Best (but oft forgotten) practices: sample size planning for powerful studies

Best (but oft forgotten) practices: Efficient sample sizes for commonly used trial designs

Designing studies such that they have a high level of power to detect an effect or association of interest is an important tool to improve the quality and reproducibility of findings from such studies. Since resources (research subjects, time, and money) are scarce, it is important to obtain sufficient power with minimum use of such resources. For commonly used randomized trials of the treatment effect on a continuous outcome, designs are presented that minimize the number of subjects or the amount of research budget when aiming for a desired power level. This concerns the optimal allocation of subjects to treatments and, in case of nested designs such as cluster-randomized trials and multicenter trials, also the optimal number of centers versus the number of persons per center. Since such optimal designs require knowledge of parameters of the analysis model that are not known in the design stage, in particular outcome variances, maximin designs are presented. These designs guarantee a prespecified power level for plausible ranges of the unknown parameters and minimize research costs for the worst-case values of these parameters. The focus is on a 2-group parallel design, the AB/BA crossover design, and cluster-randomized and multicenter trials with a continuous outcome. How to calculate sample sizes for maximin designs is illustrated for examples from nutrition. Several computer programs that are helpful in calculating sample sizes for optimal and maximin designs are discussed as well as some results on optimal designs for other types of outcomes.

Input modality matters: A comparison of touch, speech, and gesture based in-vehicle interaction

Innovative input devices are being available for in-vehicle information systems (IVISs). While they have the potential to provide enjoyable driving by enabling drivers to perform non-driving related tasks (NDRTs) in more natural ways, the associated distracting effects should be paid with more attention. The purpose of this exploratory study was to compare the effects of three novel input modalities, i.e., touchscreen-based interaction (TBI), speech-based interaction (SBI), and gesture-based interaction (GBI), on driving performance and driver visual behaviors. Moreover, we examined if the influence of different modalities would be moderated by the difficulty level of NDRTs. A total of 36 participants were invited to a simulated driving experiment where they were randomly assigned to one of the four groups (TBI, GBI, SBI or baseline) and completed three driving trials. The results showed that TBI led to the worse driving performance, as indicated by the significantly prolonged reaction time, reduced minimum time-to-collision, and increased variations in both longitudinal and lateral vehicle control. The deteriorated driving performance could be attributed, at least partially, to the intense visual demand induced by looking towards the touchscreen, as indicated by more and longer off-the-road glances. The adverse impacts of GBI were relatively smaller, but it still posed great crash risk by leading to a shorter minimum time-to-collision and less stable vehicle control compared to the baseline. SBI, although not completely equivalent to the baseline group, showed the minimum influence on driving and visual performance. Only very few interaction effects were found, suggesting that the effects of modality were quite robust across different NDRTs. It was concluded that SBI and GBI provided safer alternatives to in-vehicle interaction than TBI.

Strengthening Research that Answers Nutrition Questions of Public Health Importance: Leveraging the Experience of the USDA Nutrition Evidence Systematic Review Team

The Nutrition Evidence Systematic Review (NESR) team conducts nutrition- and public health-related systematic reviews and is within the USDA's Center for Nutrition Policy and Promotion. NESR has collaborated with scientific experts to conduct systematic reviews on nutrition and public health topics for more than a decade and is uniquely positioned to share recommendations with the research community to strengthen research quality and impact, especially the evidence base that supports public health nutrition guidance, including future editions of the Dietary Guidelines for Americans. Leveraging the expertise of NESR and its systematic review process resulted in the following recommendations for the research community: a) use the strongest study design feasible with sufficient sample size(s); b) enroll study participants who reflect the diversity of the population of interest and report participant characteristics; c) use valid and reliable dietary assessment methods; d) describe the interventions or exposures of interest and use standard definitions to promote consistency; e) use valid and reliable health outcome measures; f) account for variables that may impact the relationship between nutrition-related interventions or exposures and health outcomes; g) carry out studies for a sufficient duration and include repeated measures, as appropriate; and h) report all relevant information to inform accurate interpretation and evaluation of study results. Implementing these recommendations can strengthen nutrition and public health evidence and increase its utility in future public health nutrition systematic reviews. However, implementation will require additional support from the entire research community, including scientific journals and funding agencies.

Do you see what I see? The influence of self-objectification on appearance anxiety, intrinsic motivation, interoceptive awareness, and physical performance

Self-objectification negatively impacts body image, mental health, and cognitive performance outcomes in women and has been correlated with poorer physical performance. The purpose of this study was to determine if trying on a swimsuit (versus a sweater) impacted state self-objectification, body shame, social physique anxiety, intrinsic motivation, bodily awareness, and physical performance in university women. Female undergraduate students (N = 52) were randomly assigned to try on either a swimsuit (objectification condition) or sweater (control condition) and complete measures of self-objectification, body shame, appearance anxiety, intrinsic motivation, and interoceptive awareness, and perform a series of balance tasks. Women in the swimsuit condition reported higher state self-objectification, body-related shame and appearance anxiety, and lower intrinsic motivation compared to women in the sweater condition. In addition, women in the swimsuit condition restricted body movements during a 1-legged stand balance task. Consistent with objectification theory, women may have made smaller physical movements in an attempt to hide or cover up the body. Findings could have implications for promoting positive experiences during physical activity for women, such as in sport, exercise or rehabilitation settings.

