Machine learning predicts peak oxygen uptake and peak power output for customizing cardiopulmonary exercise testing using non-exercise features.

PURPOSE: Cardiopulmonary exercise testing (CPET) is considered the gold standard for assessing cardiorespiratory fitness. To ensure consistent performance of each test, it is necessary to adapt the power increase of the test protocol to the physical characteristics of each individual. This study aimed to use machine learning models to determine individualized ramp protocols based on non-exercise features. We hypothesized that machine learning models will predict peak oxygen uptake ( V ˙ O2peak) and peak power output (PPO) more accurately than conventional multiple linear regression (MLR). METHODS: The cross-sectional study was conducted with 274 (♀168, ♂106) participants who performed CPET on a cycle ergometer. Machine learning models and multiple linear regression were used to predict V ˙ O2peak and PPO using non-exercise features. The accuracy of the models was compared using criteria such as root mean square error (RMSE). Shapley additive explanation (SHAP) was applied to determine the feature importance. RESULTS: The most accurate machine learning model was the random forest (RMSE: 6.52 ml/kg/min [95% CI 5.21-8.17]) for V ˙ O2peak prediction and the gradient boosting regression (RMSE: 43watts [95% CI 35-52]) for PPO prediction. Compared to the MLR, the machine learning models reduced the RMSE by up to 28% and 22% for prediction of V ˙ O2peak and PPO, respectively. Furthermore, SHAP ranked body composition data such as skeletal muscle mass and extracellular water as the most impactful features. CONCLUSION: Machine learning models predict V ˙ O2peak and PPO more accurately than MLR and can be used to individualize CPET protocols. Features that provide information about the participant's body composition contribute most to the improvement of these predictions. TRIAL REGISTRATION NUMBER: DRKS00031401 (6 March 2023, retrospectively registered).

Cardiorespiratory fitness (V̇O2peak) across the adult lifespan in persons with multiple sclerosis and matched healthy controls.

OBJECTIVES: Cardiorespiratory fitness (measured as peak oxygen uptake; V̇O2peak) is a well-established health predictor in the general population and in persons with multiple sclerosis (pwMS). We aimed to investigate differences in V̇O2peak between age groups and the prevalence of V̇O2peak impairments across the adult lifespan in pwMS compared to age- and sex-matched healthy controls (HC). DESIGN: Cross-sectional study. METHODS: Data from 469 pwMS (EDSS range 1.0-7.0), who carried out graded cardiopulmonary exercise testing during their rehabilitation stay at the Valens clinic from 07/2010 to 10/2022, were retrospectively analyzed. Data from 21,063 HC were extracted from previously published studies containing normative reference values. RESULTS: With advanced age (i.e., across age groups), a continuous deterioration of V̇O2peak was observed in both pwMS and HC. Within all age groups, V̇O2peak was reduced in pwMS compared to HC with deficits ranging from 29 % to 40 % for females (p < .05), and from 30 % to 41 % for males (p < .05). However, no age ∗ group interaction was observed in neither males (p = .626) nor females (p = .557). With V̇O2peak impairments defined as values below the 5th percentile of HC, a high prevalence was observed in pwMS, with values ranging from 48 % to 100 % across age groups. CONCLUSIONS: The present data provide evidence for a parallel deterioration of V̇O2peak in pwMS and matched controls with advancing age, coinciding with a high prevalence of impairments in V̇O2peak already present in young adulthood in pwMS. Understanding the extent of impairments as well as the age trajectories of cardiorespiratory fitness in pwMS is crucial for designing optimal rehabilitative and preventive interventions.

On the impulsivity path: Examining the unique and conjoint relations between emotion- and non-emotion-related impulsivity, internalizing symptoms, alcohol use, and physical health parameters.

