Why carve up your continuous data?

Tumor-Specific Activity of Precision Medicines in the NCI-MATCH Trial

PURPOSE

National Cancer Institute Molecular Analysis for Therapy Choice (NCI-MATCH) is a precision medicine basket trial designed to test the effectiveness of treating cancers based on specific genetic changes in patients' tumors, regardless of cancer type. Multiple subprotocols have each tested different targeted therapies matched to specific genetic aberrations. Most subprotocols exhibited low rates of tumor shrinkage as evaluated across all tumor types enrolled. We hypothesized that these results may arise because these precision cancer therapies have tumor type-specific efficacy, as is common among other cancer therapies.

EXPERIMENTAL DESIGN

To test the hypothesis that certain tumor types are more sensitive to specific therapies than other tumor types, we applied permutation testing to tumor volume change and progression-free survival data from 10 published NCI-MATCH subprotocols (together n = 435 patients). FDR was controlled by the Benjamini-Hochberg procedure.

RESULTS

Six of ten subprotocols exhibited statistically significant evidence of tumor-specific drug sensitivity, four of which were previously considered negative based on response rate across all tumors. This signal-finding analysis highlights potential uses of FGFR tyrosine kinase inhibition in urothelial carcinomas with actionable FGFR aberrations and MEK inhibition in lung cancers with BRAF non-V600E mutations. In addition, it identifies low-grade serious ovarian carcinoma with BRAF v600E mutation as especially sensitive to BRAF and MEK co-inhibition (dabrafenib plus trametinib), a treatment that received accelerated FDA approval for advanced solid tumors with BRAF v600E mutation.

CONCLUSIONS

These findings support the value of basket trials because even when precision medicines do not have tumor-agnostic activity, basket trials can identify tumor-specific activity for future study.

Researchers in rheumatology should avoid categorization of continuous predictor variables

BACKGROUND

Rheumatology researchers often categorize continuous predictor variables. We aimed to show how this practice may alter results from observational studies in rheumatology.

METHODS

We conducted and compared the results of two analyses of the association between our predictor variable (percentage change in body mass index [BMI] from baseline to four years) and two outcome variable domains of structure and pain in knee and hip osteoarthritis. These two outcome variable domains covered 26 different outcomes for knee and hip combined. In the first analysis (categorical analysis), percentage change in BMI was categorized as ≥ 5% decrease in BMI, < 5% change in BMI, and ≥ 5% increase in BMI, while in the second analysis (continuous analysis), it was left as a continuous variable. In both analyses (categorical and continuous), we used generalized estimating equations with a logistic link function to investigate the association between the percentage change in BMI and the outcomes.

RESULTS

For eight of the 26 investigated outcomes (31%), the results from the categorical analyses were different from the results from the continuous analyses. These differences were of three types: 1) for six of these eight outcomes, while the continuous analyses revealed associations in both directions (i.e., a decrease in BMI had one effect, while an increase in BMI had the opposite effect), the categorical analyses showed associations only in one direction of BMI change, not both; 2) for another one of these eight outcomes, the categorical analyses suggested an association with change in BMI, while this association was not shown in the continuous analyses (this is potentially a false positive association); 3) for the last of the eight outcomes, the continuous analyses suggested an association of change in BMI, while this association was not shown in the categorical analyses (this is potentially a false negative association).

CONCLUSIONS

Categorization of continuous predictor variables alters the results of analyses and could lead to different conclusions; therefore, researchers in rheumatology should avoid it.

Druggable transcriptomic pathways revealed in Parkinson's patient-derived midbrain neurons

