Effect of High-Intensity Exercise Training on Dietary Intake in Patients with Chronic Obstructive Pulmonary Disease: A Prospective Observational Study Authors: Tomoyuki Ikeuchi, Kazuaki Suyama, Miyako Yagita, Shunichiro Mori, Daichi Mori, Tetsuya Kawano, Hideaki Takahata, Terumitsu Sawai, Toru Tsuda, Ryo Kozu Published: 2026 ABSTRACT AbstractPurposeChronic obstructive pulmonary disease (COPD) is frequently complicated by nutritional disorders, and low body weight is an independent prognostic factor. High-intensity exercise training may affect leptin, an appetite-regulating hormone, and dietary intake. However, longitudinal evidence regarding these associations in COPD is limited. This study aimed to examine changes in dietary intake and leptin levels during pulmonary rehabilitation in patients with COPD, with exploratory analyses according to attainment of high-intensity exercise training.Patients and MethodsIn this prospective observational study, patients with COPD participated in a 6-week pulmonary rehabilitation program. High-intensity exercise training (≥ 80% prescribed intensity) was performed whenever feasible. Dietary intake (primary outcome) and leptin were assessed before and after the program. In exploratory subgroup analyses, participants were classified as high-intensity “attainers” (Attainment group) or “nonattainers” (Nonattainment group), and within-group before and after changes as well as between-group differences in change scores were compared. Furthermore, multivariate linear regression analysis was performed to examine whether exercise intensity was independently associated with changes in dietary intake and leptin.ResultsOf the 69 participants, 55 completed the study and were included in the analysis. Dietary intake increased significantly from 1367 to 1484 kcal/day (p = 0.01), whereas leptin did not change significantly in the overall cohort. Between-group comparisons of change scores revealed a significant difference in leptin (p < 0.01) but not in dietary intake (p = 0.16). Multivariable regression analysis showed that exercise intensity attainment was associated with changes in leptin (B = −4.47, p < 0.01) but not dietary intake (p = 0.12).ConclusionDietary intake increased following pulmonary rehabilitation in patients with COPD. Although exploratory analyses suggested an association between attainment of high-intensity exercise training and changes in leptin levels, the relationship with dietary intake remains unclear.Keywords: exercise training, dietary intake, leptin, chronic obstructive pulmonary disease Introduction Chronic obstructive pulmonary disease (COPD) is the third leading global cause of death.1 According to the global initiative for chronic obstructive lung disease (GOLD), 30–60% of hospitalized people with COPD are considered to be malnourished and 50% are below 90% of their ideal body weight.1 Further, both low body weight and weight loss are prognostic factors independent of the severity of airflow limitation, highlighting the importance of strategies to prevent weight loss in this population.2,3 Malnutrition is caused by an energy imbalance. Energy expenditure in COPD is 120 to 140% higher than predicted, due to increased ventilatory costs from the obstructive ventilatory impairment and systemic inflammation.4,5 When compared to healthy individuals, people with COPD have lower dietary intake.6 The decrease in dietary intake in COPD is caused by a breakdown in the interaction between the appetite-suppressing hormone leptin and the appetite-enhancing hormone ghrelin.7,8 Exercise training may promote energy compensation in emaciated individuals and increase dietary intake, but there is large individual variability.9,10 A combination of high-intensity endurance and strength exercise training decreases leptin in people with COPD and furthermore, 6 weeks of such training has been reported to increase dietary intake in those with COPD who are underweight.11,12 However, it remains unclear whether high-intensity exercise training during pulmonary rehabilitation is associated with changes in leptin levels and dietary intake in COPD. The magnitude of energy expenditure during exercise training depends on intensity and duration.13 From a body weight perspective, both exercise training and a high level of physical activity may increase energy expenditure and decrease body weight.14 However, reports in people with COPD have shown that moderate to high-intensity exercise training increases body mass index (BMI) compared to low-intensity exercise training.15 Thus, it is not clear whether exercise training causes weight loss or gain. Our previous randomized controlled trial of pulmonary rehabilitation with different exercise intensities demonstrated that continuing high-intensity exercise, in combination with appropriate nutritional support, improved nutritional status, body composition, energy balance, and dietary intake in patients with COPD.16 However, that study primarily focused on clinical outcomes, and the role of leptin in the association