A case series evaluation of standardized exercise dosage using a cycle ergometer on body weight and patient-reported outcome measures among overweight female patients with knee osteoarthritis

Abstract

[Purpose] This case series aimed to evaluate the effects of a standardized exercise dosage using a cycle ergometer on body weight and patient-reported outcome measures among overweight female patients with knee osteoarthritis. [Participants and Methods] We conducted a two-phase study. Exercise dosage was standardized using kilocalories per kilogram of body weight per week. Study 1 verified the feasibility and safety of progressive dosage in a single participant, while Study 2 used an ABCB-type single-case design involving three participants. The participants performed a cycle ergometer exercise with standard physical therapy. Body weight was the primary outcome, and the Japan Knee Osteoarthritis Measure was the secondary outcome. [Results] In Study 1, the participant safely achieved an energy expenditure of up to 10 kilocalories per kilogram of body weight per week; however, symptoms resembling knee buckling occurred at an expenditure of 11 kilocalories per kilogram of body weight per week. In Study 2, no adverse events were observed. One participant significantly reduced body weight during Phase B2, and all participants exhibited an improvement in Japan Knee Osteoarthritis Measure scores. [Conclusion] Standardized exercise dosage based on kilocalories per kilogram of body weight per week provides a safe and effective method for overweight female patients with knee osteoarthritis.

INTRODUCTION

Osteoarthritis (OA) is a multifactorial disease influenced by aging, being overweight, and joint injury. Its prevalence, along with the associated social burden, continues to increase globally. Current estimates indicate that approximately 250 million individuals worldwide are affected by OA¹⁾. In Japan, the number of patients with knee OA—the most prevalent form of the disease—is estimated at roughly 18 million^{2, 3)}. Knee OA is clinically characterized by persistent knee pain, lower-limb muscle weakness, functional impairment, and diminished quality of life⁴⁾. Body Mass Index (BMI) has been identified as a significant predictor of disease-specific quality of life⁵⁾.

Established risk factors for knee OA include being overweight, female sex, advanced age, prior knee injury, and occupations that impose substantial mechanical stress on the knee joint²⁾. Among these, being overweight has emerged as a particularly important and modifiable determinant⁶⁾. Furthermore, comorbidities such as hypertension, dyslipidemia, and impaired glucose metabolism have been shown to contribute to both the onset and the progression of knee OA. The risk increases in proportion to the number of comorbid conditions present⁷⁾. Being overweight is closely linked to these metabolic and cardiovascular comorbidities⁸⁾.

Therefore, improving overweight can be considered one of the primary goals of knee OA treatment. Promoting physical activity and implementing structured exercise interventions, along with dietary modifications, are essential to address this goal. Physical activity and exercise have been shown to improve obesity-related outcomes and patient-reported outcome measures (PROMs) of health⁹⁾.

Nonetheless, excessive exercise may increase mechanical loading on the knee joint and exacerbate the symptoms of knee OA. Evidence from a large-scale prospective cohort study of 27,813 patients with knee OA revealed that higher physical activity levels are associated with worsening disease and an increased likelihood of requiring surgical intervention¹⁰⁾. These findings suggest that the anecdotal principle of “more is better” does not apply in this context; instead, the critical issue is the identification of an “optimal dosage”.

An umbrella review assessing the appropriate dosage of exercise for knee OA reported insufficient evidence regarding the safety and effectiveness of exercise intensity and dosage¹¹⁾. A systematic review of randomized controlled trials in 2021 highlighted a lack of adequate reporting on adverse events related to exercise interventions and limited data concerning exercise dosage¹²⁾. Collectively, these gaps underscore the unresolved clinical challenges of ensuring safety while achieving maximal therapeutic benefit with minimal intervention.

In this exploratory case series study, our aim was to present preliminary findings evaluating the effects of standardized exercise interventions with controlled dosages on four overweight female patients with knee OA. The goal was to inform the design of future controlled trials.