Complex genetic predispositions accelerate the chronic degeneration of midbrain substantia nigra neurons in Parkinson’s disease (PD). Deciphering the human molecular makeup of PD pathophysiology can guide the discovery of therapeutics to slow the disease progression. However, insights from human postmortem brain studies only portray the latter stages of PD, and there is a lack of data surrounding molecular events preceding the neuronal loss in patients. We address this gap by identifying the gene dysregulation of live midbrain neurons reprogrammed in vitro from the skin cells of 42 individuals, including sporadic and familial PD patients and matched healthy controls. To minimize bias resulting from neuronal reprogramming and RNA-seq methods, we developed an analysis pipeline integrating PD transcriptomes from different RNA-seq datasets (unsorted and sorted bulk vs. single-cell and Patch-seq) and reprogramming strategies (induced pluripotency vs. direct conversion). This PD cohort’s transcriptome is enriched for human genes associated with known clinical phenotypes of PD, regulation of locomotion, bradykinesia and rigidity. Dysregulated gene expression emerges strongest in pathways underlying synaptic transmission, metabolism, intracellular trafficking, neural morphogenesis and cellular stress/immune responses. We confirmed a synaptic impairment with patch-clamping and identified pesticides and endoplasmic reticulum stressors as the most significant gene-chemical interactions in PD. Subsequently, we associated the PD transcriptomic profile with candidate pharmaceuticals in a large database and a registry of current clinical trials. This study highlights human transcriptomic pathways that can be targeted therapeutically before the irreversible neuronal loss. Furthermore, it demonstrates the preclinical relevance of unbiased large transcriptomic assays of reprogrammed patient neurons.

Improving power in PSA response analyses of metastatic castration-resistant prostate cancer trials

BACKGROUND

To determine how much an augmented analysis approach could improve the efficiency of prostate-specific antigen (PSA) response analyses in clinical practice. PSA response rates are commonly used outcome measures in metastatic castration-resistant prostate cancer (mCRPC) trial reports. PSA response is evaluated by comparing continuous PSA data (e.g., change from baseline) to a threshold (e.g., 50% reduction). Consequently, information in the continuous data is discarded. Recent papers have proposed an augmented approach that retains the conventional response rate, but employs the continuous data to improve precision of estimation.

METHODS

A literature review identified published prostate cancer trials that included a waterfall plot of continuous PSA data. This continuous data was extracted to enable the conventional and augmented approaches to be compared.

RESULTS

Sixty-four articles, reporting results for 78 mCRPC treatment arms, were re-analysed. The median efficiency gain from using the augmented analysis, in terms of the implied increase to the sample size of the original study, was 103.2% (IQR [89.8,190.9%]).

CONCLUSIONS

Augmented PSA response analysis requires no additional data to be collected and can be performed easily using available software. It improves precision of estimation to a degree that is equivalent to a substantial sample size increase. The implication of this work is that prostate cancer trials using PSA response as a primary endpoint could be delivered with fewer participants and, therefore, more rapidly with reduced cost.

Cut points and contexts

In research, policy, and practice, continuous variables are often categorized. Statisticians have generally advised against categorization for many reasons, such as loss of information and precision as well as distortion of estimated statistics. Here, a different kind of problem with categorization is considered: the idea that, for a given continuous variable, there is a unique set of cut points that is the objectively correct or best categorization. It is shown that this is unlikely to be the case because categorized variables typically exist in webs of statistical relationships with other variables. The choice of cut points for a categorized variable can influence the values of many statistics relating that variable to others. This essay explores the substantive trade‐offs that can arise between different possible cut points to categorize a continuous variable, making it difficult to say that any particular categorization is objectively best. Limitations of different approaches to selecting cut points are discussed. Contextual trade‐offs may often be an argument against categorization. At the very least, such trade‐offs mean that research inferences, or decisions about policy or practice, that involve categorized variables should be framed and acted upon with flexibility and humility.

In practical settings, the choice of cut points for categorizing a continuous variable is likely to entail trade‐offs across multiple statistical relationships between the categorized variable and other variables. These trade‐offs mean that no single categorization is objectively best or correct.

Improving drug safety predictions by reducing poor analytical practices

Predicting the safety of a drug from preclinical data is a major challenge in drug discovery, and progressing an unsafe compound into the clinic puts patients at risk and wastes resources. In drug safety pharmacology and related fields, methods and analytical decisions known to provide poor predictions are common and include creating arbitrary thresholds, binning continuous values, giving all assays equal weight, and multiple reuse of information. In addition, the metrics used to evaluate models often omit important criteria and models’ performance on new data are often not assessed rigorously. Prediction models with these problems are unlikely to perform well, and published models suffer from many of these issues. We describe these problems in detail, demonstrate their negative consequences, and propose simple solutions that are standard in other disciplines where predictive modelling is used.

