Disability and functioning in primary and secondary hip osteoarthritis in Benin

Background In Africa, primary hip osteoarthritis seems to be less frequent than in Europe. Sickle cell disease is responsible for aseptic osteonecrosis of the femoral head associated with secondary hip osteoarthritis. Very little evidence is available on the influence of aetiology (primary and secondary) and radiographic status on pain and disability in a Beninese population with hip osteoarthritis. Objectives The aim of this study was to compare the impacts of aetiology and radiographic status on pain, disability and quality of life in a Beninese population with hip osteoarthritis. Method This was a descriptive cross-sectional study, including participants recruited in the Clinic of Physical Medicine and Rehabilitation at the National Teaching Hospital in Cotonou. Assessment was based on the International Classification of Functioning, Disability and Health model. The main outcomes were severity of osteoarthritis, pain, range of motion, muscle strength, gait speed and quality of life. Statistical comparisons between the aetiologies were performed using a t-test or rank sum test. One-way analysis of variance was used to test the effect of radiographic status. Results Forty-nine participants (26 women and 23 men; mean age [standard deviation] 40.5 [17.9] years) were recruited. According to the aetiology (59.2% and 40.8% of primary and secondary osteoarthritis, respectively), there were no significant differences for any of the outcomes. Grades I, II, III and IV osteoarthritis were observed in 22.4%, 14.3%, 26.5% and 36.7% of the participants, respectively. Participants with grade IV osteoarthritis were more affected than those with grades I, II and III based on the Kellgren and Lawrence classification. Conclusion Aetiology did not influence pain, gait speed or quality of life. Participants with grade IV osteoarthritis had more pain, were more limited in walking and had a more impaired quality of life.


Introduction
Osteoarthritis is a chronic condition that affects 11% of the general adult population and is the most common form of arthritis (Pereira et al. 2011). Hip osteoarthritis (HOA) is among the most common joint diseases and is therefore a major social and health problem (Danielsson & Lindberg 1997). Hip osteoarthritis is one of the main causes of disability in the elderly (Loeser 2010), with prevalence increasing progressively with age (Jordan et al. 2009;Oliveria et al. 1995).
However, little is known about the prevalence of HOA in Africa. As demonstrated by Dowman et al. (2012), HOA is often seen as a minor health problem and has been neglected in research and resource allocation throughout Africa despite potential related disabilities (such as abnormal gait patterns and lower physical function) and decreased quality of life (QoL).
According to their aetiologies, there are two types of HOA: primary and secondary. The primary HOA is idiopathic with no hip malformation. It is the most common HOA in Western countries. The development of primary HOA is favoured by a probable genetic predisposition that affects cartilage metabolism, with other contributing factors such as biomechanical constraints (Englund 2010). Secondary HOA is a result of changes in the microenvironment of the cartilage. Its aetiologies include congenital hip abnormalities, metabolic defects, infections and blood disorders such as sickle cell disease (Kpadonou et al. 2011;Oniankitan et al. 2009;Ouedraogo et al. 2015). This hereditary disease is more frequent in sub-Saharan Africa, India, the Middle East and Brazil (Lespasio, Sodhi & Mont 2019). Sickle cell disease is responsible for aseptic osteonecrosis of the Background: In Africa, primary hip osteoarthritis seems to be less frequent than in Europe. Sickle cell disease is responsible for aseptic osteonecrosis of the femoral head associated with secondary hip osteoarthritis. Very little evidence is available on the influence of aetiology (primary and secondary) and radiographic status on pain and disability in a Beninese population with hip osteoarthritis.
Several factors that would contribute to disability and pain caused by HOA have been identified (Arokoski et al. 2004;Juhakoski et al. 2008;Kondo et al. 2017;Steultjens et al. 2000;Thumboo, Chew & Lewin-Koh 2002). According to Dekker et al. (1992): [R]adiographically assessed degeneration of cartilage and bone is associated with pain and disability, but it appears that the association is rather weak Thus, it seems that status of the joint is not enough to explain pain and disability in patients with osteoarthritis. Arokoski et al. (2004) have shown that men with HOA have significantly lower muscle strength than their age-and sex-matched controls. Van Baar et al. (1998) showed that in HOA, disability was associated with muscle strength, joint range of motion (ROM), pain, ability to cope with pain and psychological well-being, but the level of cartilage and bone degradation assessed by radiography did not influence the disability.
In accordance with the literature, there are few studies in Africa, and particularly in the Republic of Benin, that have examined factors associated with functioning and disabilities in patients with HOA. In sub-Saharan Africa, secondary HOA occurs in a younger population, and aseptic osteonecrosis of the femoral head is one of the most reported risk factors (Kpadonou et al. 2011;Oniankitan et al. 2009;Ouedraogo et al. 2015).
The main purpose of this study was to compare the impact of primary and secondary HOA on body structures and function, activity and QoL. The secondary purpose was to explore the impact of radiographic state in modifying body structures, function and disability in Beninese participants with HOA. We hypothesised that Beninese participants with secondary HOA would be more affected with regard to impairments, activity limitations and QoL than those with primary HOA. We further hypothesised that greater degradation of cartilage and bones, assessed by X-ray, would be associated with more impairments, activity limitations and reduced QoL.

