Using Parkrun Age Grading as an Indicator of Adolescent Health: A Pilot Cross-sectional Comparator Study

1. INTRODUCTION

Prescribing Physical Activity (PA) in nature settings such as parks to improve health and wellbeing is gaining attention globally [1] and in New Zealand (NZ) [2]. Parkrun is increasingly recognized as a valuable form of PA prescription due to its numerous health benefits, including improvements in cardiorespiratory fitness (CRF), body composition, mental well-being, and social connec- tedness [3, 4]. Parkrun is thus gaining attention as a PA prescription for health and wellbeing [5]. Improvements in CRF are of particular importance as CRF is identified as a critical predictor of all-cause mortality and cardiovascular disease (CVD) mortality [6, 7]. Low CRF levels are associated with an increased risk (~30%) of developing at least 25 common long-term physical and mental health conditions [8-10], including stress [11]. PA also improves muscular strength (STR), which is a separate indicator of health status, risk of all-cause, and CVD mortality [12].

In adolescents, CRF and STR are robust and potent health markers [13, 14]. CRF and STR are both associated with adolescent quality of life and STR with family relationships [15]. There has been a global decline in CRF and STR among adolescents [16], with obesity rates in this population tripling over the past 40 years [17]. 80% of adolescents globally are classified as insufficiently active [18]. NZ faces particular challenges, ranking second among OECD countries for the percentage of overweight and obese 5–19-year-olds (33%) [19]. This is particularly concerning in this population as the underlying origins of many long-term conditions are increasingly attributed to factors encountered in younger years [20]. A recent review [13] describes the compelling evidence associating CRF and STR in adolescence with all-cause, CVD, and cancer-specific mortality in later life. Adolescent CRF and STR levels are now a global concern [21, 22].

Differences in running performance due to age and sex are well documented [23, 24], with peak 5000m run performance occurring at 27 and 29 years of age for males and females, respectively [25]. Parkrun attempts to account for age and sex differences by including Parkrun age grading (PAG). PAG is a percentage score per- formance metric generated for each runner at each event that has been adjusted for age and sex. This allows participants to compare their results to world records for their demographic group. PAG also allows each runner to compare personal performance against all other runners, even though they might be of different age and sex. A higher PAG score indicates better performance. While PAG has been linked to self-reported health benefits [4], its usefulness as a health marker has not been thoroughly studied, particularly among adolescents.

In NZ, ongoing patterns of ethnic socio-economic deprivation are major determinants of health inequities [26]. Deprivation is a small geographical area measure of socio-economic disadvantage relative to other small geo- graphical areas, estimated across eight factors: communi- cation, employment, home-ownership, income, living space, qualifications, support, and transport [27]. NZ adolescents from higher deprivation areas, when compared to those from low deprivation areas, have a greater risk of obesity and obesity-linked morbidities [28-30]. The incidence of NZ adolescents developing Type 2 Diabetes Mellitus (T2DM) is increasing and dispro- portionally increasing in high deprivation areas [31].

This pilot study aims to assess whether PAG can serve as a health marker for adolescents, focusing on a sample from a low-deprivation area in NZ. The study aims to examine the relationships between PAG and various health markers, including CRF, STR, body composition, HbA1c levels (a marker for diabetes), stress, and family relationship functioning. The study will also compare PAG with the Ruffier Dickson Index (RDI), a well-established marker of CRF in adolescents that has been recommended for clinical use [32, 33]. This research may provide valuable insights into how Parkrun and PAG could be used to monitor and improve adolescent health, particularly in populations affected by health inequities.

2. MATERIALS AND METHODS

2.2. Testing Day

All participants attended a testing day and were instructed to avoid exercise and caffeine and to not eat in the two hours prior to their testing appointment. All participants received the same sequence of testing protocol.

2.2.1. Body Composition

Participants' height was measured using a portable stadiometer to the nearest 0.5 cm. Weight (Kg), body fat (%), and skeletal muscle mass (Kg) (SMM) were then determined using ACCUNIQ BC380 (Selvas Healthcare, Seoul, 08594, Korea) bioelectrical impedance scale. BMI was calculated as weight (kg) / height² (m²). Obesity categories were determined by BMI percentiles on the WHO growth reference for 5–19-year-olds [34].