Analysis of responder-based endpoints: improving power through utilising continuous components

BACKGROUND

Clinical trials and other studies commonly assess the effectiveness of an intervention through the use of responder-based endpoints. These classify patients based on whether they meet a number of criteria which often involve continuous variables categorised as being above or below a threshold. The proportion of patients who are responders is estimated and, where relevant, compared between groups. An alternative method called the augmented binary method keeps the definition of the endpoint the same but utilises information contained within the continuous component to increase the power considerably (equivalent to increasing the sample size by > 30%). In this article we summarise the method and investigate the variety of clinical conditions that use endpoints to which it could be applied.

METHODS

We reviewed a database of core outcome sets (COSs) that covered physiological and mortality trial endpoints recommended for collection in clinical trials of different disorders. We identified responder-based endpoints where the augmented binary method would be useful for increasing power.

RESULTS

Out of the 287 COSs reviewed, we identified 67 new clinical areas where endpoints were used that would be more efficiently analysed using the augmented binary method. Clinical areas that had particularly high numbers were rheumatology (11 clinical disorders identified), non-solid tumour oncology (10 identified), neurology (9 identified) and cardiovascular (8 identified).

CONCLUSIONS

The augmented binary method can potentially provide large benefits in a vast array of clinical areas. Further methodological development is needed to account for some types of endpoints.

Correlates of type 2 diabetes and glycaemic control in adults in Saudi Arabia a secondary data analysis of the Saudi health interview survey

BACKGROUND

There is evidence that type 2 diabetes self-management programmes may have a positive impact on health outcomes of adults living in Gulf countries. However, none of the programmes evaluated were developed using evidence about the specific needs of adults with Type 2 diabetes living in the Gulf countries. This study is part of a wider programme of research, which uses a cultural adaptation framework to generate information on how to tailor type 2 diabetes self-management to the Saudi context.

METHODS

Secondary data analysis of the Saudi Health Interview Survey (SHIS) (N = 10,821) was conducted. Bivariate and multivariate logistic regression modelling assessed factors associated with type 2 diabetes and its control / self-management including sociodemographic factors (e.g. age, gender), lifestyle (e.g. diet, physical activity), and health seeking behaviours (e.g. chronic illnesses, health services).

RESULTS

7% (N = 808) of all participants had type 2 diabetes (59% male), however it represents 35% at or above 55 years. In multivariate analysis at older age, being overweight or obese, male, having hypertension, and reporting a reduction in health status in the 12 months prior to questionnaire completion, were significantly associated with having type 2 diabetes. Participants who reported walking for more than 10 min per day were less likely to report type 2 diabetes. Unexpectedly there was a significant association between type 2 diabetes and lower frequency of fast food intake, while increased fruit and vegetable intake was associated with poor glycaemic control.

CONCLUSIONS

Being overweight and/or hypertensive are concomitant with type 2 diabetes in Saudi Arabia. Any self-management programmes for type 2 diabetes patients with either of these conditions should be tailored accordingly. Walking behaviours should be prioritised in Saudi self-management programmes. Prediabetes management programmes may be of special importance to the Saudi community.

Four simple ways to increase power without increasing the sample size

Underpowered experiments have three problems: true effects are harder to detect, the true effects that are detected tend to have inflated effect sizes and as power decreases so does the probability that a statistically significant result represents a true effect. Many biology experiments are underpowered and recent calls to change the traditional 0.05 significance threshold to a more stringent value of 0.005 will further reduce the power of the average experiment. Increasing power by increasing the sample size is often the only option considered, but more samples increases costs, makes the experiment harder to conduct and is contrary to the 3Rs principles for animal research. We show how the design of an experiment and some analytical decisions can have a surprisingly large effect on power.

Perceptions of perioperative nursing competence: a cross-country comparison

Background

Throughout many countries, professional bodies rely on yearly self-assessment of competence for ongoing registration; therefore, nursing competence is pivotal to safe clinical practice. Our aim was to describe and compare perioperative nurses’ perceptions of competence in four countries, while examining the effect of specialist education and years of experience in the operating room.

Methods

We conducted a secondary analysis of cross-sectional surveys from four countries including; Australia, Canada, Scotland, and Sweden. The 40-item Perceived Perioperative Competence Scale-Revised (PPCS-R), was used with a total sample of 768 respondents. We used a factorial design to examine the influence of country, years of experience in the operating room and specialist education on nurses’ reported perceived perioperative competence.