between exercise intensity and dietary intake remained unclear. Thus, the present study differs from our previous trial by focusing specifically on longitudinal changes in dietary intake and circulating leptin levels during pulmonary rehabilitation, and by exploring whether these changes differed according to attainment of high-intensity exercise training. Based on this rationale, we hypothesized that attainment of high-intensity exercise training during pulmonary rehabilitation would be associated with increased dietary intake and reduced circulating leptin. We therefore designed this study to examine changes in dietary intake and leptin levels during a 6-week pulmonary rehabilitation program and to explore their associations with attainment of high-intensity exercise training. Material and Methods Study Design and Participants This was a single-center, prospective observational study. The study sample comprised people with stable COPD who were admitted to Kirigaoka Tsuda Hospital (Fukuoka, Japan) for a 6-week inpatient pulmonary rehabilitation program (PRP) between December 2021 and January 2024. The diagnosis of COPD was made according to GOLD.17 The eligibility criteria were defined as individuals capable of providing written consent, capable of performing exercise training, and capable of oral intake. Exclusion criteria were an acute exacerbation of COPD within the previous 4 weeks, malignancy, cerebrovascular, cardiovascular or gastrointestinal disease, cognitive impairment that would affect the assessment of physical or nutritional status. In addition, individuals taking medications affecting dietary intake (eg, corticosteroids, opioids), diabetes, a BMI ≥ 25 kg/m2, history of gastrectomy, restricted diet due to dyspnea, or inability to take food orally were excluded. Participants who met the inclusion criteria were evaluated at baseline, within 3 days of admission, and immediately following 6-week PRP treatment. Physical activity was assessed continuously for the first 7 days following admission. This study was conducted in accordance with the Declaration of Helsinki. It was approved by the Human Ethics Review Committee of Nagasaki University Graduate School of Biomedical Sciences (approved number 21101404). All participants gave written informed consent. Study Outcomes The primary outcome measure was dietary intake before and after PRP, and the secondary outcome measure was leptin levels before and after PRP. Measurements Age, sex and smoking history were collected from the medical records. Pulmonary Function Pulmonary function was performed using spirometry (CHESTGRAPH Jr. HI-101, CHEST, Tokyo, Japan) in accordance with published guidelines.18 Body Composition We measured height using a stadiometer to the nearest 0.5 cm and body weight in 0.1 kg increments while clothed in order to calculate BMI (kg/m2). Body composition was measured by validated bioelectrical impedance analysis (InBody 270, InBody, Tokyo, Japan), and fat mass index (FMI), fat free mass index (FFMI), and body fat percentage were calculated.19 All participants refrained from eating or drinking for 4 h prior to testing and measurements were made between 11:30 a.m. and 12:00 p.m.20 Physical Function The incremental shuttle walking test (ISWT) was used to measure functional exercise capacity.20 Quadriceps muscle strength (QMS) was measured in the sitting position with the hip and knee flexed at 90 degrees using a hand-held dynamometer with a fixing belt (μ-Tas, ANIMA, Japan). Three measurements were taken, separated by one minute of rest, and the maximum value (kgf) was used in the analysis.21 Dyspnea, Health-Related Quality of Life, Anxiety, and Depression We assessed dyspnea using the modified Medical Research Council dyspnea scale. Health-related quality of life (HRQOL) was assessed using the COPD Assessment Test (CAT).22 Anxiety and depression were assessed with the Hospital Anxiety and Depression Scale (HADS).23 Physical Activity Physical activity was assessed with the activPAL (activPAL3™ micro; PAL Technologies Ltd, Glasgow, UK), which has been validated in people with COPD.24 The activPAL was worn on the right thigh using a waterproof dressing. Physical activity was measured 24 hours a day, including during bathing. Physical activity was monitored for 7 consecutive days beginning immediately after admission. Because PRP started on day 4, the measurement period included 3 days before and 4 days after the start of PRP. Blood Chemistry Blood chemistry testing was performed on admission and 6 weeks later. All participants were fasted from 7:00 p.m. the night before, and samples were collected at 7:00 a.m.25 C-reactive protein (CRP) was collected as an indicator of inflammation, serum albumin as an indicator of nutritional status, and leptin was collected as a marker of appetite. Serum leptin concentrations were centrifuged immediately after sampling and sent to the central laboratory in frozen storage. The concentrations were measured with a dedicated instrument (γ-counter ARC-950/Hitachi, Ltd.) using