Study Design
This study was a cross-sectional research design. The Cricket Health and Wellbeing Study was approved by the NHS Health Research Authority (NRES), London Stanmore Research Ethics Committee (REC 15/LO/1274).
Participants and recruitment
On March 2017, 28,152 current and former cricketers from all standards-of-play who were registered on the England and Wales Cricket Board national database, were invited by email to complete an electronic questionnaire. 2,598 people self-identified as meeting the eligibility criteria and gave written consented to participate in the Cricket Health and Wellbeing Study. Participants were eligible for inclusion in the Cricket Health and Wellbeing Study if they had played ≥1 cricket season and were aged ≥18 years. Despite consenting to participate, 365 did not meet eligibility criteria. A total of 2,233 were included in the HRQoL analyses. Due to the rarity of OA in individuals less than 30 years of age,(5) only participants aged ≥30 years (n=2,071) were included in OA and persistent joint pain analyses (Figure 1).
<Figure 1>
Questionnaire design
The Cricket Health and Wellbeing Study questionnaire was designed in collaboration with the England and Wales Cricket Board and piloted on current and former cricketers. Only small wording changes on cricket participation questions were altered, to enhance clarity, following piloting. The questionnaire was designed to evaluate five aspects of health and wellbeing (i. cricket-related injury; ii. joint pain and OA; iii. general health and disease prevalence; iv. physical activity; v. resilience, quality of life and flourishing). All participant data was de-identified and encrypted into a RedCap® (Research Electronic Data Capture) software-based database. The RedCap® software(15) used branching logic and allowed participants to save their current progress and complete at a later time.
Outcomes
Health Related Quality of Life
The Short Form 8 (SF-8) was used to assess HRQoL.(16) The SF-8 is a short version of the RAND 36-Item Health Survey (SF-36) 1.0.(17) The SF-8 is scaled and measured on the same point scale (0-100) as the SF-36, with 0 representing maximum disability and 100 representing no disability.(18) The SF-8 is an 8-item, self-reported HRQoL questionnaire comprising 4 domains that contribute to the Physical Component Score (PCS) (general health, physical functioning, role limitations due to physical health problems, bodily pain) and 4 domains that contribute to the mental component score (MCS) (vitality (energy/fatigue), social functioning, mental health, and role limitations due to emotional problems).(18) The PCS and MCS scores have high reliability (0.88 and 0.82 respectively) for use in the general United States population.(19) The PCS and MCS scores are calculated using a norm-based scoring algorithm that employs a linear T-score transformation with a mean of 50 and a standard deviation of 10, derived from 1998 United States general population norms. For summary measures, group mean scores below 47 can be interpreted as being below the average range for the general population.(19) The minimum detectable difference (MDC) for the PCS was found to be two points in a sample with lower extremity OA,(20) and the minimum clinically important difference (MCID) in the general population has been estimated to range from three to five points for the PCS and MCS.(21)
Physician Diagnosed Osteoarthritis
Osteoarthritis was assessed with the following question, ‘Have you ever been told by a doctor that you have osteoarthritis (wear and tear or joint degeneration)?’
Persistent joint pain
Persistent joint pain was assessed with the following question, ‘Have you had pain in your [left/right] [hip/groin, knee, ankle, shoulder, hand/finger, spine/back, other joint] on most days of the last month?’
Explanatory Variables
History of playing sport while injured
Participants responded to the following question, ‘Have you ever played sport injured, despite feeling like doing so might make the injury worse?’ Response options included ‘yes’, ‘no’, or ‘don’t know’. ‘Don’t know’ responses were excluded from the analyses. There was a total of 54 ‘don’t know’ responses, with no differences in participant characteristics between participants that responded ‘yes’, ‘no’, or ‘don’t know’.
Standard-of-play
Standard of play was assessed with the following question, ‘What was the highest standard of cricket that you played for at least one season?’ Response options included: international; county/premier league; academy or county age group; university; school; village or social; don’t know. Participants were stratified into recreational (university, school, village or social) and elite (international or county/premier league, academy or county age group).
Covariates
Covariates were identified through clinical reasoning and a review of the literature. Covariates included age, gender, cricket seasons played, playing status, number of joints injured, and number of orthopaedic joint surgeries. Playing status was assigned as either currently playing cricket (0) or no longer playing cricket (1). Number of joints injured was assessed with the following question, ‘Have you ever had any cricket related injuries leading to more than 4 weeks of reduced participation in exercise, training or sport? If yes, where? Please write the number of injuries for each joint and side’ Participants were stratified into never sustained a joint injury (0), and sustained a joint injury (1). Number of orthopaedic surgeries were assessed by asking the following question, ‘Have you ever had orthopaedic surgery (including bone, ligament or joint surgery)? If yes, where? Please write the number of surgeries for each joint and side.’ Participants were stratified into never had an orthopaedic surgery (0), and had an orthopaedic surgery (1).
Statistical Analyses
Continuous covariates were not assumed to linearly affect the outcome, and were modelled using fractional polynomials. As a result, multivariable linear regressions with fractional polynomial regressions were used to investigate the relationship between playing sport while injured and HRQoL (MCS and PCS scores) in all participants aged 18 years and over. Unadjusted and adjusted coefficients and 95% confidence intervals (95% CI) were calculated. All assumptions for fractional polynomial regression were evaluated and satisfied.(22) Logistic regression was used to investigate the relationship between playing sport while injured and joint health (physician diagnosed OA and the presence of persistent joint pain). Unadjusted and adjusted odds ratios (ORs) and 95% CI were calculated. All assumptions for logistic regressions were evaluated and met.(23) All regression models were adjusted for age, cricket seasons played, playing status, number of joints injured, and number of joint surgeries. All analyses were repeated in elite and recreational cricketer subgroups to address the second aim of this study.
Data were assessed for missingness prior to analysis. Missing data were calculated as total number and percentage of total data. Due to the low percentage of missing data (MCS: 6.5% PCS; 6.5%, OA: 1%, persistent joint pain: 1.1%, history of playing while injured: 2.4%, joint injury history: 3.8%, orthopaedic surgery history: 1.0%), complete case analyses were performed. ‘Don’t know’ responses (history of playing sport while injured: n= 54 (2.1%), OA: n= 67 (2.6%), persistent joint pain: n = 12 (0.5%), age: 0 (0%), playing status: 0 (0%), joint injury history: 15 (0.6%), orthopaedic surgery history: 4 (1.6%)) were not included in the regression analyses. All analyses were performed in R version 3.5.1 (R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/), using the naniar package for missingness assessment,(24) and the mfp package for fractional polynomial regression.(25)