Results

Regardless of country origin, nurses with specialist qualifications reported higher perceived perioperative competence when compared to nurses without specialist education. However, cross-country differences were dependent on nurses’ number of years of experience in the operating room. Nurses from Sweden with 6–10 years of experience in the operating room reported lower perceived perioperative competence when compared to Australian nurses. In comparing nurses with > 10 years of experience, Swedish nurses reported significantly lower perceived perioperative competence when compared to nurses from Australia, Canada and Scotland.

Conclusion

Researchers need to consider educational level and years of experience in the perioperative context when examining constructs such as competence.

Dental caries is negatively correlated with body mass index among 7-9 years old children in Guangzhou, China

BACKGROUND

Evidence linking caries in primary dentition and children's anthropometric measures is contradictory. We aimed to evaluate the prevalence of primary dental caries and its relationship with body mass index (BMI) among 7-9 years old school children in urban Guangzhou, China.

METHODS

This cross-sectional study enrolled 32,461 pupils (14,778 girls and 17,683 boys) aged 7-9 years from 65 elementary schools in Guangzhou. Dental caries was detected according to criteria recommended by the World Health Organization (WHO). The total mean decayed, missing or filled teeth (dmft) of primary dentition were assessed. Weight and height were measured and BMI was calculated. Children were classified into underweight, normal weight, overweight and obesity groups by BMI based on Chinese criteria. Z-score of BMI-for-age (BAZ) was calculated by WHO standardized procedure. Multivariable odds ratios (ORs) and 95 % confidence intervals (CIs) were calculated using logistic regression. Restricted cubic spline regression was applied to evaluate the shape of the relationship between BAZ and primary dental caries.

RESULTS

The prevalence of primary dental caries was 30.7 % in total sample. Regarding dmft values, the mean ± standard deviation (SD) in the combined sample were 1.03 ± 2.05 in boys and 0.93 ± 1.92 in girls. Both indices decreased by age. Compared with normal BMI group, children in overweight and obesity groups have 27 % (OR = 0.73, 95 % CI: 0.66-0.81, P < 0.0001) and 34 % (OR = 0.66, 95 % CI: 0.59-0.74, P < 0.0001) lower odds for the presence of primary dental caries after adjustment for age and gender, respectively. Although in general, increased BAZ was associated with decreased risk of dental caries, full-range BAZ was associated with dental caries in an A-shaped manner with a zenith at around -1.4.

CONCLUSION

Higher BMI was associated with lower odds of caries; overweight and obese children were more likely to be primary dental caries free among 7-9 years in Guangzhou, China.

Ranking, selecting, and prioritising genes with desirability functions

In functional genomics experiments, researchers often select genes to follow-up or validate from a long list of differentially expressed genes. Typically, sharp thresholds are used to bin genes into groups such as significant/non-significant or fold change above/below a cut-off value, and ad hoc criteria are also used such as favouring well-known genes. Binning, however, is inefficient and does not take the uncertainty of the measurements into account. Furthermore, p-values, fold-changes, and other outcomes are treated as equally important, and relevant genes may be overlooked with such an approach. Desirability functions are proposed as a way to integrate multiple selection criteria for ranking, selecting, and prioritising genes. These functions map any variable to a continuous 0–1 scale, where one is maximally desirable and zero is unacceptable. Multiple selection criteria are then combined to provide an overall desirability that is used to rank genes. In addition to p-values and fold-changes, further experimental results and information contained in databases can be easily included as criteria. The approach is demonstrated with a breast cancer microarray data set. The functions and an example data set can be found in the desiR package on CRAN (https://cran.r-project.org/web/packages/desiR/) and the development version is available on GitHub (https://github.com/stanlazic/desiR).

An inverse U-shaped association of late and peak insulin levels during an oral glucose load with glucose intolerance in a Japanese population: a cross-sectional study

The current study investigated the association of post-load insulin levels with glucose tolerance in a Japanese population. A total of 1450 Japanese employees who underwent a 75-g oral glucose tolerance test (OGTT) were included. Glucose tolerance was assessed by 120-min glucose levels during a 75-g OGTT. A penalized cubic regression spline model analysis revealed that the 60- and 120-min insulin levels, but not 0- or 30-min insulin levels, had an inverse U-shaped relationship to the 120-min glucose level. Furthermore, peak insulin level followed an inverse U shape in relation to the 120-min glucose level, whereas the peak of insulin appeared at a later point in time as the 120-min glucose level increased. These associations were similarly observed in both obese and non-obese subgroups, although obesity was associated with higher insulin levels. Peak insulin levels also demonstrated an inverse U shape in association with 0-min glucose levels and indices of β cell function, assessed by the disposition index and the β-cell function index. In conclusion, peak insulin levels followed an inverse U shape in relation to glucose intolerance in a Japanese population, whereas the impairment of glucose tolerance was associated with a delay in the time to reach peak insulin levels.

