Reduced Rank Regression Research Articles

Testosterone-Associated Dietary Pattern Predicts Low Testosterone Levels and Hypogonadism.

Obesity and low serum testosterone (T) levels are interrelated and strongly influenced by dietary factors, and their alteration entails a great risk of hypogonadism. Substantial evidence suggests a bidirectional relationship between nutrient metabolism (e.g., glucose, lipids, and iron) and T levels in men; however, T-related dietary patterns remain unclear. This study investigated the dietary patterns associated with serum total T levels and its predictive effect on hypogonadism and the body composition. Anthropometry, blood biochemistry, and food frequency questionnaires were collected for 125 adult men. Dietary patterns were derived using a reduced rank regression from 32 food groups. Overall prevalence rates of central obesity and hypogonadism were 48.0% and 15.7%, respectively. An adjusted linear regression showed that age, insulin, red blood cell (RBC) aggregation, and transferrin saturation independently predicted serum total T levels (all p < 0.01). The total T-related dietary pattern (a high consumption of bread and pastries, dairy products, and desserts, eating out, and a low intake of homemade foods, noodles, and dark green vegetables) independently predicted hypogonadism (odds ratio: 5.72; 95% confidence interval: 1.11‒29.51, p < 0.05) for those with the highest dietary pattern scores (Q4) compared to those with the lowest (Q1). Scores were also negatively correlated with the skeletal muscle mass (p for trend = 0.002) but positively correlated with the total body fat mass (p for trend = 0.002), visceral fat mass (p for trend = 0.001), and to a lesser extent, subcutaneous fat mass (p for trend = 0.035) after adjusting for age. Randomized controlled trials are needed to confirm that improvement in dietary pattern can improve T levels and reduce hypogonadism.

On Cross-Validation for Sparse Reduced Rank Regression

SummaryIn high dimensional data analysis, regularization methods pursuing sparsity and/or low rank have received much attention recently. To provide a proper amount of shrinkage, it is typical to use a grid search and a model comparison criterion to find the optimal regularization parameters. However, we show that fixing the parameters across all folds may result in an inconsistency issue, and it is more appropriate to cross-validate projection–selection patterns to obtain the best coefficient estimate. Our in-sample error studies in jointly sparse and rank deficient models lead to a new class of information criteria with four scale-free forms to bypass the estimation of the noise level. By use of an identity, we propose a novel scale-free calibration to help cross-validation to achieve the minimax optimal error rate non-asymptotically. Experiments support the efficacy of the methods proposed.

Dietary patterns during pregnancy derived by reduced-rank regression and their association with gestational diabetes mellitus.

Evidence for the combined association of the quality of dietary carbohydrates and fats during pregnancy with gestational diabetes mellitus (GDM) is scarce. The aim of this study was to identify dietary patterns during pregnancy, derived by the reduced-rank regression (RRR) model, associated with fiber density (g/1000 kcal) and the Thrombogenicity Index (TI) and to investigate their relationship with GDM. This was a cross-sectional study conducted with 785 pregnant women at gestational weeks 24 to 39. The diagnosis of GDM was based on the World Health Organization criteria. One 24-h dietary recall was obtained from all women and a second measurement from 73% of the sample, with the multiple source method used to estimate the usual diet. The patterns were determined by RRR, and the relationship with GDM was investigated using adjusted logistic regression models. The mean (standard deviation) age of the women was 28 y and 17.7% had GDM. Dietary pattern 1 (high rice, beans, and vegetables, with low full-fat dairy products, biscuits, and sweets) correlated positively with fiber density and negatively with TI and was inversely associated with GDM after multiple adjustments (odds ratio [OR], 0.58; 95% confidence interval [CI], 0.36-0.95; P = 0.03). Dietary pattern 2 (high red meats, full-fat dairy products, chocolate powder and fruits, with low chicken and margarine) correlated positively with both TI and fiber density, with no association with GDM found (OR, 1.48; 95% CI, 0.91-2.40; P = 0.11). These data highlight the joint importance of the quality of dietary carbohydrates and fats in the dietary patterns of pregnant women in relation to the risk for developing GDM.

Dietary Patterns in Primary School are of Prospective Relevance for the Development of Body Composition in Two German Pediatric Populations.

This study performed comparative analyses in two pediatric cohorts to identify dietary patterns during primary school years and examined their relevance to body composition development. Nutritional and anthropometric data at the beginning of primary school and two or four years later were available from 298 and 372 participants of IDEFICS-Germany (Identification and prevention of Dietary-induced and lifestyle-induced health Effects In Children and infants Study) and the KOPS (Kiel Obesity Prevention Study) cohort, respectively. Principal component analyses (PCA) and reduced rank regression (RRR) were used to identify dietary patterns at baseline and patterns of change in food group intake during primary school years. RRR extracted patterns explaining variations in changes in body mass index (BMI), fat mass index (FMI), and waist-to-height-ratio (WtHR). Associations between pattern adherence and excess gain in BMI, FMI, or WtHR (>75th percentile) during primary school years were examined using logistic regression. Among PCA patterns, only a change towards a more Mediterranean food choice during primary school years were associated with a favorable body composition development in IDEFICS-Germany (p < 0.05). In KOPS, RRR patterns characterized by a frequent consumption of fast foods or starchy carbohydrate foods were consistently associated with an excess gain in BMI and WtHR (all p < 0.005). In IDEFICS-Germany, excess gain in BMI, FMI, and WtHR were predicted by a frequent consumption of nuts, meat, and pizza at baseline and a decrease in the consumption frequency of protein sources and snack carbohydrates during primary school years (all p < 0.01). The study confirms an adverse impact of fast food consumption on body composition during primary school years. Combinations of protein and carbohydrate sources deserve further investigation.