PARTICIPANTS AND METHODS

The present study was conducted in two phases: 1) a preliminary study to verify the feasibility and safety of a standardized exercise dosage using a cycle ergometer, and 2) an ABCB-type single-case design to examine the effect of the exercise. The preliminary phase was undertaken with a single participant, whereas the subsequent phase involved three participants (Table 1).

Table 1. Participant characteristics.

		Age (years)	BW (kg)	BMI (kg/m2)	SMI (kg/m2)	K/L
Study1)		74	65.9	25.1	5.3	III
Study2)
	I	66	77.5	28.5	5.6	III
	II	72	59.8	26.6	5.1	III
	III	75	56.4	26.1	5.2	III

BW: body weight; BMI: body mass index; SMI: skeletal muscle mass index; K/L: Kellgren–Lawrence grade.

The eligibility criteria for this study included female patients aged 50 years or older diagnosed with medial knee OA and a BMI of 24 or higher and who attended the outpatient rehabilitation clinic between December 2024 and January 2025. Additionally, participants were required to have a Kellgren–Lawrence (K/L) classification of Grade 2 or higher. Exclusion criteria consisted of lateral knee OA, prior hip or knee arthroplasty, a history of myocardial infarction or angina pectoris, arrhythmia, rheumatoid arthritis, previous knee fracture or malignant tumor, inability to ambulate independently, dementia, history of central nervous system disease, and severe systemic conditions that impair mobility.

In the present study, overweight was defined as a BMI of 24 or higher, based on guidelines from the World Health Organization¹³⁾. This threshold takes into account the higher proportion of body fat typically found in Asian populations compared to that in their Western counterparts at the same BMI, which increases the risk of lifestyle-related diseases.

In this study, exercise dosage was standardized using kilocalories per kilogram body weight per week (kcal/kg body weight/week: KKW)⁵⁾. KKW is a useful measure for evaluating exercise dosage and dose-response relationships, enabling comparisons across individuals with different body compositions, and is widely employed in studies of exercise interventions. The physical activity guidelines developed by the Consensus Development Panel of the National Institutes of Health in the United States¹⁴⁾ propose exercise doses of 50% (4 KKW), 100% (8 KKW), and 150% (12 KKW). The American College of Sports Medicine guidelines recommend exercise equivalent to 8–12 KKW¹⁵⁾. For example, in an individual weighing 60 kg, 4 KKW corresponds to an energy expenditure of 240 kcal/week, 8 KKW to 480 kcal/week, and 12 KKW to 720 kcal/week.

Energy expenditure was calculated using the following formula: Energy expenditure (kcal)=1.05 × Metabolic Equivalent of Task (MET) × Time (h) × Body weight (kg). Here, 1.05 represents a conversion factor, reflecting that 1 MET is close to 1 kcal/kg/hour; time refers to the duration of each exercise session (hours), and body weight indicates the participant’s body weight (kg). MET values were derived from the Revised METs Table for Physical Activity issued by the National Institutes of Biomedical Innovation, Health and Nutrition in Japan¹⁶⁾. For example, a participant weighing 60 kg and performing 90-watt ergometer exercise equivalent to 6.8 METs for 33 minutes (0.562 h) per session would expend approximately 240 kcal (1.05 × 6.8 × 0.562 × 60 ≈ 240.7 kcal). Performing this activity twice per week yields an estimated weekly expenditure of approximately 480 kcal, corresponding to 8 KKW for a 60 kg individual.

In Study 1, the feasibility and safety of the exercise dosage outlined in the guidelines were evaluated. The dosage was progressively increased by 1 KKW per week, up to a maximum of 12 KKW, to assess its safety and confirm its practicability.

