (range 0–21)9.0 ± 1.79.1 ± 1.78.8 ± 1.80.04
Undifferentiated SpA, no.370216154 
 Age, years49 ± 1450 ± 1449 ± 130.86
 Disease duration, years10 ± 1010 ± 911 ± 100.45
 BASDAI (range 0–10)4.2 ± 2.24.6 ± 2.13.8 ± 2.20.001
 BASFI (range 0–10)2.9 ± 2.43.3 ± 2.42.4 ± 2.3< 0.001
 Fatigue (range 0–10)4.7 ± 2.85.0 ± 2.94.2 ± 2.60.01
 Pain (range 0–10)4.0 ± 2.54.2 ± 2.53.6 ± 2.40.03
 EQ-5D (range 0–1)0.76 ± 0.170.74 ± 0.160.78 ± 0.170.01
 HADS anxiety (range 0–21)9.2 ± 1.89.2 ± 1.99.1 ± 1.70.82
 HADS depression (range 0–21)9.1 ± 1.69.2 ± 1.69.0 ± 1.70.09
IBD-related arthritis, no.412813 
 Age, years57 ± 1256 ± 1258 ± 100.52
 Disease duration, years11 ± 912 ± 97 ± 60.13
 BASDAI (range 0–10)5.1 ± 1.95.2 ± 2.04.6 ± 1.10.48
 BASFI (range 0–10)3.3 ± 2.74.1 ± 2.81.6 ± 1.40.01
 Fatigue (range 0–10)5.1 ± 2.55.9 ± 2.33.5 ± 2.10.01
 Pain (range 0–10)4.4 ± 2.54.9 ± 2.43.2 ± 2.40.06
 EQ-5D (range 0–1)0.74 ± 0.150.72 ± 0.160.80 ± 0.120.04
 HADS anxiety (range 0–21)9.0 ± 2.08.5 ± 1.810.1 ± 2.20.03
 HADS depression (range 0–21)8.8 ± 1.69.0 ± 1.88.4 ± 1.20.37

Standardized RRs of meeting the WHOrec.

Within the SpA group, a total of 2,126 patients had answered the question concerning MI-PA or VI-PA and were included in this analysis. Compared to the Swedish population, the patients with SpA overall met the WHOrec to a higher degree (RR 1.09, 95% CI 1.04–1.15). The age-adjusted RR for meeting the WHOrec in women with SpA was RR 1.14 (95% CI 1.06–1.22) and in men was RR 1.05 (95% CI 0.97–1.12). Women in the youngest age group (18–29 years) met the WHOrec to the lowest degree (RR 0.94, 95% CI 0.63–1.25), while the oldest age group (65–85 years) met the recommendation more often (RR 1.26, 95% CI 1.08–1.44). The lowest RR was found in young women with PsA (RR 0.60, 95% CI 0.26–0.95). For men, the point estimates were above 1.0 for all age groups, except for those ages 30–44 years, which were just below 1.0 (RR 0.98, 95% CI 0.82–1.14) (Figure 1A). The age- and sex-adjusted RR for all SpA subtypes, including AS, PsA, uSpA, and IBD-related arthritis, showed higher rates as compared to the Swedish population (RR 1.10, 95% CI 0.99–0.21; RR 1.08, 95% CI 1.01–1.16; RR 1.09, 95% CI 0.95–1.22; and RR 1.32, 95% CI 0.87–1.77, respectively) (Figure 1B).

Figure 1.

Standardized risk ratios and 95% confidence intervals for meeting the World Health Organization recommendations of physical activity for health in women and men with spondylarthritis (SpA) in different age groups (A) and in the different subtypes of SpA (B). AS = ankylosing spondylitis; PsA = psoriatic arthritis; USpA = undifferentiated SpA; Aa-IBD = inflammatory arthritis associated with inflammatory bowel disease.

Meeting the WHO recommendations of physical activity for health.

A total of 68% (n = 1,470) of all patients with SpA met the WHOrec, more frequently in women (70% versus 66%; P = 0.007). When splitting the WHOrec into MI-PArec and VI-PArec, a total of 57% (n = 1,237) in the SpA group were exercising at a moderate intensity (≥30 minutes ≥5 days/week), which was also more frequent in women (61% versus 53%; P < 0.001). Furthermore, 32% (n = 697) of the patients with SpA met the VI-PArec (exercising at a more vigorous intensity for ≥30 minutes 2–3 times/week), more frequently in men (35% versus 29%; P = 0.01).

Multivariate analysis of exercise levels.