a reagent for research (HUMAN LEPTIN RIA KIT, EMD Millipore Corporation). Nutritional Assessments Nutritional status was assessed using the Geriatric Nutritional Risk Index (GNRI) calculated from body weight and serum albumin.26 Dietary Intake Dietary intake was calculated by the Weighed Food Records (WFR) using a scale.27 WFR documentation began the day after admission, and comprised the recording of all foods and mixed dishes for the duration of the WFR (quantity of each ingredient, cooking method, total quantity of dish prepared, weight of the pot/pan), quantity of the food consumed by weighing the food and beverages before and after consumption using a digital scale (KS-121 with 0.5-g accuracy and 1000-g capacity, DRETEC, Japan), and time and place of preparation and consumption for all food items. Patients were allowed to go out on weekends, and family members sometimes brought additional food from outside the hospital. For all foods purchased or prepared outside of the hospital, as much detail as possible was recorded, including the main ingredients, method of preparation, other distinguishing features, and quantities. Foods brought in from outside were also weighed, and their nutritional contribution was calculated. The meals brought in were also weighed and calculated.27 Dietary intake was calculated as the mean intake across three weekdays.28 The energy requirement was calculated by Harris-Benedict equation (Total Energy Expenditure [kcal] = Basal Energy Expenditure (BEE) × stress factor (SF) × activity factor (AF)), which is calculated by multiplying the BEE (Male: 66.47 + (13.75 × weight [kg]) + (5.0 × height [cm]) − (6.76 × age); Female: 655.1 + (9.56 × weight [kg]) + (1.85 × height [cm]) − (4.68 × age)) by the SF and the AF.29 The amount of food provided exceeded the energy requirement calculated using this formula. Pulmonary Rehabilitation Program The program consisted of exercise training, self-management, breathing exercises, relaxation, nutritional management, and psycho-social support provided by a multidisciplinary team (physical therapists, nurses, dietitians, medical social workers, and clinical psychologists). All participants performed endurance walking training on a treadmill (SportsArt Fitness, T652 TREADMILL, Japan) for 20 minutes at 80% predicted peak oxygen uptake derived from the distance walked on the ISWT,30 with rests taken when dyspnea was intolerable (ie, Borg rating > 5).31 Furthermore, if continuous walking training was not tolerated because of dyspnea or fatigue, interval training with short rest periods was implemented. Strength training consisted of torso flexion/extension, rowing, leg press, and leg flexion/extension using a training machine (Precor, VitalityTM series, Japan). Strength training consisted of two sets of resistance training with 10–15 repetitions at 80% intensity calculated from the 1 repetition maximum.32 Exercise training occurred five times a week on each weekday under the supervision of a physical therapist. Attendance and attainment of the prescribed exercise intensity were recorded for all sessions throughout the program. Sample Size Calculation The required sample size was calculated using G*Power for a t-test with α = 0.05, and effect size of 0.35 (Cohen’s d),33 and a power of 0.80. Because no published data were available for this calculation, pilot data from 21 participants were used based on changes in dietary intake (mean of difference of 66.1 and standard deviation 188.4). During this pilot phase, all 21 participants increased the intensity of their workouts with the goal of completing high-intensity exercise training; of these, 11 attained high-intensity exercise training and 10 did not. This calculation yielded a required sample size of 55, and 69 participants were required to allow for an anticipated dropout rate of 20%. Statistical Analysis All numerical data are presented as median [interquartile range]. After testing for normality using the Shapiro–Wilk test, the before and after training values were compared using a paired t-test or Wilcoxon signed rank test as appropriate. Effect sizes for within-group comparisons were calculated as Cohen’s d for paired t-tests and as r = |Z|/√N for Wilcoxon signed-rank tests. As part of the subgroup analysis, participants were divided into two groups based on whether they had attained high-intensity training by the end of the program: the Attainment group and the Nonattainment group. The Attainment group consisted of participants who were able to perform both endurance training and strength training at an intensity of 80% or higher and who completed at least 80% of the training sessions. All other participants were classified into the Nonattainment group. Baseline characteristics between the Attainment and Nonattainment groups were compared using the independent t-test or Mann–Whitney U-test for continuous variables and the chi-square test or Fisher’s exact test for categorical variables, as appropriate. We also compared baseline characteristics using the