Materials and method Study design and setting
This is a cross-sectional, descriptive study. Participants were recruited from a rehabilitation clinic and assessed once through a series of tests and questionnaires.

Participants
This study focused on participants with HOA in the Republic of Benin, a West African country.
Participants were recruited from the Clinic of Physical Medicine and Rehabilitation at the National Teaching Hospital in Cotonou. Participants with bilateral or unilateral symptomatic HOA diagnosed by radiography were recruited. In participants with bilateral HOA, the most painful hip was selected for analysis. The diagnosis of HOA was based on the American College of Rheumatology classification criteria and/or Kellgren and Lawrence (1957) radiographic classification (Vignon et al. 1999). Participants with total hip replacements, neurological or psychiatric disorders or any deficit that would prevent them from completing the questionnaires and assessments were excluded.

Recruitment procedure
Participants with an HOA diagnosis were initially identified from the rehabilitation care registry, focusing on the period between 01 June 2008 and 30 April 2019. From a contact list of 147 eligible participants, 69 were reached and invited to participate in the study. Out of the 69 reached, 4 died (information given by family members), 11 had total hip replacements and 2 refused to participate. The remaining 52 participants who agreed to participate were invited for a physical medicine consultation. After the screening consultation, three participants were excluded because they did not meet the selection criteria (diagnosis not consistent with HOA).
Finally, 49 participants meeting the inclusion criteria participated in the study.

Outcomes
Assessments were based on the International Classification of Functioning, Disability and Health (ICF) model (WHO 2001). Applying this model to HOA will enable identification of the impairments, activity limitations and participation restrictions experienced by the participants (Figure 1).

Measurements of International Classification of Functioning, Disability and Health body structures and function
The state of cartilage and bone was assessed by radiography with the Kellgren and Lawrence (1957) radiographic classification (Vignon et al. 1999). Pain was assessed with the Oxford hip score (OHS) and the numerical rating scale (NRS), hip ROM with goniometer and hip strength with the Medical Research Council (MRC) scale.
The OHS is a standard patient-reported outcome measure, developed to assess function and pain in participants with HOA (Martinelli et al. 2011) or undergoing total hip replacement surgery, particularly in the context of clinical trials (Dawson et al. 1996;Murray et al. 2007).
The OHS has also been used for the assessment of participant outcomes, including physical therapy and use of joint supplements (Murray et al. 2007). The scale consists of 12 items with 5 categories of response, 0-4 (worst to best), with the overall score ranging from 0 to 48, where 48 represents the best score (Murray et al. 2007). The NRS is a unidimensional measure of pain intensity in adults, including those with chronic pain resulting from rheumatic diseases. The 11-point numeric scale ranges from 0, representing no pain, to 10, representing the most intense imaginable pain.
Hip ROM was measured by the goniometer, whose validity has been established (Prather et al. 2010). Hip flexion, adduction and abduction motions were measured in the supine position, with the opposite thigh fixed in a neutral position. Internal and external rotation of the hip was measured in prone (with the hip in a neutral position and the knee in 90° flexion) and sit position (with the hip and knee in 90° flexion). Hip extension was measured with the participant lying on one side (hip in a neutral position and knee in extension). Goniometer-based measure has shown high intrarater reliability (intraclass correlation = 0.82-0.95) except for adduction (ICC = 0.50-0.72) (Holm et al. 2000).
Hip muscle strength was measured using the MRC scale with a rating of 0-5. Psychometric studies of MRC have shown high reliability for osteoarthritis (intraclass coefficient, > 0.95) (Youdas, Madson & Hollman 2010). Participants were placed in a position that enabled support for their body, so that they could focus their efforts on the tested muscle. The tested muscle was first placed in an antigravity position. If muscles were too weak to move the segment against gravity, they were tested in the horizontal plane. The proximal part of the tested limb was stabilised to reduce the risk of compensation by other muscles than those tested. Manual resistance was applied in the plane of movement for the tested muscles. Resistance was applied by progressive pressure. In most cases, there was a long lever in the arm (Appendix 1).