In the multivariate analyses, female sex and older age were associated with meeting the MI-PArec with odds ratios (ORs) ranging from 1.63–1.78 (sex), and OR 1.02 (age per year) (P < 0.001). Also, a better health-related quality of life (EQ-5D) was positively associated (OR 2.76, 95% CI 1.54–4.95) with the outcome, while patients with higher BASFI scores (worse function) were less likely to meet the MI-PArec (OR 0.94, 95% CI 0.90–0.98). SpA subtype was not statistically significantly associated with meeting the MI-PArec (Table 2).

Table 2. Results of multiple logistic regression analysis with meeting the MI-PArec (physical activity performed at moderate intensity ≥30 minutes ≥5 days/week) as the dependent variable*
VariablesBASFI (n = 1,980)BASDAI (n = 1,455)EQ-5D (n = 1,938)
OR (95% CI)POR (95% CI)POR (95% CI)P
  • *

    MI-PArec = recommendations for moderate-intensity physical activity; BASFI = Bath Ankylosing Spondylitis Functional Index; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; EQ-5D = EuroQol 5-domain measure; OR = odds ratio; 95% CI = 95% confidence interval; HADS = Hospital Anxiety and Depression Scale; AS = ankylosing spondylitis; PsA = psoriatic arthritis; uSpA = undifferentiated spondylarthritis; IBD = inflammatory bowel disease.

  • The variables were included separately in the multiple logistic regression analysis together with the other independent variables, according to high correlation.

  • Per unit increase.

Men1 1 1 
Women1.66 (1.37–2.01)< 0.0011.78 (1.42–2.23)< 0.0011.63 (1.35–1.98)< 0.001
Age (per 1 year)1.02 (1.01–1.03)< 0.0011.02 (1.01–1.03)< 0.0011.02 (1.01–1.03)< 0.001
Nonsmokers1 1 1 
Previous smokers/smokers0.95 (0.78–1.15)0.600.87 (0.69–1.09)0.210.92 (0.76–1.12)0.42
Low education level (≤12 years)1 1 1 
High education level (≥12 years)1.01 (0.83–1.23)0.900.99 (0.78–1.23)0.891.06 (0.87–1.29)0.59
HADS anxiety (range 0–21)0.96 (0.91–1.01)0.080.95 (0.90–1.00)0.060.97 (0.92–1.02)0.26
HADS depression (range 0–21)1.00 (0.95–1.06)0.991.00 (0.94–1.06)0.961.01 (0.96–1.06)0.77
Diagnosis subtype      
 AS1 1 1 
 PsA0.98 (0.79–1.23)0.871.05 (0.81–1.35)0.740.98 (0.78–1.22)0.84
 uSpA0.98 (0.74–1.30)0.900.92 (0.68–1.24)0.580.91 (0.68–1.22)0.53
 IBD-related arthritis1.10 (0.55–2.18)0.801.21 (0.54–2.70)0.641.27 (0.63–2.55)0.51
BASFI (range 0–10)0.94 (0.90–0.98)0.002    
BASDAI (range 0–10)  0.96 (0.91–1.01)0.14  
EQ-5D (range 0–10)    2.76 (1.54–4.95)0.001

Repeating the same multivariate analyses using meeting the VI-PArec as a dependent variable, we found negative associations with older age (OR 0.99, 95% CI 0.98–1.00), the diagnosis subtype PsA (OR 0.72, 95% CI 0.56–0.92), worse function (BASFI; OR 0.87, 95% CI 0.83–0.91), higher self-reported disease activity (BASDAI; OR 0.92, 95% CI 0.87–0.98), and smokers/previous smokers (OR range 0.68–0.73). Better health-related quality of life (OR 6.42, 95% CI 3.12–13.2) was positively associated with the outcome (Table 3).

Table 3. Results of multiple logistic regression analysis with meeting the VI-PArec (exercising 2 or 3 times/week at a vigorous level >30 minutes) as the dependent variable*
VariablesBASFI (n = 1,662)BASDAI (n = 1,241)EQ-5D (n = 1,627)
OR (95% CI)POR (95% CI)POR (95% CI)P
  • *

    VI-PArec= recommendations for vigorous-intensity physical activity; BASFI = Bath Ankylosing Spondylitis Functional Index; BASDAI = Bath Ankylosing Spondylitis Disease Activity Index; EQ-5D = EuroQol 5-domain; OR = odds ratio; 95% CI = 95% confidence interval; HADS = Hospital Anxiety and Depression Scale; AS = ankylosing spondylitis; PsA = psoriatic arthritis; uSpA = undifferentiated spondylarthritis; IBD = inflammatory bowel disease.

