Mastering variation: variance components and personalised medicine

Influence of Lumbar Mobilizations During the Nordic Hamstring Exercise on Hamstring Measures of Knee Flexor Strength, Failure Point, and Muscle Activity: A Randomized Crossover Trial

OBJECTIVE

The aims of this study were to quantify the effects of spinal mobilization on force production, failure point, and muscle activity of the hamstrings during the Nordic hamstring exercise (NHE), and to explore individual differences in responses.

METHODS

In a replicated randomized crossover trial, 24 asymptomatic, recreationally active men (age [mean ± standard deviation]: 27 ± 6 years; body mass: 82 ± 17 kg; height: 181 ± 8 cm) completed 2 standardized intervention trials (L4/5 zygapophyseal mobilizations) and 2 control trials. The failure point of the NHE was determined with 3D motion capture. Peak force, knee flexor torque, and electromyography (EMG) of the biceps femoris were measured. Data analyses were undertaken to quantify mean intervention response and explore any individual response heterogeneity.

RESULTS

Mean (95% confidence interval) left-limb force was higher in intervention than in control trials by 18.7 (4.6-32) N. Similarly, right-limb force was higher by 22.0 (3.4-40.6) N, left peak torque by 0.14 (0.06-0.22) N • m, and right peak torque by 0.14 (0.05-0.23) N • m/kg. Downward force angle was decreased in intervention vs control trials by 4.1° (0.5°-7.6°) on the side of application. Both peak EMG activity (P = .002), and EMG at the downward force (right; P = .020) increased in the intervention condition by 16.8 (7.1-26.4) and 8.8 (1.5-16.1) mV, respectively. Mean downward acceleration angle changed by only 0.3° (-8.9° to 9.4°) in intervention vs control trials. A clear response heterogeneity was indicated only for right force (Participant × Intervention interaction: P = .044; response heterogeneity standard deviation = 34.5 [5.7-48.4] N). Individual response heterogeneity was small for all other outcomes.

CONCLUSION

After spinal mobilization, immediate changes in bilateral hamstring force production and peak torque occurred during the NHE. The effect on the NHE failure point was unclear. Electromyographic activity increased on the ipsilateral side. Response heterogeneity was generally similar to the random trial-to-trial variability inherent in the measurement of the outcomes.

On the treatment effect heterogeneity of antidepressants in major depression: A Bayesian meta-analysis and simulation study

Background

The average treatment effect of antidepressants in major depression was found to be about 2 points on the 17-item Hamilton Depression Rating Scale, which lies below clinical relevance. Here, we searched for evidence of a relevant treatment effect heterogeneity that could justify the usage of antidepressants despite their low average treatment effect.

Methods

Bayesian meta-analysis of 169 randomized, controlled trials including 58,687 patients. We considered the effect sizes log variability ratio (lnVR) and log coefficient of variation ratio (lnCVR) to analyze the difference in variability of active and placebo response. We used Bayesian random-effects meta-analyses (REMA) for lnVR and lnCVR and fitted a random-effects meta-regression (REMR) model to estimate the treatment effect variability between antidepressants and placebo.

Results

The variability ratio was found to be very close to 1 in the best fitting models (REMR: 95% highest density interval (HDI) [0.98, 1.02], REMA: 95% HDI [1.00, 1.02]). The between-study standard deviation τ under the REMA with respect to lnVR was found to be low (95% HDI [0.00, 0.02]). Simulations showed that a large treatment effect heterogeneity is only compatible with the data if a strong correlation between placebo response and individual treatment effect is assumed.

Conclusions

The published data from RCTs on antidepressants for the treatment of major depression is compatible with a near-constant treatment effect. Although it is impossible to rule out a substantial treatment effect heterogeneity, its existence seems rather unlikely. Since the average treatment effect of antidepressants falls short of clinical relevance, the current prescribing practice should be re-evaluated.

Interpretation of chronic pain clinical trial outcomes: IMMPACT recommended considerations

Interpreting randomized clinical trials (RCTs) is crucial to making decisions regarding the use of analgesic treatments in clinical practice. In this article, we report on an Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) consensus meeting organized by the Analgesic, Anesthetic, and Addiction Clinical Trial Translations, Innovations, Opportunities, and Networks, the purpose of which was to recommend approaches that facilitate interpretation of analgesic RCTs. We review issues to consider when drawing conclusions from RCTs, as well as common methods for reporting RCT results and the limitations of each method. These issues include the type of trial, study design, statistical analysis methods, magnitude of the estimated beneficial and harmful effects and associated precision, availability of alternative treatments and their benefit-risk profile, clinical importance of the change from baseline both within and between groups, presentation of the outcome data, and the limitations of the approaches used.

Design and analysis considerations for utilizing a mapping function in a small sample, sequential, multiple assignment, randomized trials with continuous outcomes

Small sample, sequential, multiple assignment, randomized trials (snSMARTs) are multistage trials with the overall goal of determining the best treatment after a fixed amount of time. In snSMART trials, patients are first randomized to one of three treatments and a binary (e.g. response/nonresponse) outcome is measured at the end of the first stage. Responders to first stage treatment continue their treatment. Nonresponders to first stage treatment are rerandomized to one of the remaining treatments. The same binary outcome is measured at the end of the first and second stages, and data from both stages are pooled together to find the best first stage treatment. However, in many settings the primary endpoint may be continuous, and dichotomizing this continuous variable may reduce statistical efficiency. In this article, we extend the snSMART design and methods to allow for continuous outcomes. Instead of requiring a binary outcome at the first stage for rerandomization, the probability of staying on the same treatment or switching treatment is a function of the first stage outcome. Rerandomization based on a mapping function of a continuous outcome allows for snSMART designs without requiring a binary outcome. We perform simulation studies to compare the proposed design with continuous outcomes to standard snSMART designs with binary outcomes. The proposed design results in more efficient treatment effect estimates and similar outcomes for trial patients.

Response heterogeneity in a randomized controlled trial of telerehabilitation interventions among adults with multiple sclerosis

INTRODUCTION

Telerehabilitation may be effective on average but is not equally effective among all people with multiple sclerosis (MS). Thus, the purpose of this secondary analysis of a randomized controlled trial was to explore whether baseline characteristics of participants with MS influence fatigue and physical activity outcomes of three telerehabilitation interventions.

METHODS

Participants were randomized to contact-control intervention (CC), physical activity-only intervention (PA-only), and physical activity plus fatigue self-management intervention (FM+). The 12-week interventions were delivered over the phone. Sociodemographic (age and income), clinical (comorbidities, mental function and physical function), psychosocial (self-efficacy, outcome expectations and goal-setting), and behavioural baseline characteristics (step count and fatigue self-management behaviors) were used in a moderated regression analysis and a responder analysis to examine their influence on the Fatigue Impact Scale (FIS) and Godin Leisure-Time Exercise Questionnaire (GLTEQ) at post-test (i.e. immediately post-interventions).

RESULTS

No interactions terms were statistically significant in the moderation analysis. However, the responder analysis showed that baseline psychosocial characteristics and mental function were significantly different (p < 0.05) between responders and non-responders. Specifically, non-responders on the FIS at post-test in the PA-only intervention had significantly lower baseline scores in goal setting for engaging in fatigue self-management behaviours. Also, non-responders on the GLTEQ at post-test in the FM+ intervention had significantly worse baseline scores in mental function.

DISCUSSION

Further research is needed to understand the complex relationship among baseline characteristics, telerehabilitation and response heterogeneity. We discuss how research on examining response heterogeneity may be advanced by conducting mega-clinical trials, secondary analyses of big data, meta-analyses and employing non-traditional research designs.

TRIAL REGISTRATION

Clinicaltrials.gov (NCT01572714).