Measurements of International Classification of Functioning, Disability and Health activity
Activity limitations were assessed by the 10-metre walk test (10MWT). The test was performed on a 14-metre pathway, where each participant was asked to walk at his or her own comfortable speed. The participant's performance was timed over the middle 10-metre course. The initial 2 metre served to reach a constant speed. The final 2 metre prevented the participant from decelerating prematurely (Juhakoski et al. 2008;Tani et al. 2018) (Appendix 2).

Measurements of International Classification of Functioning, Disability and Health participation
To assess the QoL, the short-form 36-item (SF-36) health survey questionnaire and West Ontario and McMaster (WOMAC) osteoarthritis index (Appendices 3 and 4) were used. The SF-36 is a generic validated questionnaire, acceptable for the long-term measurements of QoL in osteoarthritis and validated in French (Hayes et al. 1995;Walters, Munro & Brazier 2001). This measure allows quantification of the physical and mental components (PC and MC) of QoL. It was shown to be reliable and valid (McHorney et al. 1994;McHorney, Ware & Raczek 1993) and consists of eight subscales: physical function, role limitations resulting from physical problems, bodily pain, general health perceptions, role limitations resulting from emotional problems, mental health, social functioning and vitality. Each score ranges from 0 to 100, where 0 is the poorest possible health state and 100 is the best health state.
The WOMAC questionnaire was developed to assess pain, joint stiffness and disability related to osteoarthritis of the knee and hip (Bellamy et al. 1988). It contains 24 questions, 5 related to pain, 2 to stiffness and 17 to physical function. It provides an excellent overview of a participant's functional capacity and complements the more objective data provided by magnetic resonance imaging, arthroscopy, cartilage biopsy and radiographs.

Data analysis
All variables with a normal distribution and equality of variance were presented as means with standard deviation (SD). For non-normally distributed variables, non-parametric descriptive statistics were used, and the results were expressed as medians and 1st and 3rd quartiles (25% -75%). Statistical analyses were performed using SigmaPlot from SPSS (version 13.0). The significance level was set at p < 0.05.
T-test or Mann-Whitney test for non-parametric variables was used to determine if participants with secondary HOA are more affected in structure and function, activity and participation than those with primary osteoarthritis. To establish if the degree of degradation in the Kellgren and Lawrence radiographic classification was associated with impairments, activity limitations and decreased QoL, we used a one-way analysis of variance (ANOVA) if the normality and equality of variance tests passed or Kruskal-Wallis one-way ANOVA on ranks (if the normality and equality of variance tests failed) with one factor (factor = grade, that is, grade I, grade II, grade III and grade IV). A post hoc test was used to identify significant differences between the different radiographic grades.

Ethical consideration
Adult participants or parents of minor participants freely gave signed consent to participate in the study. The clinical study was conducted with authorisation from the hospital institution and informed consent of the patients. This is the current regulation in Republic of Benin. This study is part of a thesis project that has received the approval of the local

Characteristics of participants
Our sample of 49 participants comprised 26 women and 23 men, with a mean age of 40.5 (SD 17.9) years (range, 11-70 years). Participants had an average body mass index (BMI) of 25.6 kg/m 2 (SD 7.4); 40 participants had unilateral and 9 had bilateral HOA. According to Kellgren-Lawrence grades, the radiographic severity was of grade I in 22.5% (n = 11) of participants, grade II in 14.3% (n = 7), grade III in 26.5% (n = 13) and grade IV in 36.7% (n = 18).