same method between participants who dropped out due to COVID-19 infection and those included in the final analysis. Within-group changes before and after PRP were analyzed using the paired t-test or Wilcoxon signed-rank test. To compare changes between groups, delta values (after – before changes) were calculated for each outcome, and between-group comparisons were performed using the independent t-test or Mann–Whitney U-test, as appropriate. Effect sizes for between-group comparisons were calculated as r = |Z|/√N.33 Multivariable linear regression analyses were performed to examine factors independently associated with changes in dietary intake and leptin. For change in dietary intake, age, sex, GOLD stage, baseline dietary intake, and exercise intensity attainment (0 = Nonattainment, 1 = Attainment) were entered as independent variables. For change in leptin, age, sex, GOLD stage, and exercise intensity attainment were entered as independent variables. GOLD stage was included because previous studies have demonstrated significant differences in energy and protein intake across GOLD stages in patients with COPD.34 Baseline leptin was not included as a covariate as it did not significantly predict change in leptin (p = 0.13) and the correlation between baseline leptin and change in leptin was weak (r = −0.24). Unstandardized regression coefficients (B), standard errors (SE), and 95% confidence intervals (CI) were reported. The subgroup analysis according to attainment of high-intensity exercise training was exploratory and was not prespecified. Some follow-up data were missing, and the number of available observations therefore differed across outcomes. Analyses were performed using available data for each outcome, and missing data were not imputed because of the observational nature of the study. Because subgroup and between-group analyses were exploratory, adjustment for multiple comparisons was not performed; therefore, these findings should be interpreted cautiously. Statistical significance was set at p < 0.05. All analyses were performed using IBM SPSS Statistics version 25.0 (IBM, Chicago, IL, USA). Results Participant Characteristics We assessed 69 people for inclusion in the study. Three were excluded (consent not obtained [n = 2], restricted diet due to dyspnea [n = 1]), and 11 discontinued the program before completion. All discontinuations were due to Coronavirus disease 2019 infections. Therefore, data from 55 participants were included in the analysis, of whom 28 and 27 satisfied the criteria for the Attainment and Nonattainment group respectively (Figure 1). To assess the potential impact of COVID-19-related attrition, baseline characteristics were compared between participants who discontinued because of COVID-19 infection and those included in the final analysis. No significant differences were observed between the groups (Supplementary Materials, Table S1). At baseline, depression was significantly higher in the Nonattainment group. No significant differences were found for any other variables (Table 1). The Attainment group was able to exercise at high-intensity for 99 ± 3% of all sessions. Flow diagram of participants. Table 1Characteristics of ParticipantsAll Participants (n = 55)Attainment Group (n = 28)Nonattainment Group (n = 27)p ValueAge, years78 [73–81]77 [72–81]78 [75–81]0.88Sex, male (%)42 (76.4)22 (78.6)20 (74.1)Smoking status, pack/year56 [38–88]56 [30–98]54 [40–74]0.87Pulmonary function FEV1, L0.90 [0.46–1.31]0.83 [0.53–1.37]0.94 [0.43–1.31]0.49 FEV1, % predicted39 [23–62]33 [24–58]40 [19–73]0.43 FVC, L1.68 [1.30–2.11]1.71 [1.36–2.17]1.66 [1.26–2.11]0.36 FEV1/FVC, %54 [39–69]54 [39–62]56 [39–69]0.53 GOLD stage, I/II/III/IV3/16/15/211/9/8/102/7/7/110.74Body composition BMI, kg/m218.8 [15.7–21.2]19.3 [15.9–21.3]18.3 [15.1–21.1]0.42 FMI, kg/m24.9 [3.0–6.1]4.5 [3.2–6.1]5.1 [2.7–6.0]0.87 FFMI, kg/m213.7 [13.0–15.0]13.7 [13.0–15.4]13.6 [12.8–14.3]0.29 Body fat, %25.0 [17.1–30.3]24.9 [16.7–31.1]26.4 [17.5–30.2]0.85Physical function ISWT, m135 [80–190]155 [93–193]110 [80–180]0.15 QMS/Body weight, %40 [33–48]43 [34–49]38 [30–46]0.18Dyspnea, HRQOL, anxiety, and depression mMRC dyspnea scale, 0/1/2/3/40/1/12/22/200/0/7/12/90/1/5/10/110.20 CAT21 [16–27]21 [16–25]21 [16–28]0.21 HADS anxiety6 [3–10]6 [2–9]6 [4–10]0.18 HADS depression9 [6–12]8 [5–10]10 [8–12]0.02Physical activity Steps per day2646 [1308–4547]2328 [1202–5579]3354 [1317–4486]0.53 Activity score, MET.h31 [31–32]31 [31–33]31 [31–32]0.47Blood chemistry Leptin, ng/mL6.5 [4.7–11.5]6.5 [4.5–12.7]5.3 [4.7–10.5]0.67 Albumin, g/dL3.8 [3.5–4.0]3.9 [3.6–4.0]3.8 [3.4–4.1]0.81 CRP, mg/dL0.13 [0.06–0.60]0.12 [0.07–0.60]0.14 [0.05–0.74]0.63Nutritional assessments and dietary intake GNRI92 [85–99]94 [87–98]89 [83–100]0.92 Dietary intake, kcal1367 [1059–1689]1456 [1037–1713]1326 [1080–1662]0.61 Protein, g61 [44–74]64 [42–75]56 [46–74]0.91 Fat, g38 [29–46]41 [28–46]37 [29–47]0.97 Carbohydrates, g221 [181–247]228 [178–251]211 [189–245]0.51Notes: Data are presented as medians and interquartile range or n (%). p value, indicates differences between Attainment and Nonattainment group.Abbreviations: BMI, body mass index; CAT, COPD assessment test; CRP, C-reactive protein; FEV1, forced expiratory volume in 1 s; FFMI, fat-free mass index; FMI, fat mass index; FVC, forced vital capacity; GNRI, geriatric nutritional risk index; GOLD, global initiative for chronic obstructive lung disease; HADS, hospital anxiety and depression scale; HRQOL, health related quality of life; ISWT, incremental shuttle walking test; MET, metabolic equivalents; mMRC, modified medical research council dyspnea scale; QMS, quadriceps muscle strength. Characteristics of Participants Notes: Data are presented as medians and interquartile range or n (%). p value, indicates differences between Attainment and Nonattainment group. Abbreviations: BMI, body mass index; CAT, COPD assessment test; CRP, C-reactive protein; FEV1, forced expiratory volume in 1 s; FFMI, fat-free mass index; FMI, fat mass index; FVC, forced vital capacity; GNRI, geriatric nutritional risk index; GOLD, global initiative for chronic obstructive lung disease; HADS, hospital anxiety and depression scale; HRQOL, health related quality of life; ISWT, incremental shuttle walking test; MET, metabolic equivalents; mMRC, modified medical research council dyspnea scale; QMS, quadriceps muscle strength. Changes in Primary Outcomes Following Pulmonary Rehabilitation Dietary intake increased by a median of 117 kcal (effect size r = 0.43, p = 0.01), while there was no change in leptin levels (p = 0.46). Body composition showed significant increases in BMI and FFMI, along with a slight increase in body fat percentage. Physical function improved, as demonstrated by increases in the ISWT and QMS. Furthermore, CAT scores and both anxiety and depression scores decreased significantly (Table 2). Table 2Comparison of Data for All Participants Before and After Pulmonary Rehabilitation (n = 55)nBeforeAfterEffect Sizep ValueBody composition BMI, kg/m25518.8 [15.7–21.2]19.8 [16.9–20.9]0.280.04 FMI, kg/m2454.9 [3.0–6.1]5.1 [3.2–6.2]0.160.24 FFMI, kg/m24513.7 [13.0–15.0]14.2 [13.2–15.3]0.50< 0.01 Body fat, %4525.0 [17.1–30.3]26.0 [19.3–30.8]0.270.05Physical function ISWT, m39135 [80–190]180 [120–250]0.53< 0.01 QMS/Body weight, %4340 [33–48]44 [37–53]0.51< 0.01HRQOL, anxiety, and depression CAT4621 [16–27]19 [13–24]0.40< 0.01 HADS anxiety466 [3–10]6 [2–8]0.300.03 HADS depression469 [6–12]8 [5–11]0.270.05Blood chemistry Leptin, ng/mL556.5 [4.7–11.5]7.3 [4.9–13.1]0.100.46 Albumin, g/dL533.8 [3.5–4.0]3.7 [3.4–4.0]0.170.21 CRP, mg/dL530.13 [0.06–0.60]0.11 [0.06–0.66]0.080.55Nutritional assessments GNRI5392 [85–99]93 [86–101]0.020.87 Dietary intake, kcal551367 [1059–1689]1484 [1258–1736]0.430.01 Protein, g5560.7 [44.0–73.7]65.7 [56.9–75.5]0.410.02 Fat, g5537.6 [28.5–46.3]42.5 [34.8–50.1]0.240.08 Carbohydrates, g55220.5 [180.8–246.8]228.5 [196.6–250.0]0.280.04Notes: Data are medians and interquartile range.Abbreviations: BMI, body mass index; CAT, COPD assessment test; CRP, C-reactive protein; FFMI, fat-free mass index; FMI, fat mass index; GNRI, geriatric nutritional risk index; HADS, hospital anxiety and depression scale; HRQOL, health related quality of life; ISWT, incremental shuttle walking test; QMS, quadriceps muscle strength. Comparison of Data for All Participants Before and After Pulmonary Rehabilitation (n = 55) Notes: Data are medians and interquartile range. Abbreviations: BMI, body mass index; CAT, COPD assessment test; CRP, C-reactive protein; FFMI, fat-free mass index; FMI, fat mass index; GNRI, geriatric nutritional risk index; HADS, hospital anxiety and depression scale; HRQOL, health related quality of life; ISWT, incremental shuttle walking test; QMS, quadriceps muscle strength. Within and Between-Group Changes by Exercise Intensity Attainment Within-group before and after changes following the PRP in the Attainment and Nonattainment groups are summarized in Table 3. Dietary intake increased only in the Attainment group. Leptin decreased in the Attainment group but increased in the Nonattainment group. Regarding body composition, FFMI increased and body fat percentage decreased only in the Attainment group. Table 3Comparison of Before and After Pulmonary Rehabilitation in the Attainment and Nonattainment Groups Within GroupBetween GroupAttainment Group (n = 28)Nonattainment Group (n = 27)nBeforeAfterChangeaEffect Sizep ValuebnBeforeAfterChangeaEffect Sizep Valuebp ValuecrBody composition BMI, kg/m22819.3 [15.9–21.3]20.0 [16.0–21.0]0.1 [−0.1–0.5]0.250.062718.3 [15.1–21.1]19.8 [17.4–20.8]0.4 [−0.4–0.7]0.140.300.940.01 FMI, kg/m2264.5 [3.2–6.1]4.9 [2.8–6.2]−0.1 [−0.5–0.1]0.230.08195.1 [2.7–6.0]5.3 [4.4–6.3]0.1 [−0.4–0.4]0.020.900.390.13 FFMI, kg/m22613.7 [13.0–15.4]14.4 [13.1–15.6]0.3 [0.1–0.7]0.46< 0.011913.6 [12.8–14.3]14.0 [13.2–14.7]0.4 [0.0–0.9]0.240.070.750.05 Body fat percentage, %2624.9 [16.7–31.1]23.4 [16.8–31.3]−1.0 [−2.5–0.4]0.310.021926.4 [17.5–30.2]28.1 [22.4–30.1]0.1 [−2.2–1.7]0.050.720.250.17Physical function ISWT, m25155 [93–193]190 [130–255]30 [10–60]0.48< 0.0114110 [80–180]180 [78–185]15 [0–55]0.240.080.460.12 QMS/Body weight, %2643 [34–48]43 [38–51]3 [0–6]0.320.021738 [30–46]47 [32–57]7 [2–9]0.41<0.010.120.24HRQOL, anxiety, and depression CAT2721 [16–25]17 [11–22]−3 [−5–1]0.39< 0.011921 [16–28]22 [13–26]−1 [−3–1]0.190.280.310.15 HADS anxiety276 [2–9]5 [2–6]−1 [−3–0]0.350.01196 [4–10]7 [4–11]−1 [−2–2]0.040.780.330.14 HADS depression278 [5–10]7 [4–9]−1 [−3–1]0.210.111910 [8–12]9 [5–12]−2 [−4–−1]0.170.220.470.11Blood chemistry Leptin, ng/mL286.5 [4.5–12.7]6.3 [4.5–10.3]−0.8 [−1.5–0.7]0.320.03275.3 [4.7–10.5]10.1 [6.1–17.1]2.2 [0.7–6.5]0.53< 0.01< 0.010.45 Albumin, g/dL283.9 [3.6–4]3.8 [3.6–4]0.0 [−0.3–0.2]0.100.45253.8 [3.4–4.1]3.5 [3.2–3.9]0.0 [−0.3–0.1]0.150.280.710.05 CRP, mg/dL280.12 [0.07–0.60]0.10 [0.05–0.29]0.01 [−0.13–0.10]0.080.56250.14 [0.05–0.74]0.18 [0.07–0.9]−0.01 [−0.10–0.13]0.030.850.910.02Nutritional assessment GNRI2894 [87–98]93 [86–101]0 [−3–4]0.040.762589 [83–100]93 [85–101]−1 [−9–4]0.090.490.450.10 Dietary intake, kcal281456 [1037–1713]1576 [1410–1803]57.5 [−11.1–341]0.550.04271326 [1080–1662]1457 [1202–1705]32 [−40–180]0.310.110.160.09 Protein, g2864 [42–75]71 [60–76]2 [0–15]0.41< 0.012756 [46–74]63 [52–74]0 [−10–8]0.150.270.100.22 Fat, g2841 [28–46]45 [36–50]1 [−2–9]0.240.082737 [29–47]40 [33–50]0 [−7–7]0.080.560.030.14 Carbohydrates, g28228 [178–251]240 [207–259]6 [−9–24]0.240.0827211 [189–245]221 [170–246]5 [−44–16]0.160.230.880.02Notes: Data are presented as medians and interquartile range. a difference between before and after; b within-group comparison p value (Wilcoxon signed-rank test); c difference in change between Attainment and Nonattainment groups (Mann–Whitney U-test).Abbreviations: BMI, body mass index; CAT, COPD assessment test; CRP, C-reactive protein; FFMI, fat-free mass index; FMI, fat mass index; GNRI, geriatric nutritional risk index; HADS, hospital anxiety and depression scale; HRQOL, health related quality of life; ISWT, incremental shuttle walking test; QMS, quadriceps muscle strength. Comparison of Before and After Pulmonary Rehabilitation in the Attainment and Nonattainment Groups Notes: Data are presented as medians and interquartile range. a difference between before and after; b within-group comparison p value (Wilcoxon signed-rank test); c difference in change between Attainment and Nonattainment groups (Mann–Whitney U-test). Abbreviations: BMI, body mass index; CAT, COPD assessment test; CRP, C-reactive protein; FFMI, fat-free mass index; FMI, fat mass index; GNRI, geriatric nutritional risk index; HADS, hospital anxiety and depression scale; HRQOL, health related quality of life; ISWT, incremental shuttle walking test; QMS, quadriceps muscle strength. For physical function, ISWT improved significantly only in the Attainment group, whereas QMS increased significantly in both groups. In addition, CAT and HADS-Anxiety scores decreased only in the Attainment group. No significant before and after changes were observed in albumin, CRP, or GNRI in either group (Table 3). Between-group comparisons of change scores are presented in Table 3. Significant between-group differences were observed for changes in leptin levels and fat intake. Compared with the Nonattainment group, the Attainment group demonstrated a greater reduction in leptin levels and a greater increase in fat intake. In contrast, no statistically significant between-group differences were observed in changes in dietary intake. No significant differences were observed in the other variables. Multivariable Analyses When changes in dietary intake were used as the dependent variable, exercise intensity attainment was not independently associated with changes in dietary intake after adjustment for covariates (Table 4). In contrast, when changes in leptin were used as the dependent variable, attainment of high-intensity exercise training remained independently associated with changes in leptin (Table 5). Table 4Multiple Regression Analysis for Change in Dietary IntakeModel 1Model 2BSE95% CIp ValueBSE95% CIp ValueAge, years−8.175.78−19.77 to 3.430.16−7.295.72−18.78 to 4.200.21Sex268.2480.92105.71 to 430.77< 0.01261.9979.83101.57 to 422.41< 0.01GOLD stage−40.9134.13−109.46 to 27.630.24−40.5733.63−108.14 to 27.010.23Dietary intake, baseline (kcal)−0.470.09−0.65 to −0.29< 0.01−0.470.09−0.64 to −0.29< 0.01Exercise intensity (Attainment)————102.2664.61−27.58 to 232.090.12R2 = 0.377, Adjusted R2 = 0.327, F = 7.551, p < 0.01R2 = 0.407, Adjusted R2 = 0.346, F = 6.724, p < 0.01Notes: Model 1: adjusted for age, sex, GOLD stage, and baseline dietary intake. Model 2: Model 1 covariates plus exercise intensity attainment (0 = Nonattainment, 1 = Attainment). n = 55 for both models.Abbreviations: B, unstandardized regression coefficient; SE, standard error; CI, confidence interval; GOLD, global initiative for chronic obstructive lung disease. Multiple Regression Analysis for Change in Dietary Intake Notes: Model 1: adjusted for age, sex, GOLD stage, and baseline dietary intake. Model 2: Model 1 covariates plus exercise