  • The variables were included separately in the multiple logistic regression analysis together with the other independent variables, according to high correlation.

  • Per unit increase.

Men1 1 1 
Women0.86 (0.70–1.07)0.170.83 (0.65–1.05)0.120.78 (0.63–0.96)0.02
Age (per 1 year)0.99 (0.98–1.00)0.060.99 (0.98–1.00)0.010.99 (0.98–0.996)0.004
Nonsmokers1 1 1 
Previous smokers/smokers0.68 (0.55–0.84)< 0.0010.69 (0.54–0.87)0.0020.73 (0.59–0.90)0.003
Low education level (≤12 years)1 1 1 
High education level (≥12 years)1.07 (0.86–1.32)0.541.06 (0.83–1.35)0.651.13 (0.91–1.39)0.28
HADS anxiety (range 0–21)0.99 (0.94–1.05)0.830.96 (0.91–1.02)0.221.01 (0.96–1.07)0.75
HADS depression (range 0–21)1.01 (0.95–1.07)0.731.01 (0.95–1.08)0.701.03 (0.97–1.10)0.31
Diagnosis subtype      
 AS1 1 1 
 PsA0.72 (0.56–0.92)0.0090.78 (0.59–1.04)0.090.79 (0.62–1.01)0.06
 uSpA0.90 (0.66–1.22)0.510.87 (0.63–1.20)0.390.95 (0.70–1.30)0.74
 IBD-related arthritis1.40 (0.68–2.88)0.361.55 (0.69–3.49)0.291.37 (0.66–2.82)0.39
BASFI (range 0–10)0.87 (0.83–0.91)< 0.001    
BASDAI (range 0–10)  0.92 (0.87–0.98)0.006  
EQ-5D (range 0–10)    6.42 (3.12–13.2)< 0.001

DISCUSSION

Seven of 10 patients with SpA met the WHO recommendations for physical activity, which was slightly higher in all patients studied in comparison with the Swedish population. Young women with SpA had a tendency to be less active than young women in the general population. Analyzing the physical activity patterns in SpA patients yielded different factors associated with being physically active at a moderate intensity several times per week compared to preferring to exercise at a vigorous intensity 2 to 3 times per week.

Several studies show that physical inactivity has considerable consequences for the individual in terms of pain, disability, and premature mortality (3–5). Furthermore, supporting physical activity is a cost-effective way to maintain health (1). The finding that the SpA cohort was more physically active compared to the general population is supported by an earlier Swedish report (26), while the opposite also has been reported (6, 31). Still, the total proportion of individuals not meeting the WHO recommendations of physical activity for health is large, both in the current SpA cohort (32%) and in the general population (35%) (7).

The RRs in the total SpA cohort indicated a higher activity rate among females compared to males in reaching the WHO recommendations, even though CIs were overlapping. This finding is not in accordance with earlier findings in the arthritis and general populations, where men and subjects with lower age more often met the WHO recommendations (6, 31–33). A possible explanation for our findings might be that, traditionally, patients with AS have been encouraged to do regular exercise, and women may be more inclined to follow these recommendations (34). An alternative explanation could be that men, in some aspects such as development of radiographic abnormalities, have a more severe disease (35), which may raise barriers for physical activity. A healthy lifestyle is fundamental for everyone, but it is even more important to identity groups at risk. The most notable finding was that the youngest female age group (18–29 years) showed the largest discrepancy from the general population, despite better disease activity and physical function scores, lower body mass index, and fewer previous smokers/smokers as compared with the older age groups found in post hoc analysis (data not shown). Splitting data on SpA subtypes showed that 44% of this youngest female age group were women with PsA. Young women with PsA in accordance with this were the least likely to meet the WHO recommendations. In a study of patients with RA, not analyzing effect by sex, younger age was associated with a higher level of physical activity (32). This finding on nonactive young women needs to be further explored and, if confirmed, may affect interventions and coaching strategies in the clinic.

There is evidence showing that exercising on a more vigorous level might be more important from a cardiovascular point of view (36). In the current questionnaire, the patients reported whether physical activity was performed at a moderate or a vigorous intensity. Comparing these 2 preferences showed different findings with regard to background factors and support that being a woman or a patient of higher age was associated with meeting the MI-PArec. In the current SpA cohort, the women reported a poorer perceived health, which is also confirmed by others (37–40). This, together with older age, may be possible explanations as to why these patients prefer to be physically active on a moderate level several times per week, while men in general were more prone to exercise at a more vigorous level (VI-PArec). Yet another explanation might concern activity preferences; women may prefer walking and gardening, while men may prefer running and other cardio-intensive sports, which increases the importance of investigating patients' motivation and preferences for exercise when coaching toward a healthier lifestyle.