In Study 2, the ABCB-type single-case design was employed. The initial phase, designated as Phase A, involved standard physical therapy; Phase B included standard physical therapy, combined with the cycle ergometer; and Phase C involved standard physical therapy combined with resistance training. The initial implementation of the bicycle ergometer intervention was designated as Phase B1, followed by Phase C, and then a subsequent repetition designated as Phase B2. The entire intervention period lasted 20 weeks, with each phase lasting 5 weeks. Participants attended two sessions per week. There was no washout period or baseline re-establishment period implemented between the phases because the interventions were conducted consecutively in 5-week blocks. Therefore, the possibility of physiological or behavioral carryover effects between phases could not be excluded.

In Phase A, only standard physical therapy sessions were conducted, each lasting 20 minutes. The therapeutic protocol adhered to the clinical practice guidelines for knee OA as delineated by van Doormaal et al¹⁷⁾. The prescribed program encompassed joint range-of-motion exercises for the hip and knee, muscle strengthening targeting the quadriceps and hip abductors, balance training including static and single-leg standing, and fundamental functional exercises such as sit-to-stand transitions and gait training, as well as activities of daily living such as stair ascent and descent.

In Phase B1, the cycle ergometer was initiated at a resistance of 1 KKW, progressively increased to 4 KKW by week 4, and maintained at this level through week 5 besides the standard physical therapy. In Phase B2, the target workload was set at 4 KKW initially, with an anticipated progression to 8 KKW by week 5. We monitored the participants’ perceived exertion using the Borg Rating of Perceived Exertion (RPE) scale (6–20 points), with a range of 12–16 on the scale. It corresponded to approximately 60–80% of V̇O₂ max, considered a moderate level of intensity. This range is widely recommended in clinical guidelines as an effective and safe zone for aerobic conditioning¹⁸⁾. Saddle height was individually optimized for participant comfort. The RPE range of 12–16 is widely endorsed in clinical guidelines as an effective and safe zone for aerobic conditioning¹⁸⁾.

Besides standard physical therapy, resistance training during Phase C was conducted for lower limb muscles such as the quadriceps and hamstrings. This training was designed based on the findings of Hsu et al.¹⁹⁾ Each set comprised 10 repetitions, with a total of five sets executed. Exercise intensity was modulated either by increasing tension on the resistance band or by substituting with a band of greater thickness. The training criterion was defined as 10 repetitions performed at an intensity corresponding to the RPE of 12–16, consistent with the parameters employed for the bicycle ergometer.

The independent variables included age, Skeletal Muscle Mass Index (SMI), and Kellgren–Lawrence (K/L) classification as demographic and clinical attributes. Additionally, step counts and caloric intake (kcal) were recognized as potential confounders and assessed daily during the last week of each phase to calculate weekly mean values. These confounding factors were measured using the GoBe3 wearable device (GoBe3, HEALBE Inc., Redwood City, CA, USA). The GoBe3 integrates a three-axis accelerometer and a bioimpedance sensor to automatically record and estimate physiological data. Food intake (kcal) was estimated through an algorithm that analyzed subtle intracellular water fluctuations derived from bioimpedance changes, thereby enabling non-invasive monitoring and eliminating the need for self-reported dietary records. The validity of this methodology has been investigated at the University of California, Davis, and the Red Cross Hospital in Guangzhou, China. Evidence has demonstrated that the device exhibits moderate concordance with traditional dietary documentation methods and indirect calorimetry in estimating caloric intake²⁰⁾. In this study, the participants were instructed to wear the GoBe3 continuously from waking until bedtime for data collection.

The primary outcome measure was body weight, and the secondary outcome was assessed using the Japan Knee Osteoarthritis Measure (JKOM). The JKOM consists of five domains: severity of knee pain, pain and stiffness, activities of daily living, usual activities, and overall health status. This instrument is a self-administered questionnaire comprising 25 items, each scored on a 0–4 scale, yielding a total score ranging from 0 to 100, with higher scores indicative of greater functional impairment. The JKOM has demonstrated strong criterion-related validity in comparison with the internationally recognized Western Ontario and McMaster Universities Osteoarthritis Index²¹⁾. Body weight and SMI were measured using the InBody270 body composition analyzer (InBody Japan Inc., Tokyo, Japan). Adverse events were defined as any exercise-related incident requiring medical intervention or resulting in pain persisting for more than two consecutive days.