2.2.2. Cardiorespiratory Fitness

Participants were then measured for the Ruffier Dickson Index (RDI) using the previously validated protocol [35]. RDI is calculated from three different heart rate (HR) beat per minute (BPM) measurements (HR1) average BPM from 15 s HR at rest, (HR2) the maximum BPM HR recorded during the immediate 15 s recovery following a standardized 30 squat test (40 squats·min^-1, knee flexion 90°, for 45 s), and (HR3) the average BPM from 15 s HR at recovery from 1 min. The calculation is:

HR during RDI testing was recorded using Polar OH1 sensors (Polar, Polar Electro, Kempele, Finland).

2.2.3. Strength

Hand grip STR was measured using a Hand Dynamometer (Jamar Digital Plus, SI Instruments, Hilton, Australia) while sitting following the American Society of Hand Therapists recommendations [36]. Three repetitions with the dominant hand were allowed, and 30 seconds of rest between repetitions, with no verbal encouragement given. The maximum value from the three repetitions was used.

2.2.4. Hb1Ac

Hb1Ac was used as a measure of blood glucose control by Cobas B101 (Roche Diagnostics, Auckland, New Zealand). A registered nurse took blood and performed the analysis as per manufacturer instructions.

2.2.5. Perceived Stress

Participants then completed the ten-item perceived stress scale, a validated questionnaire designed to measure perceived mental stress over the previous 4 weeks [37]. Stress was calculated and categorised as previously described [38]. Scores within the 0-13 range are low stress, the 14-26 range is moderate stress, while over 27 is high stress.

2.2.6. Family Assessment Device

Finally, participants completed the 12-item General Functioning Subscale of the McMaster Family Assessment Device (FAD). The subscale is validated as a single reliable measure to assess perceptions of family functioning [39]. Scores range from 1.0 (best functioning) to 4.0 (worse functioning), while any score ≥ 2.0 was used to classify a poorly functioning family.

3. RESULTS

Of the 25 participants’ 17 were female, and 8 were male. Participants' BMIs (Table 1) were underweight (n=2), healthy (n=8), overweight (n=9), and obese (n=6). 3 participants (two females and one male) did not complete RDI testing.

PAG had no association with RDI (r = -0.18, p = 0.457). PAG and RDI were examined for associations with other health markers (Table 2).

As only FAD returned a significant D'Agostino & Pearson omnibus normality test, we examined FAD inde- pendently using Spearman's rank correlation coefficient. FAD was not associated with PAG (r=-0.40, p=0.07). As neither stress nor FAD was associated with either RDI or PGA, we subsequently examined associations between PSS and FAD to gain a better understanding of these two markers in our cohort (Fig. 1). Only three participants recorded low (<14 PSS) stress and 22 participants (88%) recorded moderate (n=13) or high (n=9) stress. Seventeen participants (68%) recorded their family functioning as problematic (<2.0).

The median PAG (30.7%) and median RDI (11) were used to split the cohort into lower (n=11) and higher (n=11) health groups resulting in contrasting differences between groups (Table 3). The largest effect and difference between groups was for the HbA1c level when using PAG (Fig. 2).

4. DISCUSSION

This pilot study investigated whether PAG is a marker for health in adolescents living in a high-deprivation area. Firstly, we aimed to examine if PAG has relationships with a holistic definition of health, including CRF, STR, body composition, Hb1Ac levels, perceived stress, and family relationship functioning. We found that PAG had a positive linear relationship with SMM (r=0.47, p=0.027) and inverse linear relationships with body fat (r=-0.65, p < .001) and Hb1Ac (r=-0.45, p=0.037). Secondly, we sought to compare PAG with RDI, an established CRF marker in adolescents. Both PAG and RDI appear useful for classifying some aspects of physical health in vulnerable adolescents. However, neither PAG nor RDI are markers for all aspects of health in this population.