How alike are young non-drinkers, former-drinkers, low-risk drinkers, and hazardous drinkers?

The aim of this study was to examine whether - in terms of personality characteristics and beliefs--former-drinkers and people who have never consumed alcohol exist on a continuum that includes low-risk drinkers and hazardous drinkers, or are distinct groups. An online questionnaire hosted on a secure server was completed by 1418 young people (642 men and 776 women) aged 16-21 living in South-East England. Participants' scores on the Alcohol Use Disorders Identification Test (AUDIT) were used to classify them as non-drinkers, former-drinkers, low-risk drinkers or hazardous drinkers. Multinomial logistic regression identified eight significant multivariate correlates that explained 39% of the variance in men's AUDIT category membership (χ(2)(24) = 263.32, p < .01), and seven significant multivariate correlates that explained 41% of the variance in women's AUDIT category membership (χ(2)(21) = 332.91, p < .01). The results suggest that non-drinkers and former-drinkers are more similar to each other than they are to both low risk and hazardous drinkers. We should not, therefore, treat these groups of young people as representing a drinking continuum. In particular, interventions for high risk young drinkers may be more effective if distinguished from general campaigns intended for all young people.

Psychological distress and emergency department utilization in the United States: evidence from the Medical Expenditure Panel Survey

OBJECTIVES

Psychological distress not only has substantial health and social consequences, but is also associated with emergency department (ED) use. Previous studies have typically used cross-sectional data to focus on the relation between serious psychological distress and dichotomized ED utilization measures, without assessing the volume of ED use or examining nonserious levels of psychological distress. The objective of this study was to explore the association between ED utilization volume and the full spectrum of psychological distress.

METHODS

Data from Panel 14 of the Medical Expenditure Panel Survey (MEPS; 2009-2010, n = 9,743) provided a nationally representative sample of U.S. individuals. ED utilization volume and three specifications of the Kessler Psychological Distress Scale (K6) were analyzed: a dichotomous serious/no serious psychological distress measure, a five-category ordinal measure, and a scale measure with a range of 0 to 24. Negative binomial-logit hurdle regression models were used to analyze how the different specifications of the K6 psychological distress measure were related to ED use.

RESULTS

Adults with serious psychological distress in 2009 had 1.59 (95% confidence interval [CI] = 1.15 to 2.20) times greater adjusted odds of having one or more ED visits in 2010 than those without serious psychological distress. Nonserious psychological distress levels in 2009 were also associated with increased adjusted odds of having at least one ED visit in 2010. The K6 scores showed a dose-response relationship in terms of the adjusted odds of having one or more ED visits. The adjusted odds ratios (ORs) were 1.86 (95% CI = 1.37 to 2.54) for adults with K6 scores at or above 11, OR 1.76 (95% CI = 1.38 to 2.25) for adults with K6 scores between 6 and 10, OR 1.33 (95% CI = 1.05 to 1.68) for adults with K6 scores between 3 and 5, and OR 1.17 (95% CI = 0.92 to 1.48) for adults with K6 scores of 1 or 2. In addition, the adjusted odds of having one or more ED visits in 2010 significantly increased with increasing psychological distress in 2009 (OR = 1.04, 95% CI = 1.03 to 1.06). Each additional point added to the K6 scale results in an increase in the adjusted odds of an ED visit.

CONCLUSIONS

Even a low level of psychological distress, and not just serious psychological distress, may be an early indicator of future ED use. These results highlight the need to develop novel responses to better manage or avert ED use not only for adults with serious psychological distress but also for those who are experiencing even mild symptoms of psychological distress.

Dental caries and growth in school-age children

BACKGROUND AND OBJECTIVE

Dental caries in young children is commonly untreated and represents a public health problem. Dental caries in children is reported to affect their anthropometric outcomes, but the evidence is conflicting. Some studies found no association, whereas others found that caries was associated with underweight or overweight. The objective was to assess the relationship between dental caries status and height and weight in 6- to 8-year-old Saudi children with high caries prevalence.

