Introduction
Worldwide, the numbers of patients undergoing total hip or total knee arthroplasty (THA or TKA) for hip or knee osteoarthritis (OA) are rapidly increasing. Overall, the outcomes are favourable, with a large majority of patients having less pain and improved physical functioning after surgery [
1‐
4].
Although the benefits of THA and TKA are well documented for pain and function, relatively little is known on their impact on one specific aspect of physical functioning, i.e. physical activity (PA). Just like for any other individual, achieving and maintaining a sufficient level of PA is important for patients with hip and knee OA with respect to their potential general health benefits. Moreover, in patients who undergo THA or TKA, PA may have an additional beneficial effect on the quality of the bone, which in turn may prevent complications such as early loosening [
5‐
8]. In addition, PA may have a positive effect on muscle strength and range of motion of the affected leg [
9].
With respect to the literature on PA after THA or TKA, de Groot et al. demonstrated in 84 THA and TKA patients that 6 months post-operatively PA levels as measured with an activity monitor did not significantly differ from the preoperative activity levels [
10]. Harding et al. [
11] found similar results when measuring PA by means of an accelerometer in 63 American THA and TKA patients before surgery and 6 months post-operatively. Kahn and Schwartzkopf [
12] found no difference in PA as measured with an accelerometer between those on the waiting list for TKA and those who had TKA 2 years earlier.
By using the patient-reported University of California at Los Angeles (UCLA) activity questionnaire, Baumann et al. [
13] found that both THA and TKA patients were regularly active on moderate to high levels after on average 6–12 months after surgery. This finding is supported by Dahm et al. [
14], who reported that 5.7 years after surgery TKA patients had an average physical activity score of 7.1 out of 10, with 10 being highly active. In contrast to these findings, Kahn and Schwartzkopf [
12] observed, using an accelerometer, that adherence to health-enhancing PA guidelines was only 5% in persons with TKA. In all of these studies, a comparison with the general population was lacking.
Two Dutch studies compared patient-reported physical activity in THA [
15] and TKA [
16] patients at 1–5 years post-operatively to that of age and gender-matched controls. It was found that in THA the proportion of persons reaching the Dutch Public Health Physical Activity guideline (the “Nederlandse Norm Gezond Bewegen”, NNGB) was similar to that of matched controls [51.2% (THA) vs 48.8% (controls)] [
15], whereas in TKA patients the proportion of patients adhering to the guideline (54.5%) was significantly lower than that of the matched control population (63.7%) [
16]. However, these studies did not take BMI into account, whilst BMI is one of the determinants of physical activity [
17].
Given the lack of knowledge on post-operative PA levels after total joint arthroplasty (TJA) compared to the general population, the aim of the present study was to compare the minutes of PA and proportion meeting the public health guidelines of THA and TKA patients to those of the general Dutch population. Moreover, factors other than TJA, possibly contributing to levels of physical activity, were also evaluated.
Patients and methods
Design and recruitment of subjects
This cross-sectional, multicentre study concerned a comparison of PA levels of THA and TKA patients approximately one year after surgery with those of the general Dutch population. The data from the population of patients with THA and TKA were obtained from a study primarily aiming to make an inventory of the use of physical therapy and the presence of comorbidity [
18], whereas the data from the general population were obtained from the Dutch National Bureau of Statistics (in Dutch: Centraal Bureau voor de Statistiek, CBS).
Since the survey had to be filled in only once by patients, it was judged to fall outside the remit of the law for Medical Research Involving Human Subjects Act; MO [in Dutch; Wet medisch wetenschappelijk onderzoek met mensen (WMO)]. An exemption for medical ethical review was therefore given by the Medical Ethical Committee of the Leiden University Medical Center. The health monitoring conducted by the CBS commissioned by the Dutch Government also falls outside the remit of the WMO. The study was conducted in accordance with the Handbook for Good Clinical Research Practice of the World Health Organization [
19] and the Declaration of Helsinki principles [
20].
Patients with THA or TKA
The patient data were obtained from a cross-sectional study performed in 2012, including patients who underwent THA or TKA for hip or knee OA in 2011 in four different hospitals in the Leiden region (Leiden University Medical Center in Leiden, Rijnland Hospital in Leiderdorp, Groene Hart Hospital in Gouda and Reinier de Graaf Hospital in Delft, the Netherlands). Patients receiving THA or TKA for reasons other than end-stage OA (such as fracture or rheumatoid arthritis) were excluded from the study, as well as patients undergoing revision surgery.
Between July 2012 and October 2012, all patients operated in 2011 were approached by mail by their orthopaedic surgeon, resulting in a range of post-surgery time of 7–22 months. The orthopaedic surgeon sent all eligible persons an invitation letter, information leaflet, informed consent form, survey and pre-stamped return envelope. Patients who returned the envelope with a completed survey and signed informed consent were included in the study.