The primary outcome, body weight, was analyzed using the split-middle technique to establish a trend line (celeration line) for Phase A, subsequently comparing it with the extended trend lines. Phases B1, C, and B2 were then evaluated relative to Phase A, with data analyzed through a binomial test. Additionally, the Tau-U statistic, a non-overlap effect size, was applied to calculate effect sizes for the B1, C, and B2 phases in comparison with Phase A. Interpretation of Tau-U followed the criteria defined by Ninci et al.: Tau-U ≤ 0.2 denoted “small change (no effect),” 0.21 ≤Tau-U ≤ 0.6 denoted “moderate change (questionable effect),” 0.61 ≤Tau-U ≤ 0.8 denoted “large change (effective),” and Tau-U≥ 0.81 denoted “very large change (highly effective)”²²⁾. Statistical analyses were conducted using R version 4.5.0, with a significance threshold set at 5%.

For the secondary outcomes, JKOM and confounding factors (step counts and dietary intake) were assessed for each participant at four time points: post Phase A, post Phase B1, post Phase C, and post Phase B2. JKOM was assessed at week 5, 10, 15, and 20. Meanwhile, the confounding factors were measured daily during weeks 4, 9, 14, and 19. Weekly averages were calculated from the data collected each week and used as representative values for each phase.

The study complied with the ethical standards of the Kita Chiba Orthopedic Clinic Ethics Committee (2502-024) and the Teikyo Heisei University Ethics Committee (2024-048), with written informed consent obtained from all participants.

RESULTS

In Study 1, the participant successfully adhered to the exercise dosage, with weekly incremental increases, achieving up to an expenditure of 10 KKW by the tenth week of the 12 weeks intervention. However, when the participant performed an expenditure of 11 KKW in the week 11, the participant exhibited symptoms resembling a buckling knee when stepping over a threshold while returning home, nearly causing a fall. Consequently, the exercise dosage was reduced to an expenditure of 10 KKW and maintained thereafter, with safety prioritized. After this incident, neither adverse events nor persistent pain occurred. Furthermore, there were no withdrawals during the intervention, and the adjusted dosage of 10 KKW was continued without incident.

In Study 2, no adverse events were observed across the three cases throughout the study period. As compared to Phase A, during Phase B2, a significant reduction in body weight was observed in Case I. While Case I demonstrated a tendency toward increased step counts, no substantial differences were detected in Cases II or III. Caloric intake (kcal) remained stable in all cases, and JKOM scores consistently trended toward improvement across the study period. Detailed results are as follows.

For case I, a significant reduction in body weight was observed in Phase B2, using a binomial test with Phase A as baseline, (Fig. 1). No significant differences were detected in Phases B1 and C (Fig. 1). Tau-U revealed a very large body weight decrease in Phases B2 and C and a moderate body weight decrease in Phase B1 (Fig. 1). Step counts increased from Phase A to Phase B1, caloric intake appeared stable, and JKOM showed an improving trend across all phases (Table 2).

Fig. 1.

Case I results of the binomial test and Tau-U.

BW: body weight.

Table 2. Comparison of JKOM scores, step counts, and dietary intake.

Case	JKOM(score)				Step counts(step/day)				Dietary intake(kcal)
	Post A	Post B1	Post C	Post B2	Post A	Post B1	Post C	Post B2	Post A	Post B1	Post C	Post B2
I	92	86	80	73	4,832	5,731	6,531	6,232	1,920	1,824	1,980	1,800
II	87	84	76	72	5,188	4,888	4,333	4,721	1,944	2,044	1,882	1,902
III	68	68	63	56	4,811	4,075	3,575	4,811	1,958	1,938	1,911	1,888

JKOM was assessed at weeks 5, 10, 15, and 20, while step counts and dietary intake were recorded daily during weeks 4, 9, 14, and 19. JKOM: Japan knee osteoarthritis measure.