Effects of aetiology on body structure and function, activity and participation
In terms of activity (10MWT), no statistically significant differences were observed between participants with primary or secondary HOA (12.2 s vs. 12.7 s) (

Effects of grades of X-rays on structure and function, activity and participation
Grade of HOA significantly affected function, activity and QoL (Tables 5 and 6). No statistically significant differences for age, weight, height, BMI, hip ROM or   (Tables 4 and 5).
Regarding pain, participants with grade IV HOA reported more pain (NRS, median 7 [IQR 6-9], p = 0.02; OHS, median 19 , p < 0.001) than those with grades I, II and III. Post hoc analysis revealed for NRS a difference between participants with grades I and IV (p = 0.02); for OHS there was a significant difference between participants with grades II and IV HOA on the one hand and participants with grades I and IV HOA on the other hand (p < 0.001) ( Table 5).
Regarding hip ROM and muscle strength, there were no differences between any of the groups (Table 5).
With respect to activity, a significant statistical difference was also observed. Participants' scores on the 10MWT were 9.9, 10.2, 10.3 and 12.3 s for grades I, II, III and IV HOA, respectively. Post hoc analysis revealed a significant difference between participants with grades I and IV HOA (p = 0.001) (  (Table 6).

Discussion
The purpose of this study was to compare the impacts of the aetiology of HOA (primary and secondary) and radiographic status in three domains of the ICF model in Beninese participants with HOA. There was no significant difference between aetiology on structures and functions, activity and QoL in this study. However, the level of cartilage and bone degradation significantly impacted the participants on function, activity and QoL.   In this study, the average age was 40.5 (17.9) years. This suggests that HOA in the Republic of Benin affects young adult people. This may be attributed to the role of haemoglobinopathy in the occurrence of aseptic osteonecrosis of the femoral head, which was reported in 20 of the 49 participants. In aseptic osteonecrosis of the femoral head, bone cells or osteocytes are affected by a metabolic disorder and their nutrition modified by a simple local reduction in circulation below the threshold that ensures their survival (Lespasio et al. 2019;Narayanan et al. 2017). This local reduction in circulation is thought to result from ischemic thrombosis of the artery of the round ligament, the nourishing artery of the femoral head (Lespasio et al. 2019;Narayanan et al. 2017). This observation was reinforced by Oniankitan et al. (2009), who showed that epiphysis and aseptic osteonecrosis of the femoral head seem to constitute the main risk factors for secondary HOA in sub-Saharan Africa.
Participants were overweight with an average BMI of 25.6 (SD 7.4) kg/m 2 . Obesity can exert an increased load because of increased body weight; however, there may be differential systemic effects depending on the degree of fat versus lean mass. Body mass index may be associated with HOA in our participants. Indeed, an increasing load on a hip joint resulting from overweight may affect the appearance of HOA. This is consistent with the observations from Heliovaara et al. (1993) and Jiang et al. (2011), who reported a positive association between an increased BMI and HOA.
Participants with secondary HOA were not more severely affected, as no statistically significant differences were found for pain, ROM and strength. The young age profile of participants with HOA induced by aseptic osteonecrosis of the femoral head combined with lower BMI compared to participants with primary HOA could explain the lack of differences in scores. Future investigations should be envisaged to support our findings.
Radiographic state had an influence on the pain, activity and QoL of participants but not on hip joint ROM or muscle strength. Participants with grade IV HOA felt more pain. Pressure on the subchondral bones, sclerosis, and cysts observed in patients undergoing severe joint deterioration may be the explanation for their greater pain. This finding suggests that pain intensity may be partly related to the degree of HOA, but this assumption is only partially supported by research findings. Summers et al.'s (1988) study has shown that the degree of objective disease severity (ODS), a set of criteria for the radiographic assessment of disease severity in osteoarthritis (OA), was significantly correlated with a higher level of pain on the McGill pain questionnaire. Dougados et al. (1996) has suggested that there could be a statistically significant correlation between clinical and radiological parameters evaluating HOA. In a survey, Arokoski et al. (2004) showed that the pain score was not correlated with the degree of radiologic severity of OA. Veenhof et al.'s (2012) systematic review showed conflicting evidence in association between pain and greater joint degradation in HOA. The literature suggests that other factors, such as physical and psychosocial disability (Hopman-Rock et al. 1996), muscle weakness and ability to cope with pain (Van Baar et al. 1998) or duration of limitation in normal activities and level of education (Thumboo et al. 2002), are associated with the severity of pain in HOA. Other studies need to be conducted to support this finding.
There was no statistically significant difference according to the grades of HOA in the affected hip joint ROM and muscular strength in this study. This observation is consistent with what was reported in the existing literature indicating reduced hip joint ROM in a population with HOA (Arokoski et al. 2004;Rydevik et al. 2010;Steultjens et al. 2000). According to Arokoski et al. (2004), hip joint ROM is a sensitive marker of the radiographic severity of osteoarthritis. He showed that the worse the deterioration of the hip, the lower the ROM value of the hip detected in abduction and in internal and external rotations. More research is required to corroborate these results. However, this muscle deficiency observed in the participants may be associated with pain and limitation of activity as described in the literature (Rydevik et al. 2010;Van Baar et al. 1998).
In terms of activity, a statistically significant difference in the spontaneous gait speed was observed between participants with grades I and IV HOA. This difference may be attributed to a combination of many factors: (1) a greater BMI (mean 26 [SD 7.4) kg/m 2 ]), more intense pain . It is therefore possible that the level of joint degradation of our participants and its consequences may lead to limitation of activity, resulting in a slow spontaneous gait. This slow spontaneous gait in participants, particularly in those with grade IV HOA, may be one strategy to counterbalance pain and joint load or joint deformities resulting from HOA severity. Several authors have reported the effects of HOA on activity but no strong association with radiographic state. For a few authors, there is low evidence for the association between level of physical activity and joint degradation (Veenhof et al. 2012), and radiographic OA correlated poorly with physical function (Thumboo et al. 2002). Constantinou et al. (2017) found that patients with mild to moderate HOA walk slower and the walking distance was significantly shorter during the 6MWT (Rydevik et al. 2010).
In a population survey, Odding et al. (1996) found that restricted ROM of several joint actions was associated with the presence of locomotor disability. According to Arokoski et al. (2004), people without HOA were significantly better at walking, ascending and descending stairs, and performing a 25-metre walk test. Future researches using more efficient tests need to be considered to confirm our observations.
The participants with grades I, II and III HOA had a better QoL than those with grade IV. Indeed, participants with grade IV HOA had more pain with a slower spontaneous gait than participants with grades I, II or III HOA. In accordance, Rydevik et al. (2010) reported that significantly lower healthrelated QoL in HOA patients is attributed to a reduction of hip ROM and an increase in joint stiffness. Arokoski et al. (2004) observed that the function and scores of WOMAC correlated (p < 0.05) with the degree of radiologically estimated severity of HOA. Some previous researches have reported that WOMAC is associated with the radiological severity of HOA (Bellamy 2005;Stucki et al. 1998).