METHODS

This study was a cross-sectional survey in schoolchildren aged 6 to 8 years attending military primary schools in Jeddah, Saudi Arabia. Caries status was assessed by using the dmft (decayed, missing, filled, teeth [primary teeth]) index. Height and weight were assessed by using z scores of height-for-age (HAZ), weight-for-age (WAZ), and BMI-for-age (BAZ) calculated by World Health Organization standardized procedures. Relationships between caries and HAZ, WAZ, and BAZ were assessed by using regression models.

RESULTS

A total of 417 of the 436 eligible schoolchildren with complete data were included, with a response rate of 95.6%. Their mean dmft index was 5.7 ± 4.2. There was an inverse linear relationship between caries status and children's HAZ, WAZ, and BAZ and significantly lower anthropometric outcomes for children at each consecutive group with higher levels of caries. The associations remained significant after adjusting for dental, social, and demographic variables.

CONCLUSIONS

The inverse linear association between dental caries and all anthropometric outcomes suggests that higher levels of untreated caries are associated with poorer growth in Saudi schoolchildren.

Effects of pain controllability and discrepancy in social support on depressed mood among patients with chronic pain

BACKGROUND

Preference of types of social support may vary across recipients, and social support that is less than the amount preferred may be associated with depressed mood.

PURPOSE

This study aimed to investigate the interaction effects between pain controllability and discrepancy in social support and the additive utility of discrepancy in social support over perceived social support in predicting depressed mood among patients with chronic pain.

METHOD

A total of 173 patients seeking treatment at two outpatient pain management clinics in Hawaii participated.

RESULTS

The results indicated that (1) patients with low pain controllability preferred significantly more social support than those with high pain controllability, for all types of social support; (2) patients preferred significantly more informational and emotional support than instrumental support for both pain controllability levels; (3) discrepancy in informational or instrumental support was not associated with depressed mood, whereas discrepancy in emotional support was significantly associated with more depressed mood for low pain controllability, but not for high pain controllability; and (4) discrepancy in emotional support added a significant increment of variance in predicting depressed mood over and above perceived emotional support, whereas discrepancy in informational or instrumental support did not.

CONCLUSION

Findings suggest the relative importance of discrepancy in emotional support from a significant other, especially for patients with low pain controllability.

The relationship between self-efficacy for behaviors that promote healthy weight and clinical indicators of adiposity in a sample of early adolescents

OBJECTIVE

Examine the relationship between self-efficacy and various measures of adiposity in a sample of teens.

METHODS

A total of 132 teens were selected from schools participating in an existing research study titled Teen Eating and Activity Mentoring in Schools (TEAMS). Teens completed demographic questionnaires and healthy eating-specific and physical activity-specific measures of self-efficacy. Waist circumference (WC), triceps skinfold thickness (TSF), and body mass index (BMI) percentile scores were also obtained.

RESULTS

Regression analyses indicated that healthy eating-specific and physical activity-specific measures of self-efficacy predicted WC and TSF. ANOVA revealed significant differences in healthy eating-specific self-efficacy levels between students of recommended weight and overweight/obese status. Supplemental analyses showed significant negative relationships between a student's ideal BMI ratio and self-efficacy.

CONCLUSIONS

Because self-efficacy may be amenable to change, these findings could inform future efforts aimed at increasing behaviors that promote healthy weight status among early adolescents.

Pooling dietary data using questionnaires with open-ended and predefined responses: implications for comparing mean intake or estimating odds ratios

In the current era of diet-gene analyses, large sample sizes are required to uncover the etiology of complex diseases. As such, consortia form and often combine available data. Food frequency questionnaires, which commonly use 2 different types of responses about the frequency of intake (predefined responses and open-ended responses), may be pooled to achieve the desired sample size. The common practice is to categorize open-ended responses into the predefined response categories. A problem arises when the predefined categories are noncontiguous: possible open-ended responses may fall in gaps between the predefined categories. Using simulated data modeled from frequency of intake among 1,664 controls in a lung cancer case-control study at The University of Texas M. D. Anderson Cancer Center (Houston, Texas, 2000-2005), the authors describe the effect of different categories of open-ended responses that fall in between noncontiguous, predefined response sets on estimates of the mean difference in intake and the odds ratios. A significant inflation of false positives appears when comparing mean differences of intake, while the bias in estimating odds ratios may be acceptably small. Therefore, if pooling data cannot be restricted to the same type of response, inferences should focus on odds ratio estimation to minimize bias.