Precision Medicine in Type 2 Diabetes: Using Individualized Prediction Models to Optimize Selection of Treatment

Despite the known heterogeneity of type 2 diabetes and variable response to glucose lowering medications, current evidence on optimal treatment is predominantly based on average effects in clinical trials rather than individual-level characteristics. A precision medicine approach based on treatment response would aim to improve on this by identifying predictors of differential drug response for people based on their characteristics and then using this information to select optimal treatment. Recent research has demonstrated robust and clinically relevant differential drug response with all noninsulin treatments after metformin (sulfonylureas, thiazolidinediones, dipeptidyl peptidase 4 [DPP-4] inhibitors, glucagon-like peptide 1 [GLP-1] receptor agonists, and sodium-glucose cotransporter 2 [SGLT2] inhibitors) using routinely available clinical features. This Perspective reviews this current evidence and discusses how differences in drug response could inform selection of optimal type 2 diabetes treatment in the near future. It presents a novel framework for developing and testing precision medicine-based strategies to optimize treatment, harnessing existing routine clinical and trial data sources. This framework was recently applied to demonstrate that "subtype" approaches, in which people are classified into subgroups based on features reflecting underlying pathophysiology, are likely to have less clinical utility compared with approaches that combine the same features as continuous measures in probabilistic "individualized prediction" models.

Time to reality check the promises of machine learning-powered precision medicine

Machine learning methods, combined with large electronic health databases, could enable a personalised approach to medicine through improved diagnosis and prediction of individual responses to therapies. If successful, this strategy would represent a revolution in clinical research and practice. However, although the vision of individually tailored medicine is alluring, there is a need to distinguish genuine potential from hype. We argue that the goal of personalised medical care faces serious challenges, many of which cannot be addressed through algorithmic complexity, and call for collaboration between traditional methodologists and experts in medical machine learning to avoid extensive research waste.

Individual response to antidepressants for depression in adults-a meta-analysis and simulation study

BACKGROUND

The observation that some patients appear to respond better to antidepressants for depression than others encourages the assumption that the effect of antidepressants differs between individuals and that treatment can be personalized.

OBJECTIVE

To compare the outcome variance in patients receiving antidepressants with the outcome variance in patients receiving placebo in randomized controlled trials (RCTs) of adults with major depressive disorder (MDD) and to illustrate, using simulated data, components of variation of RCTs.

METHODS

From a dataset comprising 522 RCTs of antidepressants for adult MDD, we selected the placebo-controlled RCTs reporting outcomes on the 17 or 21 item Hamilton Depression Rating Scale or the Montgomery-Asberg Depression Rating Scale and extracted the means and SDs of raw endpoint scores or baseline to endpoint changes scores on eligible depression symptom rating scales. We conducted inverse variance random-effects meta-analysis with the variability ratio (VR), the ratio between the outcome variance in the group of patients receiving antidepressants and the outcome variance in the group receiving placebo, as the primary outcome. An increased variance in the antidepressant group would indicate individual differences in response to antidepressants.

RESULTS

We analysed 222 RCTs that investigated 19 different antidepressants compared with placebo in 345 comparisons, comprising a total of 61144 adults with an MDD diagnosis. Across all comparisons, the VR for raw endpoint scores was 0.98 (95% CI 0.96 to 1.00, I2 = 0%) and 1.00 (95% CI 0.99 to 1.02, I2 = 0%) for baseline-to-endpoint change scores.

CONCLUSION

Based on these data, we cannot reject the null hypothesis of equal variances in the antidepressant group and the placebo group. Given that RCTs cannot provide direct evidence for individual treatment effects, it may be most reasonable to assume that the average effect of antidepressants applies also to the individual patient.

An appraisal of the SDIR as an estimate of true individual differences in training responsiveness in parallel-arm exercise randomized controlled trials

Calculating the standard deviation of individual responses (SD_IR) is recommended for estimating the magnitude of individual differences in training responsiveness in parallel‐arm exercise randomized controlled trials (RCTs). The purpose of this review article is to discuss potential limitations of parallel‐arm exercise RCTs that may confound/complicate the interpretation of the SD_IR. To provide context for this discussion, we define the sources of variation that contribute to variability in the observed responses to exercise training and review the assumptions that underlie the interpretation of SD_IR as a reflection of true individual differences in training responsiveness. This review also contains two novel analyses: (1) we demonstrate differences in variability in changes in diet and physical activity habits across an intervention period in both exercise and control groups, and (2) we examined participant dropout data from six RCTs and found that significantly (P < 0.001) more participants in control groups (12.8%) dropped out due to dissatisfaction with group assignment compared to exercise groups (3.4%). These novel analyses raise the possibility that the magnitude of within‐subject variability may not be equal between exercise and control groups. Overall, this review highlights that potential limitations of parallel‐arm exercise RCTs can violate the underlying assumptions of the SD_IR and suggests that these limitations should be considered when interpreting the SD_IR as an estimate of true individual differences in training responsiveness.

Methods and Challenges for Assessing Heterogeneity

The widely acknowledged homogeneity assumption limits progress in refining clinical diagnosis, understanding mechanisms, and developing new treatments for mental health disorders. This homogeneity assumption drives both a comorbidity and a heterogeneity problem, where two different approaches tackle the problems. One, a unifying approach, tackles the comorbidity problem by assuming that a single general psychopathology factor underlies multiple disorders. Another, a multifactorial approach, tackles the heterogeneity problem by assuming that disorders comprise multiple subtypes driven by multiple discrete factors. We show how each of these approaches can make useful contributions to mental health-related research and clinical practice. For example, the unifying approach can develop a rapid assessment tool that may be clinically valuable for triaging cases. The multifactorial approach can reveal subtypes that are differentially responsive to treatments and highlight distinct mechanisms leading to similar phenotypes. Because both approaches tackle different problems, both have different limitations. We describe the statistical frameworks that incorporate and adjudicate between both approaches (e.g., the bifactor model, normative modeling, and the functional random forest). Such frameworks can identify whether sets of disorders are more affected by heterogeneity or comorbidity. Therefore, future studies that incorporate such frameworks can provide further insight into the nature of psychopathology.

Individual Differences in Response to Antidepressants: A Meta-analysis of Placebo-Controlled Randomized Clinical Trials

IMPORTANCE

Antidepressants are commonly used worldwide to treat major depressive disorder. Symptomatic response to antidepressants can vary depending on differences between individuals; however, this variability may reflect nonspecific or random factors.

OBJECTIVES

To investigate the assumption of systematic variability in symptomatic response to antidepressants and to assess whether this variability is associated with severity of major depressive disorder, antidepressant class, or year of study publication.

DATA SOURCES

Data used were from a recent network meta-analysis of acute treatment with licensed antidepressants in adults with major depressive disorder. The following databases were searched from inception to January 8, 2016: the Cochrane Central Register of Controlled Trials, CINAHL, Embase, LILACS database, MEDLINE, MEDLINE In-Process, and PsycINFO. Additional sources were international trial registries, drug approval agency websites, and key scientific journals.

STUDY SELECTION

Analysis was restricted to double-blind, randomized placebo-controlled trials with available data at the study's end point.

DATA EXTRACTION AND SYNTHESIS

Baseline and end point means, SDs, number of participants in each group, antidepressant class, and publication year were extracted. The data were analyzed between August 14 and November 18, 2019.