intensity attainment (0 = Nonattainment, 1 = Attainment). n = 55 for both models. Abbreviations: B, unstandardized regression coefficient; SE, standard error; CI, confidence interval; GOLD, global initiative for chronic obstructive lung disease. Table 5Multiple Regression Analysis for Change in LeptinModel 1Model 2BSE95% CIp ValueBSE95% CIp ValueAge, years−0.100.13−0.36 to 0.160.46−0.130.12−0.37 to 0.110.27Sex2.341.77−1.21 to 5.890.192.591.63−0.68 to 5.860.12GOLD stage0.190.76−1.34 to 1.710.810.180.70−1.22 to 1.580.80Exercise intensity (Attainment)————−4.471.38−7.25 to −1.69< 0.01R2 = 0.046, Adjusted R2 = −0.010, F = 0.818, p = 0.49R2 = 0.211, Adjusted R2 = 0.147, F = 3.333, p = 0.02Notes: Model 1: adjusted for age, sex, and GOLD stage. Model 2: Model 1 covariates plus exercise intensity attainment (0 = Nonattainment, 1 = Attainment). Baseline leptin was not included as a covariate as it did not significantly predict Δleptin (p =0.134) and the correlation between baseline leptin and Δleptin was weak (r = −0.241), indicating minimal regression-to-the-mean effect. n = 55 for both models.Abbreviations: B, unstandardized regression coefficient; SE, standard error; CI, confidence interval; GOLD, global initiative for chronic obstructive lung disease. Multiple Regression Analysis for Change in Leptin Notes: Model 1: adjusted for age, sex, and GOLD stage. Model 2: Model 1 covariates plus exercise intensity attainment (0 = Nonattainment, 1 = Attainment). Baseline leptin was not included as a covariate as it did not significantly predict Δleptin (p =0.134) and the correlation between baseline leptin and Δleptin was weak (r = −0.241), indicating minimal regression-to-the-mean effect. n = 55 for both models. Abbreviations: B, unstandardized regression coefficient; SE, standard error; CI, confidence interval; GOLD, global initiative for chronic obstructive lung disease. Discussion The present study demonstrated that dietary intake increased after a 6-week PRP in COPD. In addition, exploratory analyses suggested that attainment of high-intensity exercise training was associated with changes in leptin levels. This is a valuable longitudinal report investigating changes in dietary intake and leptin levels during pulmonary rehabilitation in COPD. Consistent with an effective PRP, exercise capacity, muscle strength, symptoms, and psychological outcomes improved, with changes in ISWT, QMS, and CAT approaching established minimal clinically important differences.35–37 However, caution is required when interpreting the results of this study due to missing values. The increase in dietary intake observed across the overall cohort in this study may reflect behavioral and/or physiological changes accompanying PRP, because no nutritional supplements were provided during the intervention period. The participants were relatively lean (median BMI 18.8 kg/m2), and exercise training may have facilitated compensatory increases in intake in this underweight-lean phenotype. In the overall cohort, leptin did not change significantly; however, subgroup analyses revealed a decrease in leptin in the Attainment group and an increase in the Nonattainment group. Prior studies have reported reductions in leptin after high-intensity exercise training in healthy individuals and in COPD, and our findings in the Attainment group are consistent with these observations.11,38 Given the reported positive association between leptin and body fat percentage,8 the lower body fat percentage observed in the Attainment group may have contributed to the reduction in leptin. In addition, the present multivariable analysis demonstrated that attainment of high-intensity exercise training remained independently associated with changes in leptin after adjustment for clinically relevant covariates. These findings suggest that attainment of higher exercise intensity during PRP was associated with leptin responses in patients with COPD; however, residual confounding cannot be excluded. Although differences in leptin response were observed depending on the level of high-intensity exercise training, no significant differences in dietary intake were found between the groups. These findings suggest that the relationship among exercise intensity, leptin regulation, and dietary intake may be more complex than initially assumed. Although leptin is a key hormone involved in appetite regulation, dietary intake in COPD is influenced by multiple factors, including dyspnea, fatigue, depression, systemic inflammation, social environment, and pharmacotherapy.39 Therefore, during a short-term rehabilitation period, a decrease in leptin levels alone does not necessarily lead directly to an increase in dietary intake. In addition, appetite itself was not directly assessed using measures such as visual analog scales or appetite questionnaires. Future studies incorporating direct appetite assessment, additional hormonal measurements, and mediation analyses are needed to better clarify the relationships among exercise intensity, leptin responses, and dietary