Comparing two small samples with an unstable, treatment-independent baseline

Due to time and resource constraints, small samples (N=3-7 cases per group) are often used in neurobiological studies that employ multiple techniques. In a simulation study, five statistical tests were used to compare two small samples (treated and control) with an unstable, additive baseline. These five tests differed in the way that they used the baseline variable (B) to adjust or normalize the variable affected by the treatment (Y). We conclude that, if N=3 or 4, the independent t-test on Y-B tends to have the highest power; if N> or =7, ANCOVA on Y with B as the covariate tends to have the highest power; and both tests have comparably high power if N=5 or 6. The Wilcoxon rank-sum test (or, equivalently, the Mann-Whitney test) has precisely zero power if one group has 3 cases and the other has 3 or 4 cases. Some other problems of small-sample analysis are considered.

Why we should use simpler models if the data allow this: relevance for ANOVA designs in experimental biology

Background

Analysis of variance (ANOVA) is a common statistical technique in physiological research, and often one or more of the independent/predictor variables such as dose, time, or age, can be treated as a continuous, rather than a categorical variable during analysis – even if subjects were randomly assigned to treatment groups. While this is not common, there are a number of advantages of such an approach, including greater statistical power due to increased precision, a simpler and more informative interpretation of the results, greater parsimony, and transformation of the predictor variable is possible.

Results

An example is given from an experiment where rats were randomly assigned to receive either 0, 60, 180, or 240 mg/L of fluoxetine in their drinking water, with performance on the forced swim test as the outcome measure. Dose was treated as either a categorical or continuous variable during analysis, with the latter analysis leading to a more powerful test (p = 0.021 vs. p = 0.159). This will be true in general, and the reasons for this are discussed.

Conclusion

There are many advantages to treating variables as continuous numeric variables if the data allow this, and this should be employed more often in experimental biology. Failure to use the optimal analysis runs the risk of missing significant effects or relationships.

Concern about petrochemical health risk before and after a refinery explosion

On March 23, 2005, a large explosion at an oil refinery in Texas City, Texas caused 15 deaths and approximately 170 injuries. Little is known about how such an industrial accident influences concern about environmental health risks. We used measures of environmental health concern about nearby petrochemical production with a sample of Texas City residents to understand patterns of concern and change in concern after an industrial accident, as well as individual and contextual factors associated with those patterns. Survey interviews with residents of Texas City, Texas (N= 315) both pre- and postexplosion using a brief Concern About Petrochemical Health Risk Scale (CAPHRS) and other questions were used to collect pertinent predictor information. CAPHRS baseline, postexplosion, and change scores were compared and modeled using ordinary least squares (OLS) regression and a mixed model. Higher preexplosion CAPHRS scores were predicted by younger adults, foreign-born Hispanics, non-Hispanic blacks, lower- and middle-income groups, and those who live with someone who has worked at the petrochemical plants. Higher CAPHRS change scores are predicted by the same variables (except income), as well as proximity to, or perception of, the explosion, and reports of neighborhood damage. Findings suggest these groups' concern scores could indicate a greater vulnerability to psychological and physical harm generated by concern and stress arising from local petrochemical activities. A clearer understanding of concern about actual environmental health risks in exposed populations may enhance the evolving theory of stress and coping and eventually enable public health professionals to develop appropriate mitigation strategies.

Do illustrations enhance preschoolers’ memories for stories? Age-related change in the picture facilitation effect

This study investigated whether illustrations facilitate story recall in preschoolers (N=58) 46 to 63 months of age. Each child was exposed to either a verbal story narrative with illustrations (Verbal and Picture condition), the narrative alone (Verbal Only condition), the narrative with uninformative illustrations (Verbal and Irrelevant Picture condition), or the illustrations alone (Picture Only condition). Children recalled the story immediately and after a 1-week delay. With increasing age, the Verbal and Picture group increasingly outperformed the other verbal groups, particularly when recalling central details conveyed by both narrative and illustrations. Performance did not vary among the three verbal groups at lower values of age, whereas the Verbal and Picture group recalled more than the other verbal groups at the older ages. Regardless of age, all verbal groups outperformed the Picture Only group, suggesting that preschoolers are unlikely to spontaneously generate a narrative from pictures alone.