MAIN OUTCOMES AND MEASURES

With the use of validated methods, coefficients of variation were derived for antidepressants and placebo, and their ratios were calculated to compare outcome variability between antidepressant and placebo. Ratios were entered into a random-effects model, with the expectation that response to antidepressants would be more variable than response to placebo. Analysis was repeated after stratifying by baseline severity of depression, antidepressant class (selective serotonin reuptake inhibitors: citalopram, escitalopram, fluoxetine, fluvoxamine, paroxetine, sertraline, and vilazodone; serotonin and norepinephrine reuptake inhibitors: desvenlafaxine and venlafaxine; norepinephrine-dopamine reuptake inhibitor: bupropion; noradrenergic agents: amitriptyline and reboxetine; and other antidepressants: agomelatine, mirtazapine, and trazodone), and publication year.

RESULTS

In the 87 eligible randomized placebo-controlled trials (17 540 unique participants), there was significantly more variability in response to antidepressants than to placebo (coefficients of variation ratio, 1.14; 95% CI, 1.11-1.17; P < .001). Baseline severity of depression did not moderate variability in response to antidepressants. Variability in response to selective serotonin reuptake inhibitors was lower than variability in response to noradrenergic agents (coefficients of variation ratio, 0.88; 95% CI, 0.80-0.97; P = .01), as was the variability in response to other antidepressants compared with noradrenergic agents (coefficients of variation ratio, 0.87; 95% CI, 0.79-0.97; P = .001). Variability also tended to be lower in studies that were published more recently, with coefficients of variation changing by a value of 0.005 (95% CI, 0.002-0.008; P = .003) for every year a study is more recent.

CONCLUSIONS AND RELEVANCE

Individual differences may be systematically associated with responses to antidepressants in major depressive disorder beyond placebo effects or statistical factors. This study provides empirical support for identifying moderators and personalizing antidepressant treatment.

Variations in strength-related measures during the menstrual cycle in eumenorrheic women: A systematic review and meta-analysis

OBJECTIVES

To systematically review the current body of research that has investigated changes in strength-related variables during different phases of the menstrual cycle in eumenorrheic women.

DESIGN

Systematic review and meta-analysis.

METHODS

A literature search was conducted in Pubmed, SPORTDiscus and Web of Science using search terms related to the menstrual cycle and strength-related measures. Two reviewers reached consensus that 21 studies met the criteria for inclusion. Methodological rigour was assessed using the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies. Random effects meta-analyses were used to compare the early-follicular, ovulatory and mid-luteal phases for maximal voluntary contraction, isokinetic peak torque, and explosive strength.

RESULTS

The assessment of study quality showed that a high level of bias exists in specific areas of study design. Non-significant and small or trivial effect sizes (p≥0.26, Hedges g≤0.35) were identified for all strength-related variables in each comparison between phases. 95% confidence intervals for each comparison suggested the uncertainty associated with each estimate extends to a small effect on strength performance with unclear direction (-0.42≤g≤0.48). The heterogeneity for each comparison was also small (p≥0.83, I²=0%).

CONCLUSIONS

Strength-related measures appear to be minimally altered (g≤0.35) by the fluctuations in ovarian sex hormones that occur during the menstrual cycle. This finding should be interpreted with caution due to the methodological shortcomings identified by the quality assessment.

Individual participant data meta-analysis to examine interactions between treatment effect and participant-level covariates: Statistical recommendations for conduct and planning

Precision medicine research often searches for treatment‐covariate interactions, which refers to when a treatment effect (eg, measured as a mean difference, odds ratio, hazard ratio) changes across values of a participant‐level covariate (eg, age, gender, biomarker). Single trials do not usually have sufficient power to detect genuine treatment‐covariate interactions, which motivate the sharing of individual participant data (IPD) from multiple trials for meta‐analysis. Here, we provide statistical recommendations for conducting and planning an IPD meta‐analysis of randomized trials to examine treatment‐covariate interactions. For conduct, two‐stage and one‐stage statistical models are described, and we recommend: (i) interactions should be estimated directly, and not by calculating differences in meta‐analysis results for subgroups; (ii) interaction estimates should be based solely on within‐study information; (iii) continuous covariates and outcomes should be analyzed on their continuous scale; (iv) nonlinear relationships should be examined for continuous covariates, using a multivariate meta‐analysis of the trend (eg, using restricted cubic spline functions); and (v) translation of interactions into clinical practice is nontrivial, requiring individualized treatment effect prediction. For planning, we describe first why the decision to initiate an IPD meta‐analysis project should not be based on between‐study heterogeneity in the overall treatment effect; and second, how to calculate the power of a potential IPD meta‐analysis project in advance of IPD collection, conditional on characteristics (eg, number of participants, standard deviation of covariates) of the trials (potentially) promising their IPD. Real IPD meta‐analysis projects are used for illustration throughout.

Population enrichment for critical care trials: phenotypes and differential outcomes

PURPOSE OF REVIEW

Sepsis and acute respiratory distress syndrome (ARDS) are two heterogenous acute illnesses where numerous RCTs have indeterminate results. We present a narrative review on the recent developments in enriching patient populations for future sepsis and ARDS trials.

RECENT FINDINGS

Many researchers are actively pursuing enrichment strategies to reduce heterogeneity to increase the sensitivity of future trials. Enrichment refers to the use of measurable patient characteristics, known before randomisation, to refine trial populations. Biomarkers could increase the diagnostic certainty of sepsis, whereas chest radiology training to enhance reliability of interpretation and stabilisation period of mechanical ventilation have been considered to increase the diagnostic certainty of ARDS. Clinical and biomarker data analyses identifies four to six sepsis clinical phenotypes and two ARDS clinical phenotypes. Similarly, leukocyte gene expression data identifies two to four sepsis molecular phenotypes. Use of a test-dose identifies ARDS subpopulations who are likely to benefit from higher PEEP. Early-phase trials report how a biomarker that is altered by the intervention, such as lymphocyte count for recombinant interleukin-7 therapy and higher check point inhibitor expression for anti-check point treatments in sepsis, could identify a higher treatment effect population for future trials.

SUMMARY

Enrichment reduces heterogeneity and will enhance the sensitivity of future trials. However, enrichment, even when it identifies more homogenous populations, may not be efficient to deploy in trials or clinical practice.

Enrichment designs using placebo nonresponders

Enrichment designs that select placebo nonresponders have gained much attention during the last years in areas with high placebo response rates, eg, in depression. Proposals were made that re-randomize patients who did not respond to placebo during a first study phase as the sequential parallel design (SPD). This design uses in a second phase an enriched patient population where the treatment effect is expected to be more pronounced. This may be problematic if an effect in the overall population is claimed. Proposals were made to combine the treatment effects in the overall population from study phase 1 and the enriched population from study phase 2, alleviating but not solving the issue of a potential selection bias. This paper shows how this bias corresponding to the effect difference between the overall population and the enriched population depends on the variability of a potential subject-by-treatment interaction. Sample sizes are given, which lead to a significant result in the combining test with a given probability if actually the average effect in the overall population is zero. If, on the other hand, no subject-by-treatment interaction is given, the enrichment is shown to be inefficient. We conclude that enrichment designs using placebo nonresponders are not able to claim a positive average effect in the overall population if a subject-by-treatment interaction cannot be excluded. It cannot be used to demonstrate positive efficacy in the overall population in a pivotal phase III trial but may be used in early phases to demonstrate varying treatment effects between patients.

What are the chances for personalised treatment with antidepressants? Detection of patient-by-treatment interaction with a variance ratio meta-analysis

OBJECTIVES

To investigate if the treatment effect of antidepressants in patients with depression substantially varies in each patient (patient-by-treatment interaction or treatment heterogeneity), a necessary but largely unexplored prerequisite of personalised antidepressant treatment.

DESIGN

Meta-analytic variance comparison of treatment outcome between drug arms and placebo arms of clinical trials, based on the assumption that patient-by-treatment interaction should lead to larger variances in drug arms than placebo arms. To put the results into context, we run simple simulations, assuming different definitions and rates of those who respond especially well to antidepressants.