For case II, no significant reductions in body weight were observed in Phases B1, C, and B2 (Fig. 2). Tau-U indicated a small body weight decrease in Phase B1, a small body weight increase in Phase C, and a moderate body weight decrease in Phase B2 (Fig. 2). No essential differences in step counts were observed, caloric intake appeared stable, and JKOM showed an improving trend across all phases (Table 2).

Fig. 2.

Case II results of the binomial test and Tau-U.

Body weight data for week 2 were missing, and analysis was conducted after excluding the value. BW: body weight.

For case III, no significant reductions in body weight were observed in Phases B1, C, and B2 (Fig. 3). Tau-U revealed a moderate body weight decrease in Phase B1, a moderate body weight increase in Phase C, and a very large body weight decrease in Phase B2 (Fig. 3). No essential differences in step counts were observed, caloric intake appeared stable, and JKOM showed an improving trend across all phases, consistent with the other cases (Table 2).

Fig. 3.

Case III results of the binomial test and Tau-U.

Body weight data for weeks 1 and 4 were missing, and analysis was conducted after excluding these values. BW: body weight.

DISCUSSION

In this case series study, we investigated the effect and safety of the standardized exercise dosage intervention utilizing a bicycle ergometer in overweight female patients with knee OA. Study 1 comprised a pilot trial employing a structured, stepwise increase of exercise dosage, while Study 2 adopted the ABCB-type single-case design to examine changes in body weight and patient-reported outcomes.

In Study 1, exercise dosage was progressively increased according to the KKW metric, with the feasibility of achieving an expenditure of up to 12 KKW being measured. The participant successfully reached an expenditure of 11 KKW; however, upon returning home following the intervention, she experienced symptoms resembling knee buckling while stepping, leading to a near-fall incident. Regarding this event, as no similar incidents had been reported in patients who had been receiving continuous intervention prior to the start of the study, this occurrence was determined not as coincidental but rather a result of the 11 KKW expenditure being overloaded. This episode was likely attributable to compromised knee joint stability resulting from cumulative fatigue or transient impairment in neuromuscular control. In patients with knee OA, pain and muscle weakness may compromise postural stability, with symptoms potentially becoming more pronounced after high-intensity exercise. Hassanlouei et al. reported that exercise-induced fatigue following high-intensity exercise on a cycle ergometer may impair postural control around the knee²³⁾.

The exercise dosage was reduced to 10 KKW following this event, and the intervention was safely continued afterward. These findings suggest that although exercise prescriptions based on the objective KKW index are valuable, individualized adjustments accounting for physical condition and recovery capacity are indispensable. Moreover, the KKW-based exercise dosage recommendations outlined in existing guidelines are derived primarily from research in Western populations and may not be directly applicable to Asian populations. Asians, even at equivalent BMI levels, typically present with higher body fat percentages and lower muscle mass compared to their Western counterparts. This difference potentially results in relatively greater physiological and metabolic stress during exercise. Consequently, high-intensity interventions such as those involving 12 KKW expenditure may impose an excessive burden on elderly Japanese patients with knee OA. In light of these considerations, tailoring exercise dosage to individual medical history and physical capacity is imperative.

Furthermore, previous studies have indicated that abrupt increases in exercise dosage among overweight or obese individuals may elicit heightened appetite, thereby predisposing the individuals to overeating tendencies²⁴⁾. This phenomenon is hypothesized to arise from physiological mechanisms whereby elevated energy expenditure modulates hormonal regulation and neural reward pathways. This modulation may result in compensatory increases in food intake. In the present study, episodes of body weight gain exceeding 1 kg were also observed in Cases II and III. Causality remains uncertain given that dietary intake and daily step counts were not concurrently assessed across the study period. However, further investigations of the dose–response relationship in people with knee OA are essential to mitigate the risk of overeating induced by excessive exercise dosages. Moreover, comprehensive bidirectional monitoring of both exercise behavior and dietary intake will be indispensable for a precise evaluation of body weight reduction.