BACKGROUND: Researchers have increasingly differentiated trait-like tendencies toward impulsivity occurring during emotional states (emotion-related impulsivity [ERI]) from impulsivity not tied to emotion (non-ERI). Relative to non-ERI, ERI has shown robust correlations with psychopathology and mild to moderate associations with physical health parameters (e.g., physical activity, poor sleep quality, body mass index [BMI]). Therefore, we first aimed to investigate the unique contributions of ERI and non-ERI to psychopathology symptoms while controlling for neuroticism. Second, we sought to explore the combined associations of physical health parameters with several impulsivity forms. METHODS: German-speaking adults (N = 350, 35.9 ± 14.6 years, 69.1% female, BMI: 24.0 ± 4.8 kg/m2 , mostly students or employees) completed measures of impulsivity, psychopathology symptoms, neuroticism, and physical health. We gathered measures of two ERI forms: Feelings Trigger Action and Pervasive Influence of Feelings. As a control comparison, we gathered a measure of non-ERI, the Lack of Follow-Through scale. We conducted separate path models for Aims 1 and 2. RESULTS: For Aim 1, Pervasive Influence of Feelings showed strong links with internalizing symptoms. Feelings Trigger Action and Lack of Follow-Through showed small links with alcohol use. For Aim 2, poor sleep quality was related to all three impulsivity factors, while physical activity was only related to Pervasive Influence of Feelings and Lack of Follow-Through. BMI showed a curvilinear association with impulsivity. CONCLUSIONS: ERI is more directly relevant than non-ERI for psychopathology symptoms, emphasizing the need to differentiate between the two ERI types. The association of ERI and non-ERI with physical activity and poor sleep quality may serve as potential treatment targets for impulsivity-related problems.

Physical activity is related to disease severity and fatigue, but not to relapse rate in persons with relapsing remitting multiple sclerosis - a self-reported questionnaire based study.

Introduction: Based on theoretical models, physical activity has been introduced as a promoting method to mitigate the disease severity, fatigue and relapse rate in multiple sclerosis. The primary objective of the study was to investigate the relation between self-reported physical activity level and disease severity, fatigue and relapse rate in persons with relapsing remitting multiple sclerosis (RRMS). Methods: A survey was offered to persons with RRMS from March 2019 to August 2021 (n = 253). Physical activity level, fatigue and disease severity were determined using the Godin Leisure-Time Questionnaire (GLTEQ), the Patient Determined Disease Steps (PDDS) scale and the Fatigue Scale for Motor and Cognitive Functions (FSMC). Additionally, participants' relapse rate was recorded. Results: Bivariate correlations revealed an inverse relation between physical activity level and PDDS (ρ = -0.279; p < 0.001) as well as between physical activity and FSMC (r = -0.213, p < 0.001), but not between physical activity and relapse rate (r = 0.033, p > 0.05). Multiple linear regression analyses explained 12.6% and 5.2% of the variance of PDDS and FSMC. Conclusion: Our findings confirm a relation between self-reported physical activity, disease severity and fatigue in persons with RRMS. However, self-reported physical activity level does not seem to affect the annualised relapse rate.

Complexity and pitfalls in maximal exercise testing for persons with multiple sclerosis.

BACKGROUND AND PURPOSE: Valid measurements of cardiorespiratory fitness in persons with multiple sclerosis (pwMS) are essential during inpatient rehabilitation for a precise evaluation of the current health status, for defining appropriate exercise intensities, and for evaluation of exercise intervention studies. We aim (i) to examine the proportion of pwMS who attain the American College of Sports Medicine (ACSM) criteria for maximal effort during graded cardiopulmonary exercise testing (CPET) and (ii) to provide insight into participant characteristics that limit maximal exercise performance. METHODS: This cross-sectional study comprises a retrospective examination of ACSM criteria for maximal effort during graded CPET of n = 380 inpatient pwMS (mean age = 48 ± 11 years, 66% female). Chi-squared or Fisher's exact tests were conducted to compare differences in the distribution of criteria achieved. Participants' characteristics were examined as potential predictors using binary logistic regression. RESULTS: Only 60% of the overall sample attained a respiratory exchange ratio ≥ 1.10. With regard to the definition applied, only 24% or 40% of the participants achieved an oxygen consumption plateau, and 17% or 50% attained the heart rate criterion. Forty-six percent met at least two of three criteria. Disability status, gender, disease course, and body mass index were associated with the attainment of maximal effort. CONCLUSIONS: Our findings suggest that a relevant proportion of inpatient pwMS do not attain common criteria utilized to verify maximal oxygen consumption. Identified predictors for criteria attainment can be used to create models to predict cardiorespiratory fitness and to optimize CPET protocols in restrictive groups of pwMS.

