Effects of a combined exercise and dietary intervention on body composition, physical functioning and fatigue in patients with ovarian cancer: results of the PADOVA trial

Abstract

Background

Guidelines recommend to include exercise and dietary advice in standard care for patients with cancer, based on evidence primarily derived from patients with breast cancer. Its applicability to patients with ovarian cancer is uncertain due to differences in patient characteristics and treatments. The PADOVA trial examined the effectiveness of a combined exercise and dietary intervention on fat-free mass (FFM), physical functioning, and fatigue.

Methods

In total, 81 patients with ovarian cancer were randomised to the exercise and dietary intervention (n = 40) or control (n = 41) group. Measurements were performed before chemotherapy, after chemotherapy, and 12 weeks later. FFM was assessed by bioelectrical impedance analysis, and physical functioning and fatigue were assessed using questionnaires. Intervention effects were assessed on an intention-to-treat basis using linear mixed models.

Results

FFM and physical functioning increased, and fatigue decreased significantly over time in both groups. No significant difference between the groups were found for FFM (β = −0.5 kg; 95% CI = −3.2; 2.1), physical functioning (β = 1.4; 95% CI = −5.4; 8.3) and fatigue (β = 0.7; 95% CI = −1.5; 2.8).

Conclusions

During treatment, both groups improved in FFM, physical functioning, and fatigue. The intervention group, however, did not demonstrate additional benefits compared to the control group. This highlights the need for caution when extrapolating findings from different cancer populations to patients with ovarian cancer.

Introduction

Ovarian cancer is the seventh most common cancer in women [1]. Because the disease is usually detected at an advanced stage, ovarian cancer has a low overall 5-year survival rate of 30–40% [2]. Standard treatment of ovarian cancer includes cytoreductive surgery with (neo)adjuvant chemotherapy consisting of platinum and taxane, administered for 6 three-week cycles [3]. This can lead to numerous physical and psychosocial problems, including reduced physical functioning, fatigue, sarcopenia and malnutrition, which can significantly compromise patients’ health-related quality of life (HRQoL) [4, 5].

In general, it has been shown that exercise and dietary interventions have significant benefits on body composition, fatigue, physical fitness, and HRQoL in patients with cancer [6–14]. This has resulted in international guidelines recommending exercise and dietary support in cancer care [15–17]. However, much of the current knowledge regarding the effects of exercise and dietary interventions on health outcomes is based on studies conducted among patients with breast cancer receiving curative treatment [10]. Thus, it remains unclear whether findings from studies among patients with other types of cancer are applicable to patients with ovarian cancer. Patients with ovarian cancer have different patient characteristics, treatment types and trajectories, and are usually diagnosed at an advanced stage [2], setting them apart from patients with other types of cancer. Some pilot studies have demonstrated the feasibility of moderate-intensity exercise interventions or combined exercise and dietary interventions during chemotherapy in patients with ovarian cancer [18–21]. However, a well-designed randomised controlled trial investigating the effects of these interventions has not yet been conducted. The importance of such a trial is highlighted by the low physical activity levels and suboptimal diet reported in patients with ovarian cancer [22], which can negatively impact physical functioning HRQOL [22–24].

The aim of this study was to compare the effectiveness of an intervention that included both exercise, as well as dietary support, with usual care for patients with ovarian cancer receiving (neo)adjuvant chemotherapy. The primary outcomes were body composition, physical functioning and fatigue. Secondary outcomes included physical fitness, HRQoL, anxiety and depression, symptoms of neuropathy, sleep disturbances, physical activity and dietary intake.

Methods

Design and participants

The Physical Activity and Dietary Intervention in OVArian cancer (PADOVA) trial [25] was a two-arm multicenter randomised controlled trial in which patients were randomised to a combined exercise and dietary intervention in addition to usual care or a usual care control group. Between 2018 and 2022, patients were recruited from three gynaecological cancer centres and their affiliated peripheral hospitals in The Netherlands. After diagnosis and before the start of neoadjuvant chemotherapy or adjuvant chemotherapy, the gynaecological oncologist informed patients about the PADOVA study. Patients aged >18 years with primary epithelial ovarian cancer who were scheduled for (neo)-adjuvant chemotherapy treatment were invited to participate. Patients were excluded from this study if they had a prior cancer diagnosis up to 5 years ago, were not able to perform basic activities of daily living, had a contraindication for exercise (e.g., heart failure), had a cognitive disorder or severe emotional instability (e.g., schizophrenia, Alzheimer), were unable to read/or write Dutch, or had a life expectancy of less than 3 months. Patients who did not wish to participate in the trial were invited for a one-time questionnaire. Written informed consent was obtained from all patients prior to participation. Measurements were performed at baseline before the first or second chemotherapy cycle and before randomisation (T0), after the last cycle of chemotherapy (T1) and at 12 weeks after T1 (T2). The intervention started at the first 3-week cycle of chemotherapy and continued until the last chemotherapy cycle (between T0 and T1). The study was approved by the Medical Ethics Committee of the Amsterdam UMC (reference: 018) and registered in the Netherlands Trial Register (NTR6300). Details of the study protocol are described elsewhere [25].