DATA SOURCES

163 randomised, placebo-controlled trials (51 396 patients) with complete results for pre-post differences, selected from a recently published systematic review.

ANALYSIS

Variance ratios (VRs) and coefficients of variance ratios (CVRs) of individual trials were meta-analytically combined. The analysis was repeated for classes of antidepressants and specific antidepressants.

RESULTS

VRs (VR=1.01, CI 0.99 to 1.02) and CVRs (CVR=0.82, CI 0.80 to 0.84) of the antidepressant-treatment arms were comparable or smaller than in placebo arms. Similar results were observed for classes of antidepressants and for specific antidepressants. Our simulation analysis confirmed that equal VRs can only be obtained if they are not more than a few patients who respond slightly above average.

CONCLUSIONS

The lack of increased treatment-outcome variance in the antidepressants versus placebo groups in randomised controlled trials indicates that no or only very small subgroups of patients respond particularly well to antidepressants. Thus, the scope for personalised treatment with antidepressants seems to be limited.

Don’t abandon RCTs in IVF. We don’t even understand them

The conclusion of the Human Fertilisation and Embryology Authority that ‘add-on’ therapies in IVF are not supported by high-quality evidence has prompted new questions regarding the role of the randomized controlled trial (RCT) in evaluating infertility treatments. Critics argue that trials are cumbersome tools that provide irrelevant answers. Instead, they argue that greater emphasis should be placed on large observational databases, which can be analysed using powerful algorithms to determine which treatments work and for whom. Although the validity of these arguments rests upon the sciences of statistics and epidemiology, the discussion to date has largely been conducted without reference to these fields. We aim to remedy this omission, by evaluating the arguments against RCTs in IVF from a primarily methodological perspective. We suggest that, while criticism of the status quo is warranted, a retreat from RCTs is more likely to make things worse for patients and clinicians.

Exploration of associations between the FTO rs9939609 genotype, fasting and postprandial appetite-related hormones and perceived appetite in healthy men and women

BACKGROUND

The fat mass and obesity-associated gene (FTO) rs9939609 A-allele has been associated with obesity risk. Although the exact mechanisms involved remain unknown, the FTO rs9939609 A-allele has been associated with an impaired postprandial suppression of appetite.

OBJECTIVES

To explore the influence of FTO rs9939609 genotype on fasting and postprandial appetite-related hormones and perceived appetite in a heterogeneous sample of men and women.

DESIGN

112 healthy men and women aged 18-50-years-old completed three laboratory visits for the assessment of FTO rs9939609 genotype, body composition, aerobic fitness, resting metabolic rate, visceral adipose tissue, liver fat, fasting leptin, and fasting and postprandial acylated ghrelin, total PYY, insulin, glucose and perceived appetite. Participants wore accelerometers for seven consecutive days for the assessment of physical activity and sedentary behaviour. Multivariable general linear models quantified differences between FTO rs9939609 groups for fasting and postprandial appetite outcomes, with and without the addition of a priori selected physiological and behavioural covariates. Sex-specific univariable Pearson's correlation coefficients were quantified between the appetite-related outcomes and individual characteristics.

RESULTS

95% confidence intervals for mean differences between FTO rs9939609 groups overlapped zero in unadjusted and adjusted general linear models for all fasting (P ≥ 0.28) and postprandial (P ≥ 0.19) appetite-related outcomes. Eta² values for explained variance attributable to FTO rs9939609 were <5% for all outcomes. An exploratory correlation matrix indicated that associations between fasting and postprandial acylated ghrelin, total PYY and general or abdominal adiposity were also small (r = -0.23 to 0.15, P ≥ 0.09). Fasting leptin, glucose and insulin and postprandial insulin concentrations were associated with adiposity outcomes (r = 0.29 to 0.81, P ≤ 0.033).

CONCLUSIONS

Associations between the FTO rs9939609 genotype and fasting or postprandial appetite-related outcomes were weak in healthy men and women.

Cognitive functional therapy compared with a group-based exercise and education intervention for chronic low back pain: a multicentre randomised controlled trial (RCT)

Background

One-size-fits-all interventions reduce chronic low back pain (CLBP) a small amount. An individualised intervention called cognitive functional therapy (CFT) was superior for CLBP compared with manual therapy and exercise in one randomised controlled trial (RCT). However, systematic reviews show group interventions are as effective as one-to-one interventions for musculoskeletal pain. This RCT investigated whether a physiotherapist-delivered individualised intervention (CFT) was more effective than physiotherapist-delivered group-based exercise and education for individuals with CLBP.

Methods

206 adults with CLBP were randomised to either CFT (n=106) or group-based exercise and education (n=100). The length of the CFT intervention varied according to the clinical progression of participants (mean=5 treatments). The group intervention consisted of up to 6 classes (mean=4 classes) over 6–8 weeks. Primary outcomes were disability and pain intensity in the past week at 6 months and 12months postrandomisation. Analysis was by intention-to-treat using linear mixed models.

Results

CFT reduced disability more than the group intervention at 6 months (mean difference, 8.65; 95% CI 3.66 to 13.64; p=0.001), and at 12 months (mean difference, 7.02; 95% CI 2.24 to 11.80; p=0.004). There were no between-group differences observed in pain intensity at 6 months (mean difference, 0.76; 95% CI -0.02 to 1.54; p=0.056) or 12 months (mean difference, 0.65; 95% CI -0.20 to 1.50; p=0.134).

Conclusion

CFT reduced disability, but not pain, at 6 and 12 months compared with the group-based exercise and education intervention. Future research should examine whether the greater reduction in disability achieved by CFT renders worthwhile differences for health systems and patients.

Trial registration number

ClinicalTrials.gov registry (NCT02145728).

Evaluation of Differences in Individual Treatment Response in Schizophrenia Spectrum Disorders: A Meta-analysis

IMPORTANCE

An assumption among clinicians and researchers is that patients with schizophrenia vary considerably in their response to antipsychotic drugs in randomized clinical trials (RCTs).

OBJECTIVE

To evaluate the overall variation in individual treatment response from random variation by comparing the variability between treatment and control groups.

DATA SOURCES

Cochrane Schizophrenia, MEDLINE/PubMed, Embase, PsycINFO, Cochrane CENTRAL, BIOSIS Previews, ClinicalTrials.gov, and World Health Organization International Clinical Trials Registry Platform from January 1, 1955, to December 31, 2016.

STUDY SELECTION

Double-blind, placebo-controlled, RCTs of adults with a diagnosis of schizophrenia spectrum disorders and prescription for licensed antipsychotic drugs.

DATA EXTRACTION AND SYNTHESIS

Means and SDs of the Positive and Negative Syndrome Scale pretreatment and posttreatment outcome difference scores were extracted. Data quality and validity were ensured by following the PRISMA guidelines.

MAIN OUTCOMES AND MEASURES

The outcome measure was the overall variability ratio of treatment to control in a meta-analysis across RCTs. Individual variability ratios were weighted by the inverse-variance method and entered into a random-effects model. A personal element of response was hypothesized to be reflected by a substantial overall increase in variability in the treatment group compared with the control group.

RESULTS

An RCT was simulated, comprising 30 patients with schizophrenia randomized to either the treatment or the control group. The different components of variation in RCTs were illustrated with simulated data. In addition, we assessed the variability ratio in 52 RCTs involving 15 360 patients with a schizophrenia or schizoaffective diagnosis. The variability was slightly lower in the treatment compared with the control group (variability ratio = 0.97; 95% CI, 0.95-0.99; P = .01).

CONCLUSIONS AND RELEVANCE

In this study, no evidence was found in RCTs that antipsychotic drugs increased the outcome variance, suggesting no personal element of response to treatment but instead indicating that the variance was slightly lower in the treatment group than in the control group; although the study cannot rule out that subsets of patients respond differently to treatment, it suggests that the average treatment effect is a reasonable assumption for the individual patient.