Cardiopulmonary exercise testing is the definitive method to assess CRF [42], though it is rarely used in adolescent populations because it requires specialist gas analysis equipment and trained personnel, and it is also labour-intensive [43]. Our local ethics committee deemed this method unethical in our vulnerable population. In this study, we used both PAG and RDI to classify CRF. RDI has been previously shown to have validity and repeatability [35] and may be useful in clinical practice [32, 33]. Our findings suggest that PAG and RDI capture different aspects of health. In our study, only RDI had a relationship with strength (r=-0.51, p=0.02). In contrast, only PAG had relationships with body fat (r=-0.65, p<0 .001), SMM r=0.47, p=0.027), and HbA1c (r=-0.45, p=0.037). When RDI was used to categorise health status, our high health group, compared with the low health group, exhibited significantly (p=0.029) lower BMI (24.1 vs. 28.3 kg/m²) and significantly (p=0.036) higher strength (36.2 vs 30.5 kg). In contrast, when PAG was used to categorise health status, the high health group, compared with the low health group, exhibited significantly (p=0.006) lower mean HbA1c (34.4 vs 38.1 mol/mol). We speculate that these differences may be partially explained by the fact that RDI performance relies on strength to a greater degree than PAG. RDI requires 30 parallel squats, and stronger individuals and/or those with a lower BMI who encounter less resistance may find squatting less challenging. We suggest both markers are potentially useful as measures of health status, though capture different aspects of health.

The participants self-identified ethnicity in our study were > 90% Māori or Pasifika. In NZ, Māori and Pasifika experience a greater risk of obesity, CVD, and T2DM [30, 44]. The HbA1c differences (34.4 vs 38.1 mol/mol, p=0.006) between low and high PAG groups may be of particular interest to clinicians. While PAG has been previously suggested as a CRF marker [4], our results suggest it is also a marker for blood glucose homeostasis in this population. CRF is the best single predictor for CVD outcomes [6, 7], and a reduced risk for T2DM is also reported [45]. We speculate that both healthcare professionals and patients may find PAG a useful gauge for health via improved CRF. Parkrun calculates and allows individual current PAG and changes in PAG to be easily tracked via a free website. However, as Parkrun uses a variety of courses and terrain, including hills on some occasions, PAG will not be interchangeable across all locations. Rather, recommending a focus on improving PAG at the same local Parkrun should be the focus for improved health when seeking to implement PAG as a marker.

In this study, neither PAG nor RDI had any relationship to stress or FAD. However, it’s worth noting that the PAG relationship with FAD was close to significance (p=0.07), and in a larger sample, this may have been significant. Regular PA is known to decrease psychological stress [11], and such stress is associated with a growing number of physical diseases [46]. In NZ, stress measures are worsening [47], and stress levels have recently been compounded by the COVID-19 epidemic [48]. However, Parkrun mitigation during the epidemic was successful [49]. In our study, stress and FAD had a moderate positive, linear relationship (r=0.62), with 35% of the variance in stress explained by FAD. The lack of any relationship between PA measures and stress in adolescents has been previously reported [50]. Clinicians should be aware that PA levels may not be a major factor for stress in similar populations.

The major limitation of our study is that we were not able to measure CRF in a laboratory using the gold standard method as we originally intended due to ethical concerns. Further limitations are a small sample size due to this being a pilot study. Further research is now required to examine if an increase in PAG improves health. The strength of the study is that PAG was calculated independently by Parkrun, and participants and researchers were blind to PAG when testing for health.

CONCLUSION

In conclusion, we have shown for the first time that PAG is a marker for health in adolescents living in high-deprivation areas. We demonstrated that PAG has relationships with body composition and Hb1Ac in this vulnerable population. Both PAG and RDI appear useful for classifying some aspects of health in vulnerable adolescents. However, stress and FAD were not associated with either of these markers. PAG is thus a marker for physical health, although it does not capture all aspects of health. This research has provided valuable insights into how Parkrun and PAG could be used to monitor and improve adolescent health.

REFERENCES