Randomisation and blinding

After baseline measurement, participants were stratified by FIGO stage (low (I/II) vs high (III/IV) stage) and treatment regimen (primary surgery followed by adjuvant chemotherapy vs neoadjuvant chemotherapy followed by interval debulking and adjuvant chemotherapy), after which we applied a random allocation to the intervention or the control group within each of the strata. The randomisation was conducted by an independent researcher using a table of randomly generated numbers arranged in blocks of four. The random numbers were provided by an independent statistician. The allocation sequence was concealed from the research and clinical staff. After randomisation, patients in both study arms received a brochure with general information on physical activity, diet and body weight recommendations for cancer survivors [26].

Intervention and usual care

Exercise and dietary intervention

Overall, the intervention encompassed a comprehensive approach that addressed both exercise and dietary intake throughout the duration of chemotherapy treatment (±18 weeks). Details of the intervention have been described elsewhere [25]. In short, the exercise programme included two 1-h sessions per week, including moderate- to high-intensity resistance and aerobic exercise, which were supervised by a physical therapist specialised in oncology in a physical therapy practice close to the patients’ home. The resistance exercise programme targeting large muscle groups included six exercises, with two sets of ten repetitions, at 70–80% of the one-repetition maximum (1RM) estimated with an indirect 1RM measurement. Aerobic exercises were performed for 30 min per session, with an intensity of 50–80% (gradually increasing over the training period) of the maximal workload as estimated by the steep ramp test [27]. The 1RM and steep ramp tests were repeated every 3 and 6 weeks, respectively, to ensure an adequate training load. During training, training load could be adjusted based on the Borg Scale of the rate of perceived exertion [28]. When the Borg scale exceeded 15, the load was decreased by one step and when the Borg scale decreased to <12, the load was increased. In addition, participants were advised to be physically active at moderate intensity for at least 30 min, three times per week, complementary to the supervised exercise programme [29].

Dietary counselling was delivered by dietitians specialised in oncology once every 3 weeks during 30–45 min face-to-face sessions in the hospital or by telephone using motivational interviewing techniques [30]. Counselling was tailored to the nutritional needs of each individual patient, taking into account factors like body composition determined through body weight and BIA measurements, nutritional status, and dietary intake assessed through 24-h recalls carried out by the dietitian during the counselling sessions. The primary focus of the counselling was to prevent weight loss by ensuring an adequate intake. In participants where weight loss was not a concern, the focus shifted to align with the dietary guidelines set by the World Cancer Research Fund (WCRF)/American Institute for Cancer Research (AICR) [29]. In addition, all participants were advised to consume at least 1.2 g of protein per kilogram body weight per day [17] and at least 25 g of protein per meal, since regular protein intake is expected to optimise muscle protein synthesis [31]. Subsequent sessions included feedback on body weight, body composition, diet quality and achievement of protein targets, creating a comprehensive and personalised approach to patient nutrition.

To promote health behaviours, physical therapists and dietitians applied behavioural change techniques which were based on Bandura’s Social Cognitive Theory [32] and primarily focused on improving self-efficacy, addressing sociostructural factors (barriers and facilitators), managing outcome expectations and setting goals to improve health behaviours [32]. Furthermore, this study was adequately tailored to this patient-specific group by adjusting the intervention to comorbidities, and disease- and treatment-induced effects, as thoroughly described elsewhere [33].

Usual care control group

Women in the control group received usual care during chemotherapy, which could include a referral to a dietitian when malnutrition was detected by the gynaecological oncologist. To prevent non-participation and dropout, patients in the usual care control group were offered a maximum of three exercises and three dietary counselling sessions in the period between T1 and T2.

Outcome measurements

Primary outcomes

Details of all outcome measures are described elsewhere [25]. In short, primary outcomes were body composition, physical functioning and fatigue, which were assessed at all time points. Physical tests were performed by trained assessors who were blinded to group allocation, and participants were instructed not to reveal their group allocation. Body weight and height were measured by trained researchers and body mass index (BMI; kg/m²) was calculated. Fat mass and fat-free mass in kg (FFM) were assessed with a non-invasive bioelectrical impedance assessment (BIA; Bodystat 1500) [34]. FFM was calculated using the sex-specific formula of Steiner et al. [35]. While our initial plan involved assessing body composition through routine CT imaging conducted for diagnostic purposes, logistic constraints resulted in only n = 12 CT scans in the intervention group and n = 15 CT scans in the control group. As a pragmatic alternative, we opted for the more readily accessible BIA method. However, this shift to BIA may have resulted in insufficient statistical power to detect differences in body composition between groups.

Physical functioning was assessed using the physical functioning subscale of the validated European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core 30 (EORTC QLQ-C30) [36]. The physical subscale comprises five items with four response options: “Not at All”, “A Little”, “Quite a Bit”, and “Very Much”. The raw EORTC physical functioning score was converted into a continuous score (0–100), with a higher score representing better levels of physical functioning.

General fatigue was assessed with the general fatigue subscale of the Multidimensional Fatigue Inventory (MFI) questionnaire [37]. The MFI is a validated questionnaire and consists of 20 items divided into five subscales: general fatigue, physical fatigue, reduced physical activity, reduced motivation and mental fatigue. Participants were asked to indicate, on a 1–5 scale, to what extent a particular item applied to them, with a maximum sum score of 20 points per subscale. A higher score indicated more fatigue.