Future of Personalized Cardiovascular Medicine: JACC State-of-the-Art Review

Previous decades have seen significant progress in the biological understanding of cardiovascular disease, as well as major advances in computational and information technologies. However, anticipated improvements in outcomes, quality, and cost of cardiovascular medicine at the individual and population levels from these advances have lagged expectations. Further, trends showing widening gaps in the pace of technological development and its successful uptake and application in practice suggests that substantial systemic changes are needed. Recent declines in key U.S. health outcomes have added further urgency to seek scalable approaches that deliver the right treatment to the right patient and to develop information-driven policies that improve health. The clinical care and research enterprises are currently in the midst of assimilating changes entrained by a "fourth industrial revolution" marked by the convergence of biology, physical sciences, and information science. These changes, if managed appropriately, can simultaneously enable cost-effective personalized medical care as well as more effective and targeted population health interventions. In this paper derived from a lecture in honor of cardiologist Paul Dudley White, the author explores how White's prescient insights into prevention and treatment continue to resonate today as we seek to assimilate ubiquitous computing, sophisticated sensor technologies, and bidirectional digital communication into the practice of cardiology. How the ongoing acceleration in basic science and information technologies can be wedded to the principles articulated by White as we pursue scalable approaches to personalized medicine is also examined.

How to deliver personalized cardiac resynchronization therapy through the precise measurement of the acute hemodynamic response: Insights from the iSpot trial

INTRODUCTION

New pacing technologies offer a greater choice of left ventricular pacing sites and greater personalization of cardiac resynchronization therapy (CRT). The effects on cardiac function of novel pacing configurations are often compared using multi-beat averages of acute hemodynamic measurements. In this analysis of the iSpot trial, we explore whether this is sufficient.

MATERIALS AND METHODS

The iSpot trial was an international, prospective, acute hemodynamic trial that assessed seven CRT configurations: standard CRT, MultiSpot (posterolateral vein), and MultiVein (anterior and posterior vein) pacing. Invasive and noninvasive blood pressure, and left ventricular (LV) dP/dtmax were recorded. Eight beats were recorded before and after an alternation from AAI to the tested pacing configuration and vice-versa. Eight alternations were performed for each configuration at each of the five atrioventricular delays.

RESULTS

Twenty-five patients underwent the full protocol of eight alternations. Only four (16%) patients had a statistically significant >3 mm Hg improvement over conventional CRT configuration (posterolateral vein, distal electrode). However, if only one alternation was analyzed (standard multi-beat averaging protocol), 15 (60%) patients falsely appeared to have a superior nonconventional configuration. Responses to pacing were significantly correlated between the different hemodynamic measures: invasive systolic blood pressure (SBP) vs noninvasive SBP r = 0.82 (P < .001); invasive SBP vs LV dP/dt r = 0.57, r2  = 0.32 (P < .001).

CONCLUSIONS

Current standard multibeat acquisition protocols are unfortunately unable to prevent false impressions of optimality arising in individual patients. Personalization processes need to include distinct repeated transitions to the tested pacing configuration in addition to averaging multiple beats. The need is not only during research stages but also during clinical implementation.

True Interindividual Variability Exists in Postprandial Appetite Responses in Healthy Men But Is Not Moderated by the FTO Genotype

BACKGROUND

After meal ingestion, a series of coordinated hormone responses occur concomitantly with changes in perceived appetite. It is not known whether interindividual variability in appetite exists in response to a meal.

OBJECTIVES

The aim of this study was to 1) assess the reproducibility of appetite responses to a meal; 2) quantify individual differences in responses; and 3) explore any moderating influence of the fat mass and obesity associated (FTO) gene.

METHODS

Using a replicated crossover design, 18 healthy men (mean ± SD age: 28.5 ± 9.8 y; BMI: 27.0 ± 5.0 kg/m2) recruited according to FTO genotype (9 AA, 9 TT) completed 2 identical control and 2 identical standardized meal conditions (5025 kJ) in randomized sequences. Perceived appetite and plasma acylated ghrelin, total peptide YY (PYY), insulin, and glucose concentrations were measured before and after interventions as primary outcomes. Interindividual differences were explored using Pearson's product-moment correlations between the first and second replicates of the control-adjusted meal response. Within-participant covariate-adjusted linear mixed models were used to quantify participant-by-condition and genotype-by-condition interactions.

RESULTS

The meal suppressed acylated ghrelin and appetite perceptions [standardized effect size (ES): 0.18-4.26] and elevated total PYY, insulin, and glucose (ES: 1.96-21.60). For all variables, SD of change scores was greater in the meal than in the control conditions. Moderate-to-large positive correlations were observed between the 2 replicates of control-adjusted meal responses for all variables (r = 0.44-0.86, P ≤ 0.070). Participant-by-condition interactions were present for all variables (P ≤ 0.056). FTO genotype-by-condition interactions were nonsignificant (P ≥ 0.19) and treatment effect differences between genotype groups were small (ES ≤ 0.27) for all appetite parameters.

CONCLUSIONS

Reproducibility of postprandial appetite responses is generally good. True interindividual variability is present beyond any random within-subject variation in healthy men but we detected no moderation by the FTO genotype. These findings highlight the importance of exploring individual differences in appetite for the prevention and treatment of obesity. This trial was registered at clinicaltrials.gov as NCT03771690.

Demonstrating Heterogeneity of Treatment Effects Among Patients: An Overlooked but Important Step Toward Precision Medicine

Although heterogeneity in the observed outcomes in clinical trials is often assumed to reflect a true heterogeneous response, it could actually be due to random variability. This retrospective analysis of four randomized, double-blind, placebo-controlled multiperiod (i.e., episode) crossover trials of fentanyl for breakthrough cancer pain illustrates the use of multiperiod crossover trials to examine heterogeneity of treatment response. A mixed-effects model, including fixed effects for treatment and episode and random effects for patient and treatment-by-patient interaction, was used to assess the heterogeneity in patients' responses to treatment during each episode. A significant treatment-by-patient interaction was found for three of four trials (P < 0.05), suggesting heterogeneity of the effect of fentanyl among different patients in each trial. Similar analyses in other therapeutic areas could identify conditions and therapies that should be investigated further for predictors of treatment response in efforts to maximize the efficiency of developing precision medicine strategies.

Issues in the determination of 'responders' and 'non-responders' in physiological research

NEW FINDINGS

What is the topic for this review? We discuss the dichotomization of continuous-level physiological measurements into 'responders' and 'non-responders' when interventions/treatments are examined in robust parallel-group studies. What advances does it highlight? Sample responder counts are biased by pre-to-post within-subject variability. Sample differences in counts may be explained wholly by differences in mean response, even without individual response heterogeneity and even if test-retest measurement error informs the choice of response threshold. A less biased and more informative approach uses the SD of individual responses to estimate the chance a new person from the population of interest will be a responder.

ABSTRACT

As a follow-up to our 2015 review, we cover more issues on the topic of 'response heterogeneity', which we define as clinically important individual differences in the physiological responses to the same treatment/intervention that cannot be attributed to random within-subject variability. We highlight various pitfalls with the common practice of counting the number of 'responders', 'non-responders' and 'adverse responders' in samples that have been given certain treatments or interventions for research purposes. We focus on the classical parallel-group randomized controlled trial and assume typical good practice in trial design. We show that sample responder counts are biased because individuals differ in terms of pre-to-post within-subject random variability in the study outcome(s) and not necessarily treatment response. Ironically, sample differences in responder counts may be explained wholly by sample differences in mean response, even if there is no response heterogeneity at all. Sample comparisons of responder counts also have relatively low statistical precision. These problems do not depend on how the response threshold has been selected, e.g. on the basis of a measurement error statistic, and are not rectified fully by the use of confidence intervals for individual responses in the sample. The dichotomization of individual responses in a research sample is fraught with pitfalls. Less biased approaches for estimating the proportion of responders in a population of interest are now available. Importantly, these approaches are based on the SD for true individual responses, directly incorporating information from the control group.