It was found that better physical function, health-related quality of life, and lower disease activity were associated with meeting both the MI-PArec and VI-PArec. It has previously been shown that physical activity and exercise are associated with improvements in physical function and better health-related quality of life, both in arthritis diseases and in the population (41–43). More recently, increasing evidence describing associations between physical activity, physical fitness, and cardiovascular risk factors/events has been shown, even though a causal relationship is yet to be established in the arthritis population (44). Due to the cross-sectional nature of this study, the causality is not clear, but it may well be that poorer perceived physical function, health-related quality of life, and experiencing a high disease activity explains why patients are not active at either a moderate or a vigorous level. These findings present a challenge to all health professionals to tailor activities suitable and manageable for these patients.

In the regression analysis it was also found that the subtype PsA was negatively associated with meeting the VI-PArec. The heterogeneous clinical pictures in the SpA subtypes may be an explanation, and, unfortunately, we do not know the proportion of patients with PsA experiencing axial symptoms and peripheral joint symptoms, respectively. Typically, patients with PsA more often experience inflammations in peripheral joints compared to the other SpA subtypes (45), which may affect the possibility of being physically active to a greater extent. Furthermore, the AS group has historically received advice on exercising as an important part of the treatment despite pain, which may be in contrast to patients with PsA, where advice concerning exercise has been more cautious (22).

One strength of this study is that we included patients within a well-defined geographic region that includes almost 1.2 million inhabitants, which enables us to study rather detailed age and sex characteristics of the different subtypes of SpA. In contrast to most previous studies based on patients identified at rheumatologic clinics, we also included cases mainly seen in primary care and thus minimized selection bias.

Since the study had a cross-sectional design, causality is uncertain for the demonstrated associations. Furthermore, the heterogeneity of the SpA group (except perhaps AS) makes interpretation and implementation of the findings difficult. Nonresponders may affect the generalizability. On the other hand, the response rate was similar to other questionnaire-based population studies (7, 11) and similar to the response rate in the national health survey with a rate of 50%, which was the study used for calculating the RRs in the current study (7, 11). Both studies had a large number of nonresponders, but if one makes the assumption that the nonresponders have similar patterns of physical activity in the current survey, as in the national health survey, the RRs would be valid despite substantial nonresponder rates. The large sample sizes give very robust results with relatively narrow CIs for all subgroups except the smallest, which is IBD-related arthritis. The latter was still included in the presentation in order to make the SpA family complete (21). Yet another limitation concerns the validity of patient-reported physical activity versus observed physical activity (e.g., with accelerometry), and it is well known that there are difficulties comparing these 2 methods (46–48) even if none of the measures are proposed as the gold standard (48). However, patient-reported physical activity is internationally accepted, supported by the WHO, and is often the only possible way to study this phenomenon in larger surveys (6, 8, 11, 25, 26, 31–33, 36, 41, 49). In studies comparing accelerometer data with patient-reported data, both measures provided qualitatively consistent information of health-enhancing physical activity at a population level, even though absolute values differed with a modest overestimation of physical activity when using self-reported information (46). Therefore, the results in the current study have been compared only with other studies using self-reported data. Even so, comparisons between studies of patient-reported physical activity measures are challenging, since slightly different measures have been used. Since the questions used in this study were similar to those used in the national health survey, we believe that such methodologic problems would have only marginal effects on our results. Unfortunately, the newly updated recommendations from the WHO on physical activity for health did not exist when the survey was conducted in 2009 (25).

In conclusion, many patients with SpA do not meet the minimum requirement of physical activity to promote and maintain health as recommended by the WHO, which was especially notable in young women with PsA. Different factors influenced the preferred intensity of exercise performance. This information is important and can guide health professionals in their work when coaching patients with SpA to achieve a healthier lifestyle.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Haglund had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Haglund, Bergman, Petersson, Jacobsson, Strömbeck, Bremander.

Acquisition of data. Haglund, Petersson, Jacobsson, Strömbeck, Bremander.

Analysis and interpretation of data. Haglund, Bergman, Jacobsson, Bremander.

Acknowledgements

The authors would like to acknowledge the support from the Faculty of Medicine, Lund University, Sweden.

Ancillary