Response to lithium in patients with bipolar disorder is associated with clinical and transdiagnostic genetic factors. The predictive combination of these variables might help clinicians better predict which patients will respond to lithium treatment.

To use a combination of transdiagnostic genetic and clinical factors to predict lithium response in patients with bipolar disorder.

This study utilised genetic and clinical data (n = 1034) collected as part of the International Consortium on Lithium Genetics (ConLi+Gen) project. Polygenic risk scores (PRS) were computed for schizophrenia and major depressive disorder, and then combined with clinical variables using a cross-validated machine-learning regression approach. Unimodal, multimodal and genetically stratified models were trained and validated using ridge, elastic net and random forest regression on 692 patients with bipolar disorder from ten study sites using leave-site-out cross-validation. All models were then tested on an independent test set of 342 patients. The best performing models were then tested in a classification framework.

The best performing linear model explained 5.1% (P = 0.0001) of variance in lithium response and was composed of clinical variables, PRS variables and interaction terms between them. The best performing non-linear model used only clinical variables and explained 8.1% (P = 0.0001) of variance in lithium response. A priori genomic stratification improved non-linear model performance to 13.7% (P = 0.0001) and improved the binary classification of lithium response. This model stratified patients based on their meta-polygenic loadings for major depressive disorder and schizophrenia and was then trained using clinical data.

Using PRS to first stratify patients genetically and then train machine-learning models with clinical predictors led to large improvements in lithium response prediction. When used with other PRS and biological markers in the future this approach may help inform which patients are most likely to respond to lithium treatment.

MD-PhD training programs train physician-scientists to pursue careers involving both clinical care and research, but decreasing numbers of physician-scientists stay engaged in clinical research. We sought to identify current clinical research training methods utilized by MD–PhD programs and to assess how effective they are in promoting self-efficacy for clinical research.

The US MD–PhD students were surveyed in April–May 2018. Students identified the clinical research training methods they participated in, and self-efficacy in clinical research was determined using a modified 12-item Clinical Research Appraisal Inventory.

Responses were received from 61 of 108 MD–PhD institutions. Responses were obtained from 647 MD–PhD students in all years of training. The primary methods of clinical research training included no clinical research training, and various combinations of didactics, mentored clinical research, and a clinical research practicum. Students with didactics plus mentored clinical research had similar self-efficacy as those with didactics plus clinical research practicum. Training activities that differentiated students who did and did not have the clinical research practicum experience and were associated with higher self-efficacy included exposure to Institutional Review Boards and participation in human subject recruitment.

A clinical research practicum was found to be an effective option for MD–PhD students conducting basic science research to gain experience in clinical research skills. Clinical research self-efficacy was correlated with the amount of clinical research training and specific clinical research tasks, which may inform curriculum development for a variety of clinical and translational research training programs, for example, MD–PhD, TL1, and KL2.

To verify the previously untested assumption that eating more salad enhances vegetable intake and determine if salad consumption is in fact associated with higher vegetable intake and greater adherence to the Dietary Guidelines for Americans (DGA) recommendations.

Individuals were classified as salad reporters or non-reporters based upon whether they consumed a salad composed primarily of raw vegetables on the intake day. Regression analyses were applied to calculate adjusted estimates of food group intakes and assess the likelihood of meeting Healthy US-Style Food Pattern recommendations by salad reporting status.

Cross-sectional analysis of data collected in 2011–2014 in What We Eat in America, the dietary intake component of the National Health and Nutrition Examination Survey.

US adults (n 9678) aged ≥20 years (excluding pregnant and lactating women).

On the intake day, 23 % of adults ate salad. The proportion of individuals reporting salad varied by sex, age, race, income, education and smoking status (P<0·001). Compared with non-reporters, salad reporters consumed significantly larger quantities of vegetables (total, dark green, red/orange and other), which translated into a two- to threefold greater likelihood of meeting recommendations for these food groups. More modest associations were observed between salad consumption and differences in intake and likelihood of meeting recommendations for protein foods (total and seafood), oils and refined grains.

Study results confirm the DGA message that incorporating more salads in the diet is one effective strategy (among others, such as eating more cooked vegetables) to augment vegetable consumption and adherence to dietary recommendations concerning vegetables.

Estimates of fruit and vegetable (FV) consumption vary depending on intake definition, which may be determined by research purpose. Researchers have used two methods to evaluate intake: epidemiological and behavioural. The present study describes FV intake by adults using epidemiological v. behavioural approaches.

One-day dietary intake data from What We Eat In America, National Health and Nutrition Examination Survey 2009–2012 were used. Sample weights were used to produce nationally representative estimates. FV intake (in cup-equivalents (CE)) was estimated using the Food Patterns Equivalents Database. The epidemiological method considered all FV after disaggregating foods and beverages. The behavioural method included foods that provided at least 0·2 CE FV per 100 g, and excluded sources high in fat, added sugar and Na.

Nationally representative survey of the US population.

Adults (n 10 563) aged ≥20 years.

For epidemiological v. behavioural, fruit intake was 1·1 v. 1·0 CE for males and 1·0 v. 0·9 CE for females. Vegetable intake was 1·8 v. 1·1 CE for males and 1·5 v. 1·0 CE for females.

The definition of FV intake affects estimates of consumption by the population and is an important consideration when planning and comparing research studies. The method used should align with research goals to assure accurate interpretation and validity of results.

To provide updated estimates of drinking water intake (total, tap, plain bottled) for groups aged ≥1 year in the USA and to determine whether intakes collected in 2005–2006 using the Automated Multiple-Pass Method for the 24 h recall differ from intakes collected in 2003–2004 via post-recall food-frequency type questions.

Cross-sectional, observational study.

What We Eat in America (WWEIA), the dietary intake component of the US National Health and Nutrition Examination Survey (NHANES).

Individuals aged ≥1 year in 2003–2004 (n 8249) and 2005–2006 (n 8437) with one complete 24 h recall.

The estimate for the percentage of individuals who reported total drinking water in 2005–2006 was significantly (P < 0·0000) smaller (76·9 %) than that for 2003–2004 (87·1 %), attributable to a lower percentage reporting tap water (54·1 % in 2005–2006 v. 67·0 % in 2003–2004; P = 0·0001). Estimates of mean tap water intake differed between the survey cycles for men aged ≥71 years.

Survey variables must be examined before combining or comparing data from multiple WWEIA/NHANES release cycles. For at least some age/gender groups, drinking water intake data from NHANES cycles prior to 2005–2006 should not be considered comparable to more recent data.