Plastic diagnostics: The remaking of disease and evidence in personalized medicine

Politically authorized reports on personalized and precision medicine stress an urgent need for finer-grained disease categories and faster taxonomic revision, through integration of genomic and phenotypic data. Developing a data-driven taxonomy is, however, not as simple as it sounds. It is often assumed that an integrated data infrastructure is relatively easy to implement in countries that already have highly centralized and digitalized health care systems. Our analysis of initiatives associated with the Danish National Genome Center, recently launched to bring Denmark to the forefront of personalized medicine, tells a different story. Through a “meta-taxonomy” of taxonomic revisions, we discuss what a genomics-based disease taxonomy entails, epistemically as well as organizationally. Whereas policy reports promote a vision of seamless data integration and standardization, we highlight how the envisioned strategy imposes significant changes on the organization of health care systems. Our analysis shows how persistent tensions in medicine between variation and standardization, and between change and continuity, remain obstacles for the production as well as the evaluation of genomics-based taxonomies of difference. We identify inherent conflicts between the ideal of dynamic revision and existing regulatory functions of disease categories in, for example, the organization and management of health care systems. Moreover, we raise concerns about shifts in the regulatory regime of evidence standards, where clinical care increasingly becomes a vehicle for biomedical research.

•

Personalized medicine has led to calls for speedy revisions of disease taxonomies.

•

A “meta-taxonomy” of disease taxonomic revisions is presented and discussed.

•

Fine-graining disease categories has regulatory implications for healthcare systems.

•

A Danish case illustrates difficulties in aligning multiple functions of disease codes.

•

We argue that the political call for speed should be treated with caution.

Heterogeneity of treatment effect by baseline risk of mortality in critically ill patients: re-analysis of three recent sepsis and ARDS randomised controlled trials

Background

Randomised controlled trials (RCTs) enrolling patients with sepsis or acute respiratory distress syndrome (ARDS) generate heterogeneous trial populations. Non-random variation in the treatment effect of an intervention due to differences in the baseline risk of death between patients in a population represents one form of heterogeneity of treatment effect (HTE). We assessed whether HTE in two sepsis and one ARDS RCTs could explain indeterminate trial results and inform future trial design.

Methods

We assessed HTE for vasopressin, hydrocortisone and levosimendan in sepsis and simvastatin in ARDS patients, on 28-day mortality, using the total Acute Physiology And Chronic Health Evaluation II (APACHE II) score as the baseline risk measurement, comparing above (high) and below (low) the median score. Secondary risk measures were the acute physiology component of APACHE II and predicted risk of mortality using the APACHE II score. HTE was quantified both in additive (difference in risk difference (RD)) and multiplicative (ratio of relative risks (RR)) scales using estimated treatment differences from a logistic regression model with treatment risk as the interaction term.

Results

The ratio of the odds of death in the highest APACHE II quartile was 4.9 to 7.4 times compared to the lowest quartile, across the three trials. We did not observe HTE for vasopressin, hydrocortisone and levosimendan in the two sepsis trials. In the HARP-2 trial, simvastatin reduced mortality in the low APACHE II group and increased mortality in the high APACHE II group (difference in RD = 0.34 (0.12, 0.55) (p = 0.02); ratio of RR 3.57 (1.77, 7.17) (p < 0.001). The HTE patterns were inconsistent across the secondary risk measures. The sensitivity analyses of HTE effects for vasopressin, hydrocortisone and levosimendan were consistent with the main analyses and attenuated for simvastatin.

Conclusions

We assessed HTE in three recent ICU RCTs, using multivariable baseline risk of death models. There was considerable within-trial variation in the baseline risk of death. We observed potential HTE for simvastatin in ARDS, but no evidence of HTE for vasopressin, hydrocortisone or levosimendan in the two sepsis trials. Our findings could be explained either by true lack of HTE (no benefit of vasopressin, hydrocortisone or levosimendan vs comparator for any patient subgroups) or by lack of power to detect HTE. Our results require validation using similar trial databases.

Electronic supplementary material

The online version of this article (10.1186/s13054-019-2446-1) contains supplementary material, which is available to authorized users.

Structural similarity networks predict clinical outcome in early-phase psychosis

Despite recent advances, there is still a major need for prediction of treatment success in schizophrenia, a condition long considered a disorder of dysconnectivity in the brain. Graph theory provides a means to characterize the connectivity in both healthy and abnormal brains. We calculated structural similarity networks in each participant and hypothesized that the "hubness", i.e., the number of edges connecting a node to the rest of the network, would be associated with clinical outcome. This prospective controlled study took place at an academic research center and included 82 early-phase psychosis patients (23 females; mean age [SD] = 21.6 [5.5] years) and 58 healthy controls. Medications were administered in a double-blind randomized manner, and patients were scanned at baseline prior to treatment with second-generation antipsychotics. Symptoms were assessed with the Brief Psychiatric Rating Scale at baseline and over the course of 12 weeks. Nodal degree of structural similarity networks was computed for each subject and entered as a predictor of individual treatment response into a partial least squares (PLS) regression. The model fit was significant in a permutation test with 1000 permutations (P = 0.006), and the first two PLS regression components explained 29% (95% CI: 27; 30) of the variance in treatment response after cross-validation. Nodes loading strongly on the first PLS component were primarily located in the orbito- and prefrontal cortex, whereas nodes loading strongly on the second PLS component were primarily located in the superior temporal, precentral, and middle cingulate cortex. These data suggest a link between brain network morphology and clinical outcome in early-phase psychosis.

Precision exercise medicine: understanding exercise response variability

There is evidence from human twin and family studies as well as mouse and rat selection experiments that there are considerable interindividual differences in the response of cardiorespiratory fitness (CRF) and other cardiometabolic traits to a given exercise programme dose. We developed this consensus statement on exercise response variability following a symposium dedicated to this topic. There is strong evidence from both animal and human studies that exercise training doses lead to variable responses. A genetic component contributes to exercise training response variability.

In this consensus statement, we (1) briefly review the literature on exercise response variability and the various sources of variations in CRF response to an exercise programme, (2) introduce the key research designs and corresponding statistical models with an emphasis on randomised controlled designs with or without multiple pretests and post-tests, crossover designs and repeated measures designs, (3) discuss advantages and disadvantages of multiple methods of categorising exercise response levels—a topic that is of particular interest for personalised exercise medicine and (4) outline approaches that may identify determinants and modifiers of CRF exercise response. We also summarise gaps in knowledge and recommend future research to better understand exercise response variability.

Protect us from poor-quality medical research

Much of the published medical research is apparently flawed, cannot be replicated and/or has limited or no utility. This article presents an overview of the current landscape of biomedical research, identifies problems associated with common study designs and considers potential solutions. Randomized clinical trials, observational studies, systematic reviews and meta-analyses are discussed in terms of their inherent limitations and potential ways of improving their conduct, analysis and reporting. The current emphasis on statistical significance needs to be replaced by sound design, transparency and willingness to share data with a clear commitment towards improving the quality and utility of clinical research.

Interindividual Responses of Appetite to Acute Exercise: A Replicated Crossover Study

PURPOSE

Acute exercise transiently suppresses appetite, which coincides with alterations in appetite-regulatory hormone concentrations. Individual variability in these responses is suspected, but replicated trials are needed to quantify them robustly. We examined the reproducibility of appetite and appetite-regulatory hormone responses to acute exercise and quantified the individual differences in responses.