Study 2 employed an ABCB-type single-case design to compare outcomes across phases, with body weight as the primary endpoint and JKOM as the secondary endpoint. The principal findings were as follows: relative to Phase A, Case I exhibited a significant reduction in body weight in Phase B2, and all cases consistently trended toward improved in JKOM scores.

In Case I, a significant reduction in body weight was observed in Phase B2, as compared with Phase A. The longitudinal comparison of average body weight revealed a reduction of 1.3 kg from Phase A to Phase B2, accompanied by a very large effect size (Tau-U=−1.000). Although, the effect in Phase B1 was limited, probably attributable to the relatively high BMI and SMI in this case, which reflect greater metabolic potential compared with that in the other participants. These findings suggest that interventions for body weight reduction yield greater efficacy in individuals with higher BMI, as reported by Barte et al²⁵⁾.

Moreover, the presence or absence of sarcopenia has been proposed as a factor influencing body weight reduction²⁶⁾. Given that skeletal muscle tissue exhibits higher metabolic activity than adipose tissue does, the preservation or augmentation of skeletal muscle mass may potentiate body weight reduction by enhancing energy expenditure. This mechanism underscores the critical role of skeletal muscle mass in determining body weight reduction efficacy. In females, an SMI below 5.4 kg/m² is indicative of sarcopenia²⁷⁾. Notably, only Case I demonstrated an SMI exceeding this threshold, further supporting the aforementioned interpretation. Although no significant body weight reduction was observed during Phase B1, a marked reduction was identified in Phase B2 of the participant, likely attributable to sufficient skeletal muscle mass (SMI 5.6 kg/m²) and sustained energy expenditure achieved through the 4–8 KKW regimen.

In Case II, body weight remained largely stable throughout the study period, with the exception of an increase exceeding 1 kg during the first week of Phase B2 relative to Phase A. This participant exhibited an SMI of 5.1 kg/m², below the diagnostic threshold for sarcopenia, thereby indicating reduced muscle mass, while a BMI of 26.6 kg/m² classified the individual as obese. According to Heymsfield et al., such a “low muscle mass–high fat percentage body composition” may present with transient body weight reduction during the early stages of body weight reduction, primarily attributable to glycogen depletion and concomitant fluid loss²⁸⁾. Moreover, comparison with Case I suggests that although individuals with greater adiposity may demonstrate more pronounced initial responses, sustaining long-term body weight reduction may be more challenging in the context of low muscle mass. Nonetheless, from week 17 onward, the body weight trended downward, accompanied by a measurable effect size. These findings suggest that while sarcopenia may attenuate body weight reduction outcomes, sustained and carefully regulated monitoring of both exercise dosage and dietary intake can yield meaningful benefits, warranting further investigation.

Similar to Case II, Case III did not exhibit significant body weight reduction in either Phase B1 or Phase B2 when compared with that in Phase A. Nonetheless, Phase B2 demonstrated a consistent effect size, with an average reduction of approximately 400 g. Moreover, in comparison with that in Phase C, significantly lower body weights were observed in both Phase B1 and Phase B2 (p=0.031, not shown in Fig. 3). This outcome may be attributable to the presence of two missing data points out of six in Phase A, likely impeding the acquisition of stable body weight values. As a result, the negative slope of the celeration line steepened, reducing the likelihood of achieving statistically significant results.

Another possible explanation is that the addition of Phase C between Phases B1 and B2 may have influenced the observed pattern of body weight changes throughout the phases, rather than simply reducing the impact of the ergometer intervention. Like in Case II, Case III likely presented with sarcopenia, indicating that extended intervention and sustained monitoring could potentially more substantially reduce body weight. Particularly, among patients with knee OA, activity limitations arising from pain and progressive declines in physical function may modify the effect of interventions. Accordingly, it is essential to consider not only the intensity and dosage of the prescribed intervention but also its continuity and adaptability.