Secondary outcome measures

Secondary outcomes were also assessed at all time points. Cardiorespiratory fitness was measured during a maximal exercise test on an electronically braked cycle ergometer using a ramp protocol, aiming to achieve peak oxygen uptake (peakVO₂, in mL/kg/min) within 8–12 min [38]. Muscle strength for elbow flexion and knee extension was measured with the break method using a hand-held dynamometer (MicroFET 2: Hoggan Health Industries Inc., Draper, USA) [39, 40].

Patient-reported outcomes focused on HRQoL (EORTC QLQ-C30 [36, 41]), anxiety and depression (Hospital Anxiety and Depression Scale [42]), symptoms of neuropathy (EORTC CIPN-20 [43]), and sleep disturbances (Pittsburgh Sleep Quality Index [44]). Physical activity was assessed using the Physical Activity Scale for the Elderly [45] questionnaire. Dietary intake was assessed by two different questionnaires; a brief food frequency questionnaire [46] to assess adherence to the WCRF/AICR lifestyle recommendations and Dutch guidelines for a healthy diet [29, 47] and a clinical questionnaire developed by dietitians to assess total energy and protein intake per day. The total intake per day was adjusted to correct for under- and over-reporting using cut-off values for minimum and maximum daily calorie intake (500–3500 kcal/day).

Assessments of covariables, session attendance, adverse events and contamination

Sociodemographic data were collected by a self-report questionnaire. Clinical information (e.g., cancer subtype, FIGO stage) was obtained from medical records. The comorbidity level was evaluated using the Charlson Comorbidity Index [48]. An elevated comorbidity level was defined as a score of ≥1 after subtracting the primary tumour, which is one component of the CCI.

Physical therapists and dietitians reported session attendance and adherence in detailed logs. Subsequently, the researchers utilised these logs to calculate attendance rates. For exercise only, the relative dose intensity was also examined by using the ratio between the total cumulative dose of completed exercises and the originally planned cumulative dose [49, 50]. Contamination was assessed by asking participants from the usual care control group at the T1 measurement if and how often they had attended supervised exercise and/or dietary counselling outside the study.

Power calculations

The sample size was initially based on the results of a previous randomised controlled trial among patients with breast cancer [25, 51]. However, a recalculation was executed based on the standard deviations of the physical functioning and fatigue scale from the first 57 participants at baseline. With 34 patients per study arm, we would be able to detect a clinically relevant between-group difference in effects directly after intervention on physical functioning (10 points; SD: 12.6), and physical fatigue (2.7 points; SD: 4.0; alpha = 0.05, power = 0.80) [19, 52–56] at T1. Taking into account a dropout of 15% [6, 51, 57], we aimed to include 40 patients per group.

Statistical analysis

Longitudinal analyses were performed using linear mixed models allowing for evaluation of changes in the primary outcomes (body composition, physical functioning, and general fatigue) and secondary outcomes (physical fitness, HRQoL, anxiety and depression, symptoms of neuropathy, sleep disturbances, physical activity, and dietary intake) over time. Separate models were built for each outcome. This model also allowed to evaluate within person and between-group differences in effect. Notably, the main intervention effect was distinctly observable at T1. Due to the design, at T2, only the potential sustainability of intervention effects can be observed. All analyses were performed based on intention-to-treat and adjusted for all balancing variables (FIGO stage and treatment regimen). In addition, potential effect modification by age, treatment regimen (neoadjuvant vs adjuvant chemotherapy) and baseline values of corresponding outcome measures were explored by adding the variable and its interaction term with the intervention into the regression model. In cases where the model fit showed significant improvement with the addition of these interaction terms, tested with the likelihood ratio test, stratified analyses were subsequently performed. To determine whether missing data was selective, univariable logistic regression analyses were conducted to examine differences in baseline demographic and clinical characteristics between patients who completed post-test assessments and those who did not. Lastly, we performed per-protocol analyses for the primary outcomes among adherent participants (i.e., ≥75% attendance of total prescribed exercise sessions and dietary consultations) [51] and usual care without contamination (i.e., no exercise training under the supervision of a physical therapist but dietary consultation was not ruled out, as it could be incorporated as part of clinical care). We considered P < 0.05 to be statistically significant. The statistical analyses were performed using RStudio version 2022.12.0.

Results

Of the 257 patients who were eligible, 81 (31.5%) participated in the trial. Additionally, 63 (24.5%) patients completed the baseline questionnaire only (Fig. 1). The average age of patients in the trial was 59 (SD: 11) years, 42.7% had a high education level, and 60.0% received primary surgery followed by adjuvant chemotherapy treatment (Table 1). No difference in age, and clinical characteristics were observed between participants in the trial and patients who solely completed the baseline questionnaire.

Fig. 1.

Flow diagram of inclusion of individuals within the PADOVA study.

Table 1.

Sociodemographic and clinical characteristics of patients with ovarian cancer.