Systematic Review of Exercise Studies in Persons with Multiple Sclerosis: Exploring the Quality of Interventions According to the Principles of Exercise Training.

INTRODUCTION: The objective of this systematic review is to explore the application and reporting of (i) the principles of exercise training in exercise trials, (ii) the components of exercise prescription, and (iii) the adherence towards the prescribed programmes in randomised controlled trials (RCTs) in persons with multiple sclerosis (pwMS). METHODS: The MEDLINE, CINAHL, SPORTDiscus, PubMed and Embase electronic databases were searched from 1 January 2000 to 16 October 2020. RCTs comprising at least 3 weeks of aerobic and/or resistance exercise intervention in pwMS that reported at least one physiological outcome and were published in peer-reviewed journals were eligible for inclusion. RESULTS: Out of 52 RCTs included in this review, 58 intervention arms were examined. None applied more than four principles of exercise training. Specificity was addressed by 85%, progression by 33%, overload by 59%, initial values by 26%, reversibility by 0% and diminishing returns by 2% of trials. Fifty-two percent of trials reported all components of exercise prescription, and 3% of trials reported the level of adherence to the prescribed exercise. CONCLUSION: This systematic review reveals that exercise training principles were not respected in the majority of included RCTs. The weak quality of reported exercise interventions limits the interpretation of the studies' results and potentially leads to an underestimation of 'exercise as medicine' in pwMS. Also, the vague descriptions of exercise prescription and adherence impede the reproducibility of results. Future studies must attend to all principles of exercise training and provide transparent information on the prescribed and performed programmes to develop specific and valid exercise recommendations for pwMS. SYSTEMATIC REVIEW REGISTRATION: CRD42020162671, 28/04/2020, PROSPERO.

Influence of combined functional resistance and endurance exercise over 12 weeks on matrix metalloproteinase-2 serum concentration in persons with relapsing-remitting multiple sclerosis - a community-based randomized controlled trial.

BACKGROUND: The relevance of regular moderate to intense exercise for ameliorating psychomotor symptoms in persons with multiple sclerosis (pwMS) is becoming increasingly evident. Over the last two decades, emerging evidence from clinical studies and animal models indicate immune regulatory mechanisms in both periphery and the central nervous system that may underlie these beneficial effects. The integrity of the blood-brain barrier as the main structural interface between periphery and brain seems to play an important role in MS. Reducing the secretion of proteolytic matrix metalloproteinases (MMP), i.e. MMP-2, as disruptors of blood-brain barrier integrity could have profound implications for MS. METHODS: In this two-armed randomized controlled trial 64 participants with relapsing-remitting MS (RRMS) (EDSS 0-4.0) will be allocated to either an intervention group or a passive wait list control group. The intervention group will perform 60 min of combined functional resistance and endurance exercises 3x per week over a period of 12 weeks in a community-based and publicly available setting. Changes in serum concentration of MMP-2 will be the primary outcome. Secondary outcomes are numbers of immune cell subsets, soluble (anti-) inflammatory factors, physical capacity, cognitive performance, physical activity behavior, gait performance, and patient-reported outcomes. All outcome measures will be assessed at baseline and after week 12 with an additional blood sampling before, during and immediately after a single training session in week 6. DISCUSSION: To our knowledge, this will be the first RCT to investigate both the acute and chronic effects of a community-based intense functional resistance and endurance exercise regimen in persons with RRMS. Combining analysis of biological and cognitive or psychological outcomes may provide a better understanding of the MS-specific symptomology. TRIAL REGISTRATION: DRKS00017091; 05th of April, 2019; International Clinical Trials Registry Platform.