Comparing the relative fit of competing models can be used to address many different scientific questions. In classical statistics one can, if appropriate, use likelihood ratio tests and information based criterion, whereas clinical medicine has tended to rely on comparisons of fit metrics like C-statistics. However, for many data adaptive modelling procedures such approaches are not suitable. In these cases, statisticians have used cross-validation, which can make inference challenging. In this paper we propose a general approach that focuses on the "conditional" risk difference (conditional on the model fits being fixed) for the improvement in prediction risk. Specifically, we derive a Wald-type test statistic and associated confidence intervals for cross-validated test sets utilizing the independent validation within cross-validation in conjunction with a test for multiple comparisons. We show that this test maintains proper Type I Error under the null fit, and can be used as a general test of relative fit for any semi-parametric model alternative. We apply the test to a candidate gene study to test for the association of a set of genes in a genetic pathway.

Background

Genome-wide association studies focusing on European-ancestry populations have identified ALL risk loci on IKZF1, ARID5B, and CEBPE. To capture the impacts of these genes on ALL risk in the California Hispanic population, we comprehensively assessed the variation within the genes and further assessed the joint effects between the genetic variation and surrogates for early-life infections (the presence of older siblings, daycare attendance, and ear infections).

Methods

Genotypic data for 323 Hispanic ALL cases and 454 controls from the California Childhood Leukemia Study were generated using Illumina OmniExpress v1 platform. Logistic regression assuming a log-additive model estimated odds ratios (OR) associated with each SNP, adjusted for age, sex, and the first five principal components. In addition, we examined potential interactions between six ALL risk alleles and surrogates for early-life infections using logistic regression models that included an interaction term.

Results

Significant associations between genotypes at IKZF1, ARID5B, and CEBPE and ALL risk were identified: rs7780012, OR 0.50, 95% confidence interval (CI) 0.35-0.71 (p = 0.004); rs7089424, OR 2.12, 95% CI 1.70-2.65 (p = 1.16 × 10(-9)); rs4982731, OR 1.69, 95% CI 1.37-2.08 (p = 2.35 × 10(-6)), respectively. Evidence for multiplicative interactions between genetic variants and surrogates for early-life infections with ALL risk was not observed.

Conclusions

Consistent with findings in non-Hispanic White population, our study showed that variants within IKZF1, ARID5B, and CEBPE were associated with increased ALL risk, and the effects for ARID5B and CEBPE were most prominent in the high-hyperdiploid ALL subtype in the California Hispanic population. Results implicate the ARID5B, CEBPE, and IKZF1 genes in the pathogenesis of childhood ALL.

Objectives

Antipsychotic drugs (APS) are widely used to treat patients with bipolar disorder (BD), but there is limited information in older-age bipolar disorder (OABD). This analysis of the Global Aging & Geriatric Experiments in Bipolar Disorder Database (GAGE-BD) investigated characteristics of OABD patients prescribed APS vs. those not prescribed APS.

Experimental design

The observational analysis used baseline, cross-sectional data from 16 international studies for adults aged ≥ 50 years with BD comprising 1,007 individuals with mean age 63.2 years (SD = 9.0), 57.4% women, and mean age of onset 31.6 years (SD = 15.0). The dependent variable was current APS treatment status. The independent variables included demographic and clinical variables, and a random effect for study, that were included in generalized mixed models.

Principal observations

46.6% of individuals (n = 469) were using APS. The multivariate model results suggest that those treated with APS were younger (p = 0.01), less likely to be employed (p < 0.001), had more psychiatric hospitalizations (p = 0.009) and were less likely to be on lithium (p < 0.001). Of individuals on APS, only 6.6% of those (n = 27) were on first-generation antipsychotics (FGAs) and experienced a greater burden of psychiatric hospitalizations (p = 0.012).

Conclusions

APS are widely prescribed in OABD, observed in nearly half of this sample with great variation across sites. Individuals with OABD on APS have more severe illness, more frequent hospitalizations and are more often unemployed vs. those not on APS. Future studies need to examine longitudinal outcomes in OABD prescribed APS to characterize underlying causal relationships.

Objective

Literature on older-age bipolar disorder (OABD) is limited. This first-ever analysis of the Global Aging & Geriatric Experiments in Bipolar Disorder Database (GAGE-BD) investigated associations among age, BD symptoms, comorbidity, and functioning.

Methods

This analysis used harmonized, baseline, cross-sectional data from 19 international studies (N = 1377). Standardized measures included the Young Mania Rating Scale (YMRS), Hamilton Depression Rating Scale (HAM-D), Montgomery-Asberg Depression Rating Scale (MADRS), and Global Assessment of Functioning (GAF).

Results

Mean sample age was 60.8 years (standard deviation [SD] 12.2 years), 55% female, 72% BD I. Mood symptom severity was low: mean total YMRS score of 4.3 (SD 5.4) and moderate-to-severe depression in only 22%. Controlled for sample effects, both manic and depressive symptom severity appeared lower among older individuals (p's < 0.0001). The negative relationship between older age and symptom severity was similar across sexes, but was stronger among those with lower education levels. GAF was mildly impaired (mean =62.0, SD = 13.3) and somatic burden was high (mean =2.42, SD = 1.97). Comorbidity burden was not associated with GAF. However, higher depressive (p < 0.0001) and manic (p < 0.0001) symptoms were associated with lower GAF, most strongly among older individuals.

Conclusions

Findings suggest an attenuation of BD symptoms in OABD, despite extensive somatic burden. Depressive symptom severity was strongly associated with worse functioning in older individuals, underscoring the need for effective treatments of BD depression in older people. This international collaboration lays a path for the development of a better understanding of aging in BD.

Objective

We investigated the association between 52 risk variants identified through genome-wide association studies and disease severity in multiple sclerosis (MS).

Methods

Ten unique MS case data sets were analyzed. The Multiple Sclerosis Severity Score (MSSS) was calculated using the Expanded Disability Status Scale at study entry and disease duration. MSSS was considered as a continuous variable and as 2 dichotomous variables (median and extreme ends; MSSS of ≤5 vs >5 and MSSS of <2.5 vs ≥7.5, respectively). Single nucleotide polymorphisms (SNPs) were examined individually and as both combined weighted genetic risk score (wGRS) and unweighted genetic risk score (GRS) for association with disease severity. Random-effects meta-analyses were conducted and adjusted for cohort, sex, age at onset, and HLA-DRB1*15:01.

Results

A total of 7,125 MS cases were analyzed. The wGRS and GRS were not strongly associated with disease severity after accounting for cohort, sex, age at onset, and HLA-DRB1*15:01. After restricting analyses to cases with disease duration ≥10 years, associations were null (p value ≥0.05). No SNP was associated with disease severity after adjusting for multiple testing.

Conclusions

The largest meta-analysis of established MS genetic risk variants and disease severity, to date, was performed. Results suggest that the investigated MS genetic risk variants are not associated with MSSS, even after controlling for potential confounders. Further research in large cohorts is needed to identify genetic determinants of disease severity using sensitive clinical and MRI measures, which are critical to understanding disease mechanisms and guiding development of effective treatments.