Characteristics	Intervention group (n = 40)	Control group (n = 41)
Sociodemographic
Age, mean (SD) year	55.5 (10.4)	59.7 (10.6)
Married/living together, n (%) yes	27 (67.5)	27 (65.9)a
Education, n (%)
Low	4 (10.0)	7 (17.1)
Intermediate	16 (40.0)	15 (36.6)
High	20 (50.0)	16 (39.0)a
Smoking, n (%) yes	2 (5.0)	1 (2.6)a
Comorbidity, n (%) ≥1	31 (77.5)	30 (73.2)
Clinical
Stage of disease, n (%)
Stage I	4 (10.0)	6 (14.6)
Stage II	6 (15.0)	7 (17.1)
Stage III	23 (57.5)	24 (58.5)
Stage IV	7 (17.5)	4 (9.8)
Type of treatment, n (yes)
Neoadjuvant chemotherapy	16 (40.0)	15 (36.6)
Adjuvant chemotherapy	24 (60.0)	26 (63.4)

^aMissing data for three participants.

Intervention adherence and contamination

In the intervention group, median attendance for the physiotherapist was 71.7% (IQR: 52.9–84.9%). Among the 39 participants with available physical activity logs, 16 participants attended more than 75% of the prescribed exercise sessions. Reasons for non-attendance was known for 91% of the non-attended sessions and main reasons included conflicting hospital appointments (15%), feeling too ill (14.2%), and being on holiday (6.6%). The median exercise relative dose intensity ranged from 0 to 97.3% (mean 72.6%, SD 16.0). In total, 28.5% of the attended sessions were modified. Of these modifications, 36.3% occurred in the resistance, 30.8% in the endurance- and 32.9% in both the resistance- and endurance exercises. In the control group, five participants indicated to follow a twice weekly exercise programme provided by a physical therapist for an unspecified duration between T0 and T1. This initiative was undertaken at the participants’ own discretion outside the study context.

Dietary records of dietary sessions were incomplete for nine participants. Dietary counselling focused on a sufficient calorie and protein intake to prevent weight loss in 70.0%, WCRF guidelines in 26.7% and both a sufficient calorie and protein intake and the WCRF guidelines in 3.3% of all attended sessions. Median attendance was 80.0% (IQR: 50.0–100.0%), and among the 31 participants from whom we have data, 18 participants attended more than 75% of the prescribed dietary sessions. In the control group, 9 participants attended sessions from a dietitian, with the number of consultations ranging from 1 to 6.

In total, seventeen participants (21%) were dropout, defined as those who did not complete post-test assessments. Among these, thirteen participants dropped-out before the first follow-up measurement and were therefore not included in the effect analyses. Lost to follow-up rates were similar in the intervention and the control group. Furthermore, it was observed that patients undergoing neoadjuvant chemotherapy were more likely to be lost to follow-up than patients receiving adjuvant chemotherapy treatment (OR = 4.1, 95% CI = 1.3–12.4).

Effects on primary and secondary endpoints

FFM, and physical functioning increased, and general fatigue decreased statistically significantly in both study arms over time, with no significant differences between groups at T1 (FFM in kg: β = −0.5; 95% CI = −3.2; 2.1, physical function: β = 1.4; 95% CI = −5.4; 8.3, fatigue: β = 0.7; 95% CI = −1.5; 2.8) and T2 (FFM: β = −0.9; 95% CI = −3.5; 1.8, physical function: β = −1.6; 95% CI = −8.7; 5.5, fatigue: β = −0.1; 95% CI = −2.3; 2.1; Table 2). No significant interaction effects were observed for age, treatment regimen and baseline value of the corresponding outcome. Per-protocol analyses showed comparable effects for body composition and fatigue to the intention-to-treat analyses, but the intervention effect on physical functioning became larger (β = 5.0; 95% CI = −2.8; 12.9, Supplementary Table 1). No significant differences between groups were found for the secondary outcomes of physical fitness, HRQoL, anxiety and depression, symptoms of neuropathy, sleep disturbances, physical activity, and dietary intake (Table 2).

Table 2.

Descriptive statistics of outcome values at baseline and post-treatment, and intervention effects for within and between-group differences at T1 and T2 in patients with ovarian cancer (n = 68).