METHODS

Fifteen healthy, recreationally active men completed two control (60-min resting) and two exercise (60-min fasted treadmill running at 70% peak oxygen uptake) conditions in randomized sequences. Perceived appetite and circulating concentrations of acylated ghrelin and total peptide YY (PYY) were measured immediately before and after the interventions. Interindividual differences were explored by correlating the two sets of response differences between exercise and control conditions. Within-participant covariate-adjusted linear mixed models were used to quantify participant-condition interactions.

RESULTS

Compared with control, exercise suppressed mean acylated ghrelin concentrations and appetite perceptions (all ES = 0.62-1.47, P < 0.001) and elevated total PYY concentrations (ES = 1.49, P < 0.001). For all variables, the standard deviation of the change scores was substantially greater in the exercise versus control conditions. Moderate-to-large positive correlations were observed between the two sets of control-adjusted exercise responses for all variables (r = 0.54-0.82, P ≤ 0.036). After adjusting for baseline measurements, participant-condition interactions were present for all variables (P ≤ 0.053).

CONCLUSIONS

Our replicated crossover study allowed, for the first time, the interaction between participant and acute exercise response in appetite parameters to be quantified. Even after adjustment for individual baseline measurements, participants demonstrated individual differences in perceived appetite and hormone responses to acute exercise bouts beyond any random within-subject variability over time.

Statistical Considerations for Exercise Protocols Aimed at Measuring Trainability

The individual response to exercise training is of great interest with methods that have been proposed to measure this response reviewed in this paper. However, individual training response estimates may be biased by various sources of variability present in exercise studies, and in particular by within-subject variability. We propose the use of protocols that can separate trainability from within-subject variability.

The Precision HYpertenSIon Care (PHYSIC) study: a double-blind, randomized, repeated cross-over study

High blood pressure is the leading risk factor for premature deaths and a major cost to societies worldwide. Effective blood pressure-lowering drugs are available, but patient adherence to them is low, likely partly due to side effects. To identify patient-specific differences in treatment effects, a repeated cross-over design, where the same treatment contrasts are repeated within each patient, is needed. Such designs have been surprisingly rarely used, given the current focus on precision medicine. The Precision HYpertenSIon Care (PHYSIC) study aims to investigate if there is a consistent between-person variation in blood pressure response to the common blood pressure-lowering drug classes of a clinically relevant magnitude, given the within-person variation in blood pressure. The study will also investigate the between-person variation in side effects of the drugs. In a double-blind, randomized, repeated cross-over trial, 300 patients with mild hypertension will be treated with four blood pressure-lowering drugs (candesartan, lisinopril, amlodipine, and hydrochlorothiazide) in monotherapy, with two of the drugs repeated for each patient. If the study indicates that there is a potential for precision hypertension care, the most promising predictors of blood pressure and side effect response to the drugs will be explored, as will the potential for development of a biomarker panel to rank the suitability of blood pressure-lowering drug classes for individual patients in terms of anticipated blood pressure effects and side effects, with the ultimate goal to maximize adherence. The study follows a protocol pre-registered at ClinicalTrials.gov with the identifier NCT02774460.

Identifying treatment responders using counterfactual modeling and potential outcomes

Individualizing treatment according to patients' characteristics is central for personalized or precision medicine. There has been considerable recent research in developing statistical methods to determine optimal personalized treatment strategies by modeling the outcome of patients according to relevant covariates under each of the alternative treatments, and then relying on so-called predicted individual treatment effects. In this paper, we use potential outcomes and principal stratification frameworks and develop a multinomial model for left and right-censored data to estimate the probability that a patient is a responder given a set of baseline covariates. The model can apply to RCT or observational study data. This method is based on the monotonicity assumption, which implies that no patients would respond to the control treatment but not to the experimental one. We conduct a simulation study to evaluate the properties of the proposed estimation method. Results showed that the predictions of the probability of being a responder were well calibrated even if we observed variability and a small bias when many parameters were estimated. We finally applied the method to a cohort study on the selection of patients for additional radiotherapy after resection of a soft-tissue sarcoma.

How to Construct, Conduct and Analyze an Exercise Training Study?

Randomized controlled trials (RCTs) can be regarded as gold standard in investigating dose-response and causal relationships in exercise science. Recommendations for exercise training routines and efficacy analyses of certain training regimen require valid data derived from robust RCTs. Moreover, meta-analyses rely on RCTs and both RCTs and meta-analyses are considered the highest level of scientific evidence. Beyond general study design a variety of methodological aspects and notable pitfalls has to be considered. Therefore, exercise training studies should be carefully constructed focusing on the consistency of the whole design "package" from an explicit hypothesis or research question over study design and methodology to data analysis and interpretation. The present scoping review covers all main aspects of planning, conducting, and analyzing exercise based RCTs. We aim to focus on relevant aspects regarding study design, statistical power, training planning and documentation as well as traditional and recent statistical approaches. We intend to provide a comprehensive hands-on paper for conceptualizing future exercise training studies and hope to stimulate and encourage researchers to conduct sound and valid RCTs in the field of exercise training.

Inter-Individual Responses of Maximal Oxygen Uptake to Exercise Training: A Critical Review

It has recently been reported how to quantify inter-individual differences in the response to an exercise intervention using the standard deviation of the change scores, as well as how to appraise these differences for clinical relevance. In a parallel-group randomised controlled trial, the key trigger for further investigation into inter-individual responses is when the standard deviation of change in the intervention sample is substantially larger than the same standard deviation derived from a suitable comparator sample. 'True' and clinically relevant inter-individual differences in response can then be plausibly expected, and potential moderators and mediators of the inter-individual differences can be explored. We now aim to critically review the research on the inter-individual differences in response to exercise training, focusing on maximal oxygen uptake (VO₂max). A literature search through the relevant bibliographic databases resulted in the identification of six relevant studies that were published prior to the influential HEalth, RIsk factors, exercise Training And GEnetics (HERITAGE) Family Study. Only one of these studies was found to include a comparator arm. Re-analysis of the data from this study, accounting for random within-subjects variation, revealed an absence of clinically important inter-individual differences in the response of VO₂max to exercise training. The standard deviation of change was, in fact, larger (±5.6 mL/kg/min) for the comparator than the intervention group (±3.7 mL/kg/min). We located over 180 publications that resulted from the HERITAGE Family Study, but we could not find a comparator arm in any of these studies. Some authors did not explain this absence, while others reasoned that only inter-individual differences in exercise response were of interest, thus the intervention sample was investigated solely. We also found this absence of a comparator sample in on-going studies. A perceived high test-retest reliability is offered as a justification for the absence of a comparator arm, but the test-retest reliability analysis for the HERITAGE Family Study was over a much shorter term than the length of the actual training period between baseline and follow-up measurements of VO₂max. We also scrutinised the studies in which twins have been investigated, resulting in concerns about how genetic influences on the magnitude of general within-subjects variability has been partitioned out (again in the absence of a comparator no-training group), as well as with the intra-class correlation coefficient approach to data analysis. Twin pairs were found to be sometimes heterogeneous for the obviously influential factors of sex, age and fitness, thereby inflating an unadjusted coefficient. We conclude that most studies on inter-individual differences in VO₂max response to exercise training have no comparator sample. Therefore, true inter-individual differences in response cannot be quantified, let alone appraised for clinical relevance. For those studies with a comparator sample, we found that the inter-individual differences in training response were not larger than random within-subjects variation in VO₂max over the same time period as the training intervention.