Data general population
Data from the Dutch general population were provided by the CBS and were derived from a nationwide survey on general health (Gezondheids-enquête) [
21]. This questionnaire is annually administered to a representative sample of ±8.000 Dutch inhabitants and is the prime health monitor tool of the Dutch government [
22].
The selection of participants is drawn from municipality registers. Persons living in institutionalized homes (e.g. nursing homes) are excluded. For the present study on physical activity, only data were selected from 2011, i.e. the same year as the data from patients with THA or TKA, and from respondents who were over 40 years of age, as none of the persons with arthroplasty was aged below 40.
Assessments
Included in both surveys were the following variables or questionnaires.
Socio-demographic and basic health characteristics
Demographic variables included: age, gender and marital status (split into either married or not married). The height and weight of the patient were asked in order to calculate the body mass index (BMI). Smoking status (non-smoker, ex-smoker and smoker) and educational level (low (elementary school, lower secondary education), medium (secondary school or college) or high (higher secondary education or university)) were recorded.
Physical activity
PA was assessed using the validated Dutch version of the short questionnaire to assess health (SQUASH) [
23,
24]. The SQUASH records the total amount of minutes per week (min/week) spent on PA in an average week in the past 12 months regarding eight different domains of active life: commuting, work activities, walking, cycling, gardening, odd jobs, household and sports. With the aid of the compendium of Ainsworth [
25], PA can subsequently be categorized into light, moderate or vigorous intensity. Using this information, it is possible to define whether an individual adhered to the Dutch Public Health recommendation (NNGB) for PA (30 min of moderate intensity PA on at least 5 days per week) [
26].
Quality of life (QoL)
QoL of the persons with a THA/TKA was assessed with the Short Form 36 (SF36) questionnaire, whilst the QoL of the general Dutch population was assessed with the Short Form 12 (SF12) [
27,
28]. The SF36 outcomes of the THA/TKA patients were transformed to SF12 outcomes. From the SF12, two summary scales were derived: the physical component scale (PCS) and the mental component scale (MCS). The higher the score on these scales, the better the physical or mental functioning.
Comorbidity
The presence of comorbidity was assessed by means of a self-reported questionnaire of the CBS which comprised 19 different comorbidities [
29]. For every comorbidity, the participants of the survey were asked to respond with either yes or no to the question “Have you received any treatment for [condition] in the past year”. The included diseases were then clustered into three groups:
Musculoskeletal comorbidities
Severe back pain (including slipped disc), severe neck or shoulder pain, severe elbow wrist or hand pain, inflammatory arthritis or other joint conditions.
Non-musculoskeletal comorbidities
Asthma or COPD (chronic obstructive pulmonary disease), (severe) cardiac disorder or coronary disease, arteriosclerosis (abdomen or legs), hypertension, (consequences of) stroke, severe bowel disorder, diabetes mellitus, migraine, psoriasis, chronic eczema, cancer and urine incontinence.
Sensory comorbidities
Hearing impairments (group and face-to-face conversation), vision impairments (short and long distance) and dizziness in combination with falling.
Statistical analyses
The demographic and health characteristics of patients undergoing THA or TKA were each compared with those of the general Dutch population by means of two sample t tests or Chi-square tests, where appropriate.
Mann–Whitney tests were conducted to compare the min/week spent on PA for each of four different age groups (aged under 65, 65–69, 70–74 and 75+).
The min/week of PA was log-transformed to reach a normal distribution. Multivariable linear regression models were used to assess whether having had a joint replacement was associated with min/week spent on PA. Each analysis was done separately for THA versus the general Dutch population and TKA versus the general Dutch population. The antilog of the effect sizes (beta’s) is reported for both the analyses with the 95% confidence interval (CI).
Multivariable logistic regression was used to assess the association between the presence of a joint replacement and adherence to the Dutch public health physical activity guideline. These results are presented as odds ratio (OR) with the 95% CI. The analyses were done separately for THA versus the general Dutch population and TKA and the general Dutch population.
All models (multivariable linear and multivariable logistic regression analyses) were constructed using a stepwise method. Potential confounders for the level of physical activity, i.e. sex, BMI, age and education level, were included in the models.
The determinants of minutes per week spent on activities categorized according to the three different levels of intensity of physical activity were determined for the arthroplasty groups and the Dutch population separately, by means of linear regression models including the variable of interest and correcting for age and sex. These analyses were performed including the variables age, sex, BMI, education, non-musculoskeletal comorbidities, musculoskeletal comorbidities, sensory comorbidities, MCS, PCS and time since surgery.
The level of statistical significance was set at P < 0.05, and analyses were performed using the SPSS statistical package (version 20.0, SPSS, Chicago, IL).
Discussion
This study demonstrated that the presence of a THA was associated with more min/week spent on PA as well as better adherence to public health recommendations for PA (NNGB) when compared to the general population. TKA was found to only be associated with adhering to the NNGB when compared to the general Dutch population.