		Descriptive over time			Within and between-group differences
		Baseline mean (SD)	Post-treatment T1 mean (SD)	Post-treatment T2 mean (SD)	Within-group differences T0 → T1	Within-group differences T0 → T2	Between-group difference at T1	Between-group difference at T2
Body composition
Body mass index (kg/m2)	Intervention	26.4 (6.0)	27.3 (6.8)	28.0 (7.1)	0.9 (0.5; 1.3)*	1.0 (0.6; 1.5)*	0.7 (−1.8; 3.3) REF	0.6 (−2.0; 3.1) REF
	Control	25.3 (3.8)	25.9 (4.0)	26.4 (2.8)	1.1 (0.6; 1.6)*	1.4 (0.9; 1.9)*
Fat mass (BIA), kg	Intervention	29.0 (10.9)	30.7 (12.8)	32.4 (13.3)	1.9 (1.0; 2.8)*	2.3 (1.3; 3.3)*	0.3 (−4.5; 5.1) REF	−0.2 (−4.6; 5.0) REF
	Control	27.9 (8.4)	29.5 (9.4)	30.2 (7.7)	2.0 (0.9; 3.1)*	2.5 (1.4; 3.6)*
Fat-free mass (BIA), kg	Intervention	43.7 (5.7)	44.3 (5.6)	44.7 (5.9)	0.5 (−0.3; 1.4)	0.5 (−0.4; 1.4)	−0.5 (−3.2; 2.1) REF	−0.9 (−3.5; 1.8) REF
	Control	43.4 (5.3)	44.0 (4.8)	45.3 (4.7)	0.9 (0.0; 1.9)*	1.2 (0.2; 2.2)*
Cardiorespiratory fitness
PeakVO2 (mL/kg/min)	Intervention	19.9 (4.9)	20.5 (5.2)	22.0 (4.9)	−0.2 (−1.4; 1.1)	1.2 (−0.2; 2.6)	−0.4 (−2.8; 2.0) REF	−1.5 (−4.0; 1.0) REF
	Control	20.6 (4.8)	20.5 (5.0)	22.8 (4.6)	−0.6 (−1.9; 0.7)	1.8 (0.4; 3.2)*
Peak power output (W)	Intervention	115.9 (38.8)	130.4 (42.6)	141.5 (45.1)	6.3 (−1.1; 14.0)	14.2 (6.3; 22.5)*	5.9 (−9.6; 21.3) REF	−8.5 (−24.5; 7.6) REF
	Control	119.4 (33.6)	120.0 (34.2)	144.0 (41.2)	−6.5 (−14.2; 1.2)	15.7 (7.5; 24.0)*
Ventilatory threshold VT1 (mL/kg/min)	Intervention	11.9 (3.6)	12.0 (3.1)	12.9 (3.1)	−0.2 (−1.4; 1.1)	0.5 (−0.8; 1.8)	−0.0 (−1.7; 1.7) REF	−0.9 (−2.7; 0.9) REF
	Control	12.9 (3.8)	12.2 (3.1)	13.3 (3.5)	−0.9 (−2.2; 0.3)	0.6 (−0.7; 2.0)
Ventilatory threshold VT2 (mL/kg/min)	Intervention	15.3 (5.0)	16.1 (4.0)	17.3 (4.7)	0.4 (−0.9; 1.7)	1.5 (0.1; 2.9)*	−0.2 (−1.9; 2.3) REF	−0.6(−2.8; 1.7) REF
	Control	16.0 (3.9)	16.1 (4.2)	17.8 (3.7)	−0.2 (−1.6; 1.1)	1.6 (0.2; 3.1)*
Muscle strength
Knee extension (Newton)	Intervention	226.4 (60.8)	248.5 (63.2)a	247.5 (65.5)a	18.7 (−4.8; 42.9)	18.0 (−7.1; 43.6)	−25.3 (−57.1; 6.7) REF	6.1 (−28.3; 40.5) REF
	Control	238.1 (59.3)a	256.8 (88.0)	235.0 (60.9)	25.9 (−0.3; 51.5)	−6.2 (−33.9; 21.4)
Elbow flection (Newton)	Intervention	173.0 (42.3)	182.4 (39.7)	182.3 (36.7)	8.8 (−3.9; 21.5)	3.9 (−9.5; 17.6)	−4.2 (−24.6; 16.3) REF	−17.1 (−38.8; 4.6) REF
	Control	186.7 (48.2)	180.8 (48.5)	191.9 (32.9)	−8.3 (−22.1; 5.5)	0.3 (15.0; 14.8)
Fatigue
Physical fatigue (MFI: range 4–20)	Intervention	12.0 (5.0)	11.8 (4.3)	8.9 (3.9)	−0.2 (−1.8; 1.4)	−3.1 (−4.8; −1.5)*	0.3 (−2.0; 2.6) REF	0.4 (−1.9; 2.8) REF
	Control	11.4 (5.1)	10.7 (4.6)	7.8 (4.0)	0.0 (−1.8; 1.7)	−3.1 (−4.9; −1.4)*
General fatigue (MFI: range 4–20)	Intervention	11.7 (4.1)	12.6 (4.1)	9.7 (4.0)	1.1 (−0.6; 2.6)	−2.0 (−3.6; −0.3)*	0.7 (−1.5; 2.8) REF	−0.1 (−2.3; 2.1) REF
	Control	10.9 (4.9)	11.4 (4.5)	9.1 (4.4)	1.2 (−0.5; 2.8)	−1.1 (−2.8; 0.6)
Reduced activity (MFI: range 4–20)	Intervention	11.6 (3.8)	10.2 (4.8)	8.0 (4.0)	−1.4 (−3.0; 0.2)	−3.7 (−5.4; −2.0)*	−0.1 (−2.3; 2.1) REF	0.3 (−1.9; 2.8) REF
	Control	11.2 (5.4)	10.4 (4.3)	7.8 (3.9)	−0.3 (−2.1; 1.4)	−3.1 (−4.9; −1.4)*
Reduced motivation (MFI: range 4–20)	Intervention	9.2 (3.6)	7.7 (2.9)	7.1 (2.2)	−1.6 (−3.1; −0.2)*	−2.3 (−3.7; −0.9)*	−0.5 (−2.3; 1.3) REF	0.1 (−1.8; 1.9) REF
	Control	9.6 (4.7)	8.1 (3.9)	7.0 (3.5)	−0.9 (−2.4; 0.6)	−2.1 (−3.7; −0.6)*
Mental fatigue (MFI; range 4–20)	Intervention	8.7 (4.3)	9.7 (4.3)	8.7 (3.7)	1.2 (−0.3; 2.6)	0.2 (−1.3; 1.7)	0.0 (−2.0; 1.9) REF	−1.3 (−3.3; 0.8) REF
	Control	9.1 (3.6)	9.5 (4.0)	9.4 (3.9)	0.7 (−0.9; 2.2)	1.0 (−0.6; 2.6)
Fatigue (EORTC)	Intervention	44.2 (23.8)	35.6 (21.1)	21.8 (17.8)	−6.5 (−14.6; 1.3)	−20.2 (−28.5; −12.2)*	6.0 (−5.0; 16.9) REF	4.9 (−6.5; 16.2) REF
	Control	34.3 (24.5)	29.1 (19.2)	17.0 (17.4)	−3.6 (−12.2; 4.7)	−16.2 (−25.0; −7.6)*
Health-related quality of life (EORTC: range 0–100)