Additionally, because there was no washout period or baseline re-establishment period included between the phases, and a resistance training phase was incorporated, cumulative or residual effects from the previous interventions may have influenced the outcomes. This is especially important for longitudinal measures like body weight and PROMs, which may change gradually over time. Therefore, interpretations specific to each phase should be approached with caution, as the changes observed may be due to physiological or behavioral adaptations across phases rather than just the effects of each intervention in isolation.

In this study, step counts and dietary intake (kcal) were measured as potential confounding variables. Weekly averages, calculated from data collected during the final week of each phase, were used as representative values for analysis. No substantial differences were observed in either the weekly averages or the daily variations. Accordingly, the influence of changes in non-intervention-related activity or dietary intake on the results appears minimal. These findings suggest that body weight changes observed during the intervention period were most likely attributable to the effects of the exercise intervention itself.

The JKOM scores trended to improve across all cases overall. Notably, the magnitude of improvement was greater in Phase B2 (4–8 KKW) compared to that in Phase B1 (1–4 KKW), suggesting that increased exercise dosage may more favorably influence functional outcomes. This observation aligns with the findings of Fransen et al.²⁹⁾, who reported that appropriate progression of exercise loading facilitates physiological adaptation and contributes to improvements in PROMs. Even in cases where changes in step counts were not significant, improvements were observed in PROMs (JKOM). This finding suggests that changes in daily activity levels (step counts) may not necessarily be directly related to improvements in subjective outcomes.

The exercise dosage in this study was defined by standardizing loads through the objective metric of KKW, thereby accommodating variations in participants’ body size. Furthermore, concurrently applying RPE to assess subjective exercise intensity likely contributed to preventing exacerbation of joint symptoms associated with excessive loading. No adverse events were reported during the intervention period. These findings suggest that a stepwise exercise intervention based on KKW represents a highly safe strategy for patients with knee OA. However, exercise dosages exceeding an expenditure of 10 KKW may risk adverse events in Japanese patients with knee OA and should therefore be cautiously implemented.

This case study is subject to several limitations. First, the findings from Study 1 (n=1) and Study 2 (n=3) have limited generalizability, necessitating validation in future research with larger sample sizes. Second, body weight data for Phase A were missing in Cases II and III, thereby constraining comparability and warranting caution in the interpretation of results. Third, the intervention period spanned 20 weeks, with each phase lasting only five weeks, which may be insufficient to thoroughly evaluate the sustainability of outcomes such as body weight reduction and PROMs (JKOM) improvements. Indeed, although a significant reduction in body weight was observed in Case I, no such differences were evident in Cases II and III, suggesting that the brevity of the intervention may have influenced the findings. Longer-term follow-up is therefore required to elucidate how the effects of exercise interventions emerge and persist over time.

This case study investigated the efficacy and safety of a quantified bicycle ergometer intervention in overweight female patients with knee OA. Case I demonstrated that, with a gradual increase in aerobic exercise dosage based on KKW, a transient knee buckling-like incident emerged after exceeding an expenditure of 11 KKW. However, following the dosage adjustment to a 10 KKW expenditure, the intervention was safely sustained. Study 2 employed the ABCB-type single-case design across three participants. A significantly reduced body weight was observed in one case relative to baseline. Furthermore, consistent improvements in PROMs (JKOM) were evident across all participants. These findings suggest that quantitative adjustment of exercise dosage using KKW can serve as a standardized approach to progressively intensifying exercise load while maintaining safety. From a clinical perspective, optimizing therapeutic benefit with the minimal effective dose is essential in terms of safety, cost-effectiveness, and patient adherence. Further interventional studies are warranted to elucidate the dose–response relationship.

Conflict of interest

There are no conflicts of interest to declare regarding this study.