Power(ful) myths: misconceptions regarding sample size in quality of life research

PURPOSE

Carefully selecting the sample size for a research study is one of the most fundamental ways to utilize resources in an ethical manner, maximize impact and replicability, and minimize research waste when investigating questions relevant to health-related quality of life (HRQOL). Despite an increased focus on sample size in the methodological literature, the topic has received limited attention in the HRQOL field, and there are still misconceptions that can weaken even well-intentioned sample size planning. This article aims to highlight common misconceptions, provide accessible and non-technical corrections to these misconceptions, and show how HRQOL researchers can benefit from a more nuanced understanding of sample size planning.

METHOD

Misconceptions were identified broadly through examples within the health, psychology, and HRQOL literatures. In examining these misconceptions, study-level (e.g., missing data, multilevel designs, multiple reported outcomes) and field-level (e.g., publication bias, replicability) issues relevant to HRQOL research were considered.

RESULTS

Misconceptions include: (a) researchers should use rules of thumb or the largest sample size possible, (b) sample size planning should always focus on power, (c) planned power = actual power, (d) there is only one level of power per study, and (e) power is only relevant for the individual researcher. Throughout the article, major themes linked to these misconceptions are mapped onto recent HRQOL studies to make the connections more tangible.

CONCLUSION

By clarifying several challenges and misconceptions regarding sample size planning and statistical power, HRQOL researchers will have the tools needed to augment the research literature in effective and meaningful ways.

Common Limitations and Challenges of Dietary Clinical Trials for Translation into Clinical Practices

The progressive development of clinical and public health nutrition has long relied on dietary clinical trials (DCTs), investigating the causal relationship between diet and multiple risk factors of non-communicable and chronic diseases. DCTs are also hallmarks for establishing dietary requirements and promoting overall nutritional health among the population. Despite their critical importance in translation into public health strategies and practices, DCTs have several limitations and challenges for study design, implementation and finding interpretation. The complex nature of nutrition interventions, collinearity between diet components, multi-target effects of the interventions, diverse dietary behaviors, and food culture are the most challenging issues. Furthermore, baseline exposure and dietary status, appropriate control groups, blinding, randomization, and poor adherence undermine the effectiveness of DCTs in translation into practices. Disruptive factors will be minimized if researchers are committed to following good clinical practice (GCP) standards available for common designs of clinical trials. Planning DCTs, however, needs careful considerations for hypothesis generation, study design development, the definition of primary and secondary outcome measures, and target population.

Clinician's Guide to Understanding Effect Size, Alpha Level, Power, and Sample Size

Effect size, α level, power, and sample size are misunderstood concepts that play a major role in the design and interpretation of studies. Effect size represents the magnitude of a change in an outcome or the strength of a relationship. Often, the effect size may be more important than just relying on the α level when interpreting results from a study because it informs a researcher of the actual magnitude of the difference or relationship. Confidence intervals can also assist in making this assessment. Power represents the probability of rejecting the null hypothesis when it is false. Effect size, power, and α level are all important in the calculation of sample size needed to conduct a study. Calculating the sample size a priori and including the required number of participants is essential. Studies with small sample sizes, relative to the needed number provided from a power analysis, may lead to false negative results. Studies with grossly large sample sizes may yield statistically significant findings with small effect sizes that may not be clinically relevant. It is beneficial when all four components are clearly presented in analytic studies. Failure to include these elements may limit the ability of other researchers to replicate the study's findings and lead to difficulty when interpreting the study's results.

DHA intake relates to better cerebrovascular and neurodegeneration neuroimaging phenotypes in middle-aged adults at increased genetic risk of Alzheimer disease

BACKGROUND

The number of APOE-ε4 alleles is a major nonmodifiable risk factor for sporadic Alzheimer disease (AD). There is increasing evidence on the benefits of dietary DHA (22:6n-3) before the onset of AD symptoms, particularly in APOE-ε4 carriers. Brain alterations in the preclinical stage can be detected by structural MRI.

OBJECTIVES

We aimed, in middle-aged cognitively unimpaired individuals at increased risk of AD, to cross-sectionally investigate whether dietary DHA intake relates to cognitive performance and to MRI-based markers of cerebral small vessel disease and AD-related neurodegeneration, exploring the effect modification by APOE-ε4 status.

METHODS

In 340 participants of the ALFA (ALzheimer and FAmilies) study, which is enriched for APOE-ε4 carriership (n = 122, noncarriers; n = 157, 1 allele; n = 61, 2 alleles), we assessed self-reported DHA intake through an FFQ. We measured cognitive performance by administering episodic memory and executive function tests. We performed high-resolution structural MRI to assess cerebral small vessel disease [white matter hyperintensities (WMHs) and cerebral microbleeds (CMBs)] and AD-related brain atrophy (cortical thickness in an AD signature). We constructed regression models adjusted for potential confounders, exploring the interaction DHA × APOE-ε4.

RESULTS

We observed no significant associations between DHA and cognitive performance or WMH burden. We observed a nonsignificant inverse association between DHA and prevalence of lobar CMBs (OR: 0.446; 95% CI: 0.195, 1.018; P = 0.055). DHA was found to be significantly related to greater cortical thickness in the AD signature in homozygotes but not in nonhomozygotes (P-interaction = 0.045). The association strengthened when analyzing homozygotes and nonhomozygotes matched for risk factors.

CONCLUSIONS

In cognitively unimpaired APOE-ε4 homozygotes, dietary DHA intake related to structural patterns that may result in greater resilience to AD pathology. This is consistent with the current hypothesis that those subjects at highest risk would obtain the largest benefits from DHA supplementation in the preclinical stage.This trial was registered at clinicaltrials.gov as NCT01835717.