Strengths and limitations
To our knowledge, this is the first study to explore the influence of aetiology and the radiographic state of HOA on the structures and functions, the activity and the participation of a Beninese population with HOA. This is a first step that may help to establish better strategies for the management of this population (policy of identification of patients with HOA, an efficient platform for the best care of patients, etc.). The knowledge of HOA's impact on the structures and function, activity and participation for a Beninese population with HOA may help clinicians to evaluate and develop rehabilitation programmes to improve functioning and reduce disability in this population.
However, this research has some limitations: (1) the sample size was small, and the recruitment was done exclusively in only one hospital reference centre; (2) other factors associated with HOA already studied in the literature were not taken into consideration; (3) the influence of environmental factors (home, work environment, social structures, communication and mobility services, laws and societal regulations, etc.) were not discussed in this study. Furthermore, the methodological choice to consider only the most painful hip in participants with bilateral HOA may constitute a bias. Further researches on the prevalence, association with other factors and treatment are needed and should be conducted on larger sample sizes. This will provide more information on the assessment and management of the population with HOA.

Conclusion
The purpose of this study was to compare the impacts of aetiology and radiographic state in three domains of the ICF model in a Beninese population with HOA. The results showed that aetiologies did not impact the body structures and function, activity or participation. The severity of OA based on radiographic state showed significant impacts on these variables. Higher degradation led to more impacts on different domains of ICF. Scores appliqués: 3 = excellent 2 = bon 1 = acceptable 0 = questionnable Présentation: Ce bilan de santé généraliste peut être utilisé en complément de bilans plus spécifiques.

Evolution du score:
Varie selon les items, afin de tester la vigilance du patient. La lecture des résultats fournit une appréciation sémantique.

Par comparaison avec il y a un an, que diriez-vous sur votre santé aujourd'hui?
Bien meilleure qu'il y a un an __ Un peu meilleure qu'il y a un an __ A peu près comme il y a un an __ Un peu moins bonne qu'il y a un an __ Pire qu'il y a un an __