Global quality of life	Intervention	60.6 (20.0)	70.4 (13.8)	76.9 (13.5)	9.8 (3.2; 16.4)*	16.6 (9.8; 23.4)*	−2.0 (−11.0; 7.0) REF	−2.3 (−11.7; 7.0) REF
	Control	65.7 (23.4)	75.2 (15.1)	81.7 (13.4)	6.6 (−0.4; 13.8)	13.7 (6.5; 21.1)*
Physical functioning	Intervention	83.7 (17.5)	85.1 (14.9)	92.3 (8.0)	0.7 (−3.9; 5.5)	7.6 (2.8; 12.5)*	1.4 (−5.4; 8.3) REF	−1.6 (−8.7; 5.5) REF
	Control	85.6 (13.1)	82.5 (12.1)	92.0 (9.0)	−2.7 (−7.7; 2.3)	7.2 (2.0; 12.4)*
Role functioning	Intervention	50.8 (26.9)	66.1 (27.6)	76.5 (17.5)	15.0 (5.2; 24.9)*	25.3 (15.2; 35.4)*	1.4 (−5.4; 8.3) REF	−11.7 (−24.8; 1.6) REF
	Control	59.3 (29.9)	71.1 (24.0)	88.3 (15.3)	11.7 (1.2; 22.3)*	33.6 (20.6; 42.4)*
Emotional functioning	Intervention	72.3 (25.1)	81.9 (13.6)	82.4 (16.1)	9.0 (2.1; 16.0)*	9.2 (2.1; 16.4)*	−0.1 (−9.1; 9.0) REF	−1.5 (−10.9; 7.9) REF
	Control	79.4 (18.5)	81.9 (16.2)	84.4 (15.0)	1.4 (−6.0; 8.9)	3.0 (−4.6; 10.8)
Cognitive functioning	Intervention	82.9 (19.8)	77.6 (24.5)	84.0 (13.5)	−5.2 (−12.7; 2.4)	1.5 (−6.2; 9.2)	−0.8 (−10.3; 8.7) REF	6.3 (−3.6; 16.2) REF
	Control	87.5 (14.6)	79.9 (23.1)	80.0 (18.3)	−7.9 (−15.9; 0.2)	−8.3 (−16.6; 0.1)
Social functioning	Intervention	65.4 (26.8)	77.0 (20.6)	85.2 (17.5)	10.8 (2.5; 19.2)*	18.9 (10.4; 27.6)*	−0.5 (−11.7; 10.7) REF	0.2 (−11.4; 11.9) REF
	Control	71.8 (25.1)	78.4 (21.5)	85.0 (18.2)	4.0 (−4.8; 13.1)	11.3 (2.2; 20.7)*
Distress (range 0–21)
Anxiety	Intervention	5.8 (3.8)	4.9 (2.4)	5.3 (3.2)	−0.7 (−1.8; 0.3)	−0.2 (−1.3; 0.8)	0.1 (−1.5; 1.7) REF	0.4 (−1.3; 2.1) REF
	Control	5.8 (4.2)	5.3 (3.3)	4.9 (3.1)	−0.2 (−1.3; 0.9)	0.0 (−1.1; 1.2)
Depression	Intervention	4.0 (3.5)	2.9 (1.9)	2.3 (2.2)	−0.9 (−2.0; 0.2)	−1.5 (−2.6; −0.4)*	−0.1 (−1.5; 1.4) REF	0.2 (−1.2; 1.7) REF
	Control	4.2 (3.4)	3.4 (3.1)	2.2 (2.3)	−0.7 (−1.9; 0.5)	−1.6 (−2.8; −0.4)*
Neuropathy (EORTC: range 0–100)
Motor	Intervention	4.7 (7.0)	11.0 (13.6)	8.2 (11.3)	6.6 (2.1; 11.0)*	4.1 (−0.5; 8.6)	−2.2 (−8.2; 3.6) REF	−1.6 (−7.7; 4.5) REF
	Control	6.0 (9.8)	12.7 (14.4)	9.7 (11.0)	7.0 (2.3; 11.8)*	3.9 (−0.9; 8.8)
Sensory	Intervention	5.2 (11.3)	20.7 (20.3)	15.8 (17.8)	15.9 (9.5; 22.1)*	10.9 (4.4; 17.4)*	3.8 (−4.5; 12.1) REF	2.3 (−6.3; 10.8) REF
	Control	4.0 (11.3)	16.2 (17.8)	12.8 (16.3)	13.7 (7.0; 20.5)*	10.4 (3.5; 17.2)*
Autonomic	Intervention	10.7 (15.5)	14.4 (21.2)	11.1 (18.5)	4.1 (−2.3; 10.4)	0.9 (−5.6; 7.4)	1.1 (−7.7; 10.0) REF	−0.3 (−9.4; 8.9) REF
	Control	9.0 (12.8)	13.1 (18.5)	10.0 (14.9)	5.2 (−1.6; 12.0)	3.5 (−3.4; 10.4)
Sleep disturbances (PSQI: range 0–21)
Sleep disturbances	Intervention	6.8 (3.5)	5.3 (2.8)	5.7 (2.7)	−1.5 (−2.5; −0.4)*	−0.9 (−2.0; 0.1)	−1.6 (−3.2; 0.1) REF	−0.5 (−2.2; 1.2) REF
	Control	7.6 (3.6)	6.9 (3.5)	5.7 (3.2)	−0.5 (−1.6; 0.7)	−1.0 (−2.1; 0.2)
Physical activity
Self-reported physical activity (PASE)	Intervention	69.6 (65.9)	110.8 (87.3)	133.7 (88.7)	43.5 (18.8; 68.2)*	65.8 (40.4; 91.1)*	−0.9 (−37.1; 35.2) REF	−38.3 (−75.7;−1.0)* REF
	Control	70.7 (48.7)	108.7 (55.8)	163.4 (83.0)	38.0 (12.0; 64.0)*	97.6 (70.8; 124.4)*
Dietary intake
Adherence to WCRF/AICR lifestyle recommendations (Range: 0–5)	Intervention	3.2 (0.6)	3.1 (0.8)	3.4 (0.8)	−0.1 (−0.4; 0.1)	0.2 (−0.1; 0.4)	−0.2 (−0.6; 0.2) REF	−0.1 (−0.5; 0.3) REF
	Control	3.5 (0.7)	3.3 (0.9)	3.6 (0.9)	−0.2 (−0.5; 0.1)	0.0 (−0.3; 0.3)
Dutch Healthy Diet Index (Range:0–80)	Intervention	49.8 (11.4)	48.8 (11.3)	52.8 (13.2)	0.1 (−3.9; 3.9)	3.4 (−0.6; 7.4)	−5.6 (−11.4; 0.2) REF	−2.1 (−8.1; 3.9) REF
	Control	53.7 (10.0)	54.9 (9.9)	55.3 (11.5)	1.6 (−2.5; 5.7)	1.5 (−2.8; 5.7)
Total caloric intake (kcal)	Intervention	1844 (600)	1794 (522)	1913 (667)	−53 (−314; 208)	85 (−178; 344)	−84 (−412; 247) REF	214 (−119; 549) REF
	Control	1771 (558)	1834 (613)	1680 (710)	61 (−222; 345)	−100 (−383; 186)
Total protein intake (gram/kg body weight)	Intervention	1.1 (0.5)	1.1 (0.4)	1.1 (0.3)	0.0 (−0.2; 0.1)	0.0 (−0.2; 0.1)	0.1 (−0.2; 0.3) REF	0.2 (−0.1; 0.4) REF
	Control	1.1 (0.4)	1.0 (0.4)	1.0 (0.5)	0.0 (−0.2; 0.2)	−0.1 (−0.3; 0.1)