Overall, it seems the Dutch population spends more minutes per week on physical activity, but since the patient group differs from the general population the comparison between these groups should be adjusted. When adjusting for age, sex, BMI and education it is found that persons with THA do spend more minutes per week on physical activity and that persons with a THA and TKA are more likely to adhere to the Dutch guideline on physical activity, NNGB. That TKA is associated with the NNGB but not to the minutes per week activity can be explained by the level of intensity of the physical activity performed.
In the general population, more associations between potential determinants of physical activity and the actual numbers of PA reached statistical significance than in the arthroplasty groups. The lack of significance is probably due to the relatively small sample sizes in the arthroplasty groups, limiting the statistical power.
Our groups spent more min/week on PA than reported by two other Dutch studies (for THA in this study 2183 min/week PA, THA in Wagenmakers et al. 1601 min/week, for TKA in this study 2153 min/week PA, TKA in Kersten et al. 1347 min/week PA) [
15,
16]. In parallel, regarding the proportion of patients adhering to the Dutch recommendation for physical activity, the outcomes were more favourable in the present study (THA in this study 75.6% and THA in Wagenmakers et al. 51.2%; TKA in this study 72.8% and TKA in Kersten et al. 55%).
Both these latter two studies were done at 1–5 years post-surgery, and our study included patients within the first 22 months after surgery. As reported earlier by our group (Peter et al.), 43.5% of the THA patients and 50.5% of the TKA patients had post-operative physiotherapy for more than 3 months [
30]. This implies that a vast amount of our patients might still have intense training with aid of physiotherapists, motivating patients to adhere to the PA. As for the other two studies (Kersten, Wagenmakers), no data on prolonged post-operative physiotherapy are present, and thus, these patients might resume easier into their old, less active activity level.
A recent systematic review on physical activity after THA or TKA measured with accelerometers showed that the post-operative PA levels were lower in the arthroplasty groups as compared to healthy control participants [
31]. The differences in outcome could be because our sample of the general population might not be totally healthy and be less active than selected healthy persons. Also, as this study used a questionnaire whilst the systematic review concerned objective measures, participants might have caused some recall bias.
The general population in our study had an adherence rate to the Dutch PA of 67.5% which is comparable to reports from CBS published (Dutch adult population, 66% adhered to the Dutch public health physical activity guideline in 2012) [
32]. The minutes per week spent on PA in our study was also consistent with the numbers reported by CBS (2589 min per week in 2012 and 2525 min per week in our study for overall physical activity for the Dutch population) [
33].
Factors we identified as influencing the level of PA of persons with hip or knee arthroplasty (BMI, increased age, physical component score) are in line with the findings in a systematic review by Stubbs et al. [
34] regarding PA in patients with hip or knee OA. The inverse association of BMI on the level of PA shows that it is an important factor, as well as age and gender, to include in any case–control study [
17].
Low-impact activities like walking or cycling seem to protect against function loss and experienced pain from OA [
35,
36], in contrast heavy load activities might be a risk factor for the development of osteoarthritis, but also early implant failure although debate exists on the latter [
37,
38]. Since contradictory evidence exists on this topic, research into this field is necessary.
Current post-operative rehabilitation after a hip or knee arthroplasty is focussed at independent ambulation and regaining a normal walking pattern, which was deteriorated in the years before surgery due to the slowly progressing osteoarthritis. Secondary to this it aims at getting physically active patient. As mentioned before, about half of our patients reported to receive physiotherapy for more than 3 months after surgery [
30]. This might imply that these patients are more motivated to be active than the general population.
Another reason for the higher levels of PA in the arthroplasty groups might be the fact that PA is a risk factor for TJA [
39]. As shown by de Groot et al. [
10], the post-operative levels of PA did not significantly differ from preoperative levels, suggesting that PA levels of TJA were probably higher than those of the general population before surgery as well.
Finally, the patients filling in the questionnaire knew that the subject of the study was PA, whilst the Dutch general population had to fill in an elaborate list of questions including all aspects of life, with only a subset on PA. Thus, the patients in our study might have overestimated their PA.
The limitations of this study are potential overestimation of outcome measures and recall bias, due to using the SQUASH questionnaire, more objective measures like accelerometers should be used in future studies. Furthermore, the preoperative levels of PA should be taken into account as an important confounder for outcome as well. Thus, more valid comparisons with the general population are possible. Also, patients in our study who refused to fill in the questionnaires were not asked about their reasons as to why they declined to participate, and therefore, we have no information about any possible self-selection bias. In addition, the comorbidities of participants were all self-reported and we were unable to confirm the presence of comorbidities both in the general Dutch population and the arthroplasty groups.
The findings of this study give insights into the movement patterns of arthroplasty patients compared to the general Dutch population. Findings show that although a part of the arthroplasty patients adhere to the Dutch public health guideline, there is still a considerable group who should increase their PA levels.
Summary conclusion
Overall OA patients with a hip prosthesis have a higher level of activity compared to the general Dutch population when adjusted for age, sex, education and BMI.