Repeated testing for the assessment of individual response to exercise training

Observed response to regular exercise training differs widely between individuals even in tightly controlled research settings. However, the respective contributions of random error and true interindividual differences as well as the relative frequency of nonresponders are disputed. Specific challenges of analyses on the individual level as well as a striking heterogeneity in definitions may partly explain these inconsistent results. Repeated testing during the training phase specifically addresses the requirements of analyses on the individual level. Here we report a first implementation of this innovative design amendment in a head-to-head comparison of existing analytical approaches. To allow for comparative implementation of approaches we conducted a controlled endurance training trial (1 yr walking/jogging, 3 days/wk for 45 min with 60% heart rate reserve) in healthy, untrained subjects ( n = 36, age = 46 ± 8 yr; body mass index 24.7 ± 2.7 kg/m2; V̇o2max 36.6 ± 5.4). In the training group additional V̇o2max tests were conducted after 3, 6, and 9 mo. Duration of the control condition was 6 mo due to ethical constraints. General efficacy of the training intervention could be verified by a significant increase in V̇o2max in the training group ( P < 0.001 vs. control). Individual training response of relevant magnitude (>0.2 × baseline variability in V̇o2max) could be demonstrated by several approaches. Regarding the classification of individuals, only 11 of 20 subjects were consistently classified, demonstrating remarkable disagreement between approaches. These results are in support of relevant interindividual variability in training efficacy and stress the limitations of a responder classification. Moreover, this proof-of-concept underlines the need for tailored methodological approaches for well-defined problems.

NEW & NOTEWORTHY This work reports a first implementation of a repeated testing training trial for the investigation of individual response. This design amendment was recently proposed to address specifically the statistical requirements of analyses on the individual level. Moreover, a comprehensive comparison of previously published methods exemplifies the striking heterogeneity of existing approaches.

Lessons learned from IDeAl - 33 recommendations from the IDeAl-net about design and analysis of small population clinical trials

BACKGROUND

IDeAl (Integrated designs and analysis of small population clinical trials) is an EU funded project developing new statistical design and analysis methodologies for clinical trials in small population groups. Here we provide an overview of IDeAl findings and give recommendations to applied researchers.

METHOD

The description of the findings is broken down by the nine scientific IDeAl work packages and summarizes results from the project's more than 60 publications to date in peer reviewed journals. In addition, we applied text mining to evaluate the publications and the IDeAl work packages' output in relation to the design and analysis terms derived from in the IRDiRC task force report on small population clinical trials.

RESULTS

The results are summarized, describing the developments from an applied viewpoint. The main result presented here are 33 practical recommendations drawn from the work, giving researchers a comprehensive guidance to the improved methodology. In particular, the findings will help design and analyse efficient clinical trials in rare diseases with limited number of patients available. We developed a network representation relating the hot topics developed by the IRDiRC task force on small population clinical trials to IDeAl's work as well as relating important methodologies by IDeAl's definition necessary to consider in design and analysis of small-population clinical trials. These network representation establish a new perspective on design and analysis of small-population clinical trials.

CONCLUSION

IDeAl has provided a huge number of options to refine the statistical methodology for small-population clinical trials from various perspectives. A total of 33 recommendations developed and related to the work packages help the researcher to design small population clinical trial. The route to improvements is displayed in IDeAl-network representing important statistical methodological skills necessary to design and analysis of small-population clinical trials. The methods are ready for use.

Response heterogeneity: Challenges for personalised medicine and big data approaches in psychiatry and chronic pain

Response rates to available treatments for psychological and chronic pain disorders are poor, and there is a considerable burden of suffering and disability for patients, who often cycle through several rounds of ineffective treatment. As individuals presenting to the clinic with symptoms of these disorders are likely to be heterogeneous, there is considerable interest in the possibility that different constellations of signs could be used to identify subgroups of patients that might preferentially benefit from particular kinds of treatment. To this end, there has been a recent focus on the application of machine learning methods to attempt to identify sets of predictor variables (demographic, genetic, etc.) that could be used to target individuals towards treatments that are more likely to work for them in the first instance. Importantly, the training of such models generally relies on datasets where groups of individual predictor variables are labelled with a binary outcome category - usually 'responder' or 'non-responder' (to a particular treatment). However, as previously highlighted in other areas of medicine, there is a basic statistical problem in classifying individuals as 'responding' to a particular treatment on the basis of data from conventional randomized controlled trials. Specifically, insufficient information on the partition of variance components in individual symptom changes mean that it is inappropriate to consider data from the active treatment arm alone in this way. This may be particularly problematic in the case of psychiatric and chronic pain symptom data, where both within-subject variability and measurement error are likely to be high. Here, we outline some possible solutions to this problem in terms of dataset design and machine learning methodology, and conclude that it is important to carefully consider the kind of inferences that particular training data are able to afford, especially in arenas where the potential clinical benefit is so large.

Response heterogeneity: Challenges for personalised medicine and big data approaches in psychiatry and chronic pain

Response rates to available treatments for psychological and chronic pain disorders are poor, and there is a substantial burden of suffering and disability for patients, who often cycle through several rounds of ineffective treatment. As individuals presenting to the clinic with symptoms of these disorders are likely to be heterogeneous, there is considerable interest in the possibility that different constellations of signs could be used to identify subgroups of patients that might preferentially benefit from particular kinds of treatment. To this end, there has been a recent focus on the application of machine learning methods to attempt to identify sets of predictor variables (demographic, genetic, etc.) that could be used to target individuals towards treatments that are more likely to work for them in the first instance. Importantly, the training of such models generally relies on datasets where groups of individual predictor variables are labelled with a binary outcome category - usually 'responder' or 'non-responder' (to a particular treatment). However, as previously highlighted in other areas of medicine, there is a basic statistical problem in classifying individuals as 'responding' to a particular treatment on the basis of data from conventional randomized controlled trials. Specifically, insufficient information on the partition of variance components in individual symptom changes mean that it is inappropriate to consider data from the active treatment arm alone in this way. This may be particularly problematic in the case of psychiatric and chronic pain symptom data, where both within-subject variability and measurement error are likely to be high. Here, we outline some possible solutions to this problem in terms of dataset design and machine learning methodology, and conclude that it is important to carefully consider the kind of inferences that particular training data are able to afford, especially in arenas where the potential clinical benefit is so large.

Does evidence support the high expectations placed in precision medicine? A bibliographic review

Background: Precision medicine is the Holy Grail of interventions that are tailored to a patient's individual characteristics. However, conventional clinical trials are designed to find differences in averages, and interpreting these differences depends on untestable assumptions. Although only an ideal, a constant effect of treatment would facilitate individual management. A direct consequence of a constant effect is that the variance of the outcome measure would be the same in the treated and control arms. We reviewed the literature to explore the similarity of these variances as a foundation for examining whether and how often precision medicine is definitively required. Methods: We reviewed parallel clinical trials with numerical primary endpoints published in 2004, 2007, 2010 and 2013. We collected the baseline and final standard deviations of the main outcome measure. We assessed homoscedasticity by comparing the variance of the primary endpoint between arms through the outcome variance ratio (treated to control group). Results: The review provided 208 articles with enough information to conduct the analysis. One out of five studies (n = 40, 19.2%) had statistically different variances between groups, implying a non-constant-effect. The adjusted point estimate of the mean outcome variance ratio (treated to control group) is 0.89 (95% CI 0.81 to 0.97). Conclusions: The mean variance ratio is significantly lower than 1 and the lower variance was found more often in the intervention group than in the control group, suggesting it is more usual for treated patients to be stable. This observed reduction in variance might also imply that there could be a subgroup of less ill patients who derive no benefit from treatment. This would require further study as to whether the treatment effect outweighs the side effects as well as the economic costs. We have shown that there are ways to analyze the apparently unobservable constant effect.