BIA bioelectrical impedance assessment, EORTC European Organization for Research and Treatment of Cancer Quality of Life Questionnaire Core, MFI Multidimensional Fatigue Inventory, PSQI Pittsburgh Sleep Quality Index, WCRF/AICR World Cancer Research Fund/American Institute for Cancer Research.

*Indicates a statistically significant effect.

Discussion

This study examined the effectiveness of a combined exercise and dietary intervention on body composition, fatigue and physical functioning as primary outcomes and explored the effects on several secondary outcomes. In contrast to our hypotheses, no differences between the intervention and control group were found on the primary outcomes. In both the intervention and control group, body composition and physical functioning increased, and fatigue decreased over time with similar trajectories.

There are several possible explanations for why we did not find a difference between the intervention and control group. Firstly, ovarian cancer is often diagnosed at an advanced stage when multiple symptoms, including pain and/or ascites are present, hampering physical functioning and HRQoL already at diagnosis. Surgery and chemotherapy help alleviate these symptoms, thereby improving physical functioning and HRQoL [58]. Apparently, exercise and dietary interventions during chemotherapy may not provide additional benefits on these outcomes in patients with ovarian cancer. This is in contrast with, for example, patients with breast cancer who usually do not report symptoms at diagnosis and for whom physical functioning and HRQoL deteriorates substantially during chemotherapy [59]. Our finding that physical activity also improved in both groups during chemotherapy treatment in patients with advanced ovarian cancer supports the theory that chemotherapy and surgery improves physical health by reducing tumour burden. This theory is also further corroborated by another longitudinal study demonstrating improved HRQoL following treatment [60].