intake in COPD. Another important point is that the subgroup analysis in this study was exploratory. Participants were classified based on their level of adherence to high-intensity exercise training during the PRP; however, this classification was not randomized. Therefore, differences between groups may reflect baseline differences in physical ability, motivation, psychological state, symptom burden, or adherence rather than the sole effect of exercise intensity itself. In fact, the Nonattainment group had higher baseline depression scores compared to the Attainment group, which may have influenced both exercise participation and dietary behavior. To address this concern, we conducted additional multivariate analyses adjusting for clinically relevant variables; however, the possibility of confounding factors cannot be ruled out. Furthermore, since approximately half of the participants were unable to perform high-intensity exercise, this highlighted the need for personalized exercise programs and support measures. Previous rehabilitation studies have similarly reported substantial heterogeneity in patient responses and emphasized that rehabilitation outcomes may be influenced by baseline clinical characteristics and adherence to training programs.40 This study has several limitations. First, this was a single-center observational study with a relatively small sample size, which may limit statistical power and generalizability. Second, the cohort consisted exclusively of Japanese patients, who may have lower body weight than Western populations;41 therefore, external validity to other ethnicities and settings should be interpreted with caution. Third, analyses were conducted using available data for each outcome, and the number of analyzed participants therefore differed across variables. This approach was chosen due to the nature of observational study, but the possibility of bias associated with missing data cannot be ruled out. Fourth, COVID-19-related dropout occurred during the study period and may have introduced additional selection bias. Fifth, physical activity was assessed over the first 7 days of hospitalization, including both the pre-rehabilitation period and the initial phase of PRP. Therefore, the measured activity level does not represent either a true baseline value or an exposure-adjusted estimate of activity throughout the rehabilitation program. However, because physical activity was collected for descriptive purposes only and was not incorporated into the primary analyses, this limitation is unlikely to have substantially affected the main findings. Finally, this study was not specifically designed for comparisons between subgroups, nor did it possess the necessary statistical power for that purpose. Although multivariable analyses were performed, the number of covariates that could be included was limited by the sample size. Therefore, potential confounding by psychological status, baseline physical function, exercise tolerance, motivation, and adherence capacity may remain. Accordingly, the exploratory findings regarding exercise intensity attainment should be interpreted cautiously, as residual confounding factors and baseline imbalances may have influenced the results. Conclusion Dietary intake increased following a 6-week pulmonary rehabilitation program. Exploratory analyses suggested that attainment of high-intensity exercise training was associated with changes in leptin levels; however, no clear association with dietary intake was observed. Further studies are needed to clarify the relationships among exercise intensity, leptin responses, and dietary intake in COPD. Acknowledgments The authors would like to express our gratitude to all the subjects who participated in this study, as well as to the staff of the Department of Respiratory Medicine and Rehabilitation at Kirigaoka Tsuda Hospital. We greatly appreciate Dr Sue Jenkins for her help with reviewing the manuscript. Funding Statement The study design, collection and analysis of the data, and preparation of the manuscript were not funded by any public, commercial, or not-for-profit sectors. Abbreviations BMI, body mass index; CAT, COPD Assessment Test; COPD, chronic obstructive pulmonary disease; CRP, C-reactive protein; FFMI, fat-free mass index; FMI, fat mass index; GNRI, Geriatric Nutritional Risk Index; GOLD, Global Initiative for Chronic Obstructive Lung Disease; HADS, Hospital Anxiety and Depression Scale; HRQOL, health related quality of life; ISWT, incremental shuttle walking test; mMRC, modified Medical Research Council dyspnea scale; QMS, quadriceps muscle strength. Data Sharing Statement The datasets in this study are available from the corresponding author upon reasonable request. Author Contributions All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work. Disclosure The authors report no conflicts of interest in this work. References Associated Data Data Availability Statement The datasets in this study are available from the corresponding author upon reasonable request.