Estimating correlation between multivariate longitudinal data in the presence of heterogeneity

15 Years of Longitudinal Genetic, Clinical, Cognitive, Imaging, and Biochemical Measures in DIAN

This manuscript describes and summarizes the Dominantly Inherited Alzheimer Network Observational Study (DIAN Obs), highlighting the wealth of longitudinal data, samples, and results from this human cohort study of brain aging and a rare monogenic form of Alzheimer's disease (AD). DIAN Obs is an international collaborative longitudinal study initiated in 2008 with support from the National Institute on Aging (NIA), designed to obtain comprehensive and uniform data on brain biology and function in individuals at risk for autosomal dominant AD (ADAD). ADAD gene mutations in the amyloid protein precursor (APP), presenilin 1 (PSEN1), or presenilin 2 (PSEN2) genes are deterministic causes of ADAD, with virtually full penetrance, and a predictable age at symptomatic onset. Data and specimens collected are derived from full clinical assessments, including neurologic and physical examinations, extensive cognitive batteries, structural and functional neuro-imaging, amyloid and tau pathological measures using positron emission tomography (PET), flurordeoxyglucose (FDG) PET, cerebrospinal fluid and blood collection (plasma, serum, and whole blood), extensive genetic and multi-omic analyses, and brain donation upon death. This comprehensive evaluation of the human nervous system is performed longitudinally in both mutation carriers and family non-carriers, providing one of the deepest and broadest evaluations of the human brain across decades and through AD progression. These extensive data sets and samples are available for researchers to address scientific questions on the human brain, aging, and AD.

Adaptive data-driven selection of sequences of biological and cognitive markers in pre-clinical diagnosis of dementia

Effective clinical decision procedures must balance multiple competing objectives such as time-to-decision, acquisition costs, and accuracy. We describe and evaluate POSEIDON, a data-driven method for PrOspective SEquentIal DiagnOsis with Neutral zones to individualize clinical classifications. We evaluated the framework with an application in which the algorithm sequentially proposes to include cognitive, imaging, or molecular markers if a sufficiently more accurate prognosis of clinical decline to manifest Alzheimer's disease is expected. Over a wide range of cost parameter data-driven tuning lead to quantitatively lower total cost compared to ad hoc fixed sets of measurements. The classification accuracy based on all longitudinal data from participants that was acquired over 4.8 years on average was 0.89. The sequential algorithm selected 14 percent of available measurements and concluded after an average follow-up time of 0.74 years at the expense of 0.05 lower accuracy. Sequential classifiers were competitive from a multi-objective perspective since they could dominate fixed sets of measurements by making fewer errors using less resources. Nevertheless, the trade-off of competing objectives depends on inherently subjective prescribed cost parameters. Thus, despite the effectiveness of the method, the implementation into consequential clinical applications will remain controversial and evolve around the choice of cost parameters.

Joint modelling of systolic and diastolic blood pressure and its associated factors among women in Ghana: Multivariate response multilevel modelling methods

Elevated blood pressure is the leading cause of cardiovascular diseases related mortality and a major contributor to non-communicable diseases globally, especially in sub-Saharan Africa where about 74.7 million people live with hypertension. In Ghana, hypertension is epidemic with prevalence of over 30% and experiencing continuing burden with its associated morbidity and mortality. Using the 2014 Ghana Demographic and Health Survey, we analyzed data on 4744 women aged 15-49 years residing in 3722 households. We employed univariate and multivariate response multilevel linear regression models to analyze predictors of systolic blood pressure (SBP) and diastolic blood pressure (DBP). Geospatial maps were produced to show the regional distribution of hypertension prevalence in Ghana. Stata version 17 and R version 4.2.1 were used to analyze the data. Of the 4744 woman, 337 (7.1%) and 484 (10.2%) were found to be hypertensive on SBP and DBP, respectively. A combined prevalence of 12.3% was found. Older ages 25-34 (OR 2.45, 95%CI: 1.27, 3.63), 35-44 (OR 8.72, 95%CI: 7.43, 10.01), 45-49 (OR 15.85, 95%CI: 14.07, 17.64), being obese (OR 5.10, 95%CI: 3.62, 6.58), and having no education (OR -2.05, 95%CI: -3.40, -0.71) were associated with SBP. For DBP, we found the associated factors to be older ages 25-34 (OR 3.29, 95%CI: 2.50, 4.08), 35-44 (OR 6.78, 95%CI: 5.91, 7.64), 45-49 (OR 10.05, 95%CI: 8.85, 11.25), being obese (OR 4.20, 95%CI: 3.21, 5.19), and having no education (OR -1.23, 95%CI: -2.14, -0.33). Substantial residual household level differences in SBP (15%) and DBP (14%) were observed. We found strong residual correlation of SBP and DBP on individual women (r = 0.73) and household-level (r = 0.81). The geospatial maps showed substantial regional differences in the observed and reported hypertension prevalence. Interventions should be targeted at the identified high-risk groups like older age groups and those who are obese, and the high-risk regions.

Variations in Physical and Competitive Performance of Highly Trained Sprinters Across an Annual Training Season

ABSTRACT

Loturco, I, Fernandes, V, Bishop, C, Mercer, VP, Siqueira, F, Nakaya, K, Pereira, LA, and Haugen, T. Variations in physical and competitive performance of highly trained sprinters across an annual training season. J Strength Cond Res XX(X): 000-000, 2022-We assessed the changes in sprint, jump, and power parameters across the annual training cycle and tested the longitudinal correlations among these variables in top-level sprinters. Thirteen sprinters training with 4 different Olympic sprint coaches were sequentially assessed over 14 months, from January 2019 to March 2020, within 4 consecutive training camps. Performance tests were conducted as follows: standing long jump, squat and countermovement jumps, 10-m and 60-m sprint time, and maximum power output in the half-squat, jump-squat, and hip-thrust exercises. The competitive results of the sprinters throughout the study period were also recorded and analyzed. A repeated measures analysis of variance was used to compare the physical measurements between different testing sessions. A Pearson product-moment correlation was applied to examine the longitudinal relationships between changes in speed-related and power-related parameters. Percentage change was computed and compared with coefficient of variation values to determine whether changes in performance metrics were higher than the test variance, thus providing an indication of whether true changes occurred on an individual basis. Overall, sprinters did not exhibit significant changes in sprint speed, jumping ability, and power output. In addition, variations in competitive times (i.e., 100 m races) followed a similar pattern, within an average range of ±1.36%, for both male and female sprinters. As expected, top-level sprinters presented only small variations in physical and competitive performance over time. Nevertheless, the use of an individual statistical technique (i.e., true changes calculation) revealed that these nonsignificant increases or decreases may represent meaningful changes in their competitive potential.