Secondly, while on average, no significant intervention effects were observed, we found large heterogeneity in effects as indicated by the large confidence intervals around the effect sizes. This may suggest that there may be subgroups who could benefit more from the intervention than others. Our explorative analysis suggested that this heterogeneity in effect is unlikely explained by baseline values of fatigue, physical functioning and body composition, nor by age or treatment.

Thirdly, the effectiveness of an exercise and dietary intervention may substantially be influenced by the level of adherence to the intervention and the risk of contamination in the control group. In this study, we observed modest adherence rates and fairly high contamination rates [6, 61]. Previous research indicated that patients with ovarian cancer often preferred assignment to the exercise and dietary intervention group. They expressed concerns that post-chemotherapy counselling sessions were insufficient and arrived too late [50], which may have prompted participants to seek help independently. However, this reasoning is unlikely to entirely account for the absence of a significant difference between the intervention and control groups.

A strength of this study is its randomised controlled design, allowing us to be the first to have sufficient statistical power to examine the effects of an exercise and dietary intervention on body composition, fatigue, and physical functioning. In addition, our intervention was developed based on exercise principles, state-of-the-art dietary guidelines, and Bandura’s Social Cognitive Theory [32], which has previously demonstrated success in improving health behaviours in patients with cancer [62]. Furthermore, this study was adequately tailored to this patient-specific group by adjusting the intervention to comorbidities and disease-induced and treatment-induced effects [63]. By adjusting the intervention to these comorbidities and disease-induced and treatment-induced effects, we aimed to maximise adherence en minimise dropout [33].

Our study also had several limitations that should be noted. First, due to logistic constraints regarding the CT scans, we opted for the more readily accessible BIA method which may have impacted the statistical power to detect differences in body composition. In addition, it is important to recognise that the BIA may have limitations in patients with ascites, which could further affect the accuracy and interpretation of our body composition results. Second, ovarian cancer is often diagnosed at a higher stage of disease, with a median age for diagnosis ranging from 50 to 79 years, varying by region [64, 65]. Given this context, it was expected that both recruitment and completion rates would be lower compared to studies involving patients diagnosed at an earlier stage or at a younger age such as for breast cancer [66]. The current participation rate of 31.5% aligns with the range observed in similar studies, typically ranging from 16 to 63% [21, 22]. In addition, dropout from the study was somewhat higher (21%) compared to the anticipated 15% [6, 51, 57], but in line with dropout rates of patients with advanced cancer [21, 22]. Notably, patients receiving neoadjuvant chemotherapy treatment were more likely to drop out compared to patients who received surgery first. This finding suggests that patients for whom primary surgery is withheld represent a group who are more challenging to retain in exercise interventions, requiring special attention when offering exercise programmes to this specific subgroup.

Nevertheless, it is worth noting that combined exercise and dietary intervention can be safely recommended to patients with ovarian cancer during and shortly after completion of cancer treatment, as no adverse advents were recorded during our study. In addition, it is important to acknowledge that, despite the absence of statistically significant differences, participants in the exercise and dietary intervention group expressed that they would recommend this intervention to other patients, citing improved quality of life, maintenance of physical fitness during chemotherapy, and quicker recovery after surgery and chemotherapy [50]. This underscores that the qualitative assessments conducted previously captured aspect that our quantitative measures may not have fully captured. This suggests that our current measurement tools may not fully capture the nuanced impacts of these interventions.

In conclusion, both groups exhibited enhancements in body composition, physical functioning, and fatigue levels during chemotherapy. Hence, the intervention group did not demonstrate additional benefits when compared to the control group, although the wide confidence intervals suggest that some individuals may have derived greater benefit than others. The unique characteristics and needs of patients with ovarian cancer necessitates tailored research and interventions specific to this population. Furthermore, the differential effects of exercise and diet during chemotherapy as compared with other cancer populations underscore the importance of conducting research in understudied populations, as results may not be directly applicable or generalisable across diverse cancer types.

Author contributions

LB, MH, GK and SS designed the study. SS and YH coordinated the trial, collected and assembled the data. MK analysed the data, and LB and MK drafted the manuscript. CB, MtT, LvL, WvD, RB and PO contributed to patient recruitment and measurements. MH and LB assisted with conceptualising the analyses and interpreting the results. All authors critically revised the manuscript and approved the final manuscript.

Funding

The PADOVA study is funded by the Dutch Cancer Society, grant number VU 2015–7950. The Dutch Cancer Society was not involved in the conceptualisation, design, data collection, analysis, decision to publish, nor in preparation of the manuscript.

Data availability

The datasets used and/or analysed during this study are available from the corresponding author on reasonable request.

Competing interests

The authors declare no competing interests.

Ethics approval and consent to participate

All participants provided written informed consent before participating. The study was approved by the medical ethical committees of the Amsterdam UMC and participating hospitals.

Consent for publication

Not applicable.

Supplementary information

The online version contains supplementary material available at 10.1038/s41416-024-02694-8.