High-Sugar, High-Saturated-Fat Dietary Patterns Are Not Associated with Depressive Symptoms in Middle-Aged Adults in a Prospective Study

The consumption of unhealthy "Western" dietary patterns has been previously associated with depressive symptoms in different populations. We examined whether high-sugar and high-saturated-fat dietary patterns are associated with depressive symptoms over 5 y in a British cohort of men and women. We used data from the Whitehall II study in 5044 individuals (aged 35-55 y). Diet was assessed at phase 7 (2003-2004) using a validated food-frequency questionnaire. Dietary patterns were derived by using reduced rank regression with sugar, saturated fat, and total fat as response variables. The Center for Epidemiological Studies-Depression (CES-D) scale was used to assess depressive symptoms (CES-D sum score ≥16 and/or use of antidepressant medication) at phase 7 and at phase 9 (2008-2009). We applied logistic regression analyses to test the association between dietary patterns and depressive symptoms. All analyses were stratified by sex. In total, 398 cases of recurrent and 295 cases of incident depressive symptoms were observed. We identified 2 dietary patterns: a combined high-sugar and high-saturated-fat (HSHF) and a high-sugar dietary pattern. No association was observed between the dietary patterns and either incidence of or recurrent depressive symptoms in men or women. For example, higher consumption of the HSHF dietary pattern was not associated with recurrent depressive symptoms in men (model 3, quartile 4: OR: 0.67; 95% CI: 0.36, 1.23; P-trend = 0.13) or in women (model 3, quartile 4: OR: 1.26; 95% CI: 0.58, 2.77; P-trend = 0.97). Among middle-aged men and women living in the United Kingdom, dietary patterns containing high amounts of sugar and saturated fat are not associated with new onset or recurrence of depressive symptoms.

Reduced rank modeling for functional regression with functional responses

This article considers regression problems where both the predictor and the response are functional in nature. Driven by the desire to build a parsimonious model, we consider functional reduced rank regression in the framework of reproducing kernel Hilbert spaces, which can be formulated in the form of linear factor regression with estimated multivariate factors, and achieves dimension reduction in both the predictor and the response spaces. The convergence rate of the estimator is derived. Simulations and real datasets are used to demonstrate the competitive performance of the proposed method.

Group sparse reduced rank regression for neuroimaging genetic study.

The neuroimaging genetic study usually needs to deal with high dimensionality of both brain imaging data and genetic data, so that often resulting in the issue of curse of dimensionality. In this paper, we propose a group sparse reduced rank regression model to take the relations of both the phenotypes and the genotypes for the neuroimaging genetic study. Specifically, we propose designing a graph sparsity constraint as well as a reduced rank constraint to simultaneously conduct subspace learning and feature selection. The group sparsity constraint conducts feature selection to identify genotypes highly related to neuroimaging data, while the reduced rank constraint considers the relations among neuroimaging data to conduct subspace learning in the feature selection model. Furthermore, an alternative optimization algorithm is proposed to solve the resulting objective function and is proved to achieve fast convergence. Experimental results on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset showed that the proposed method has superiority on predicting the phenotype data by the genotype data, than the alternative methods under comparison.

Dietary patterns related to exposure to persistent organic pollutants based on the Ewha Birth and Growth Cohort

Persistent organic pollutants (POPs) are endocrine disruptors that are typically ingested in food. Using data from the Ewha Birth and Growth Cohort study, we identified dietary patterns related to blood persistent organic pollutant (POP) levels in children. In total, 188 Korean children (95 boys and 93 girls) aged 7–9 years were included in this study. Daily food intake data derived from a semi-quantitative food-frequency questionnaire were grouped into 33 food groups. Dietary patterns were analyzed using the reduced rank regression method, with blood POP levels as dependent variables.The marker polychlorinated biphenyl (PCB) pattern showed more weighted loading factors in the intake of shellfish/salted seafood, cheese, nuts and seeds, and total PCBs showed similar results. The dioxin-like PCB pattern was characterized by a higher intake of yogurt, beverages, and fruit and a lower intake of grain, seaweeds, and processed meat. Additionally, the total organochlorine pesticide (OCP) pattern showed positive factor loading values in shrimp and beverages, and negative factor loading values in seaweeds and processed meat. Each dietary pattern accounted for 25.0% of total OCPs, 22.9% of dioxin-like PCBs, 21.6% of marker PCBs, and 21.3% of total PCBs in blood. Dietary pattern scores were linearly related to blood POP levels, and this association remained after adjusting for sex, age, follow-up survey year, maternal educational level, body mass index, and total energy. Dietary patterns differed slightly among the individual POPs. Further studies are needed to develop an intervention strategy for children.

Association between dietary patterns derived by reduced rank regression and incident diabetes: From the korean genome epidemiology study

Rationale: Using data from the Korean Genome Epidemiology Study, we assessed dietary patterns related to clinical indicators of diabetes.

Red Blood Cell Aggregation-Associated Dietary Pattern Predicts Hyperlipidemia and Metabolic Syndrome.

Red blood cell (RBC) aggregation and iron status are interrelated and strongly influenced by dietary factors, and their alterations pose a great risk of dyslipidemia and metabolic syndrome (MetS). Currently, RBC aggregation-related dietary patterns remain unclear. This study investigated the dietary patterns that were associated with RBC aggregation and their predictive effects on hyperlipidemia and MetS. Anthropometric and blood biochemical data and food frequency questionnaires were collected from 212 adults. Dietary patterns were derived using reduced rank regression from 32 food groups. Adjusted linear regression showed that hepcidin, soluble CD163, and serum transferrin saturation (%TS) independently predicted RBC aggregation (all p < 0.01). Age-, sex-, and log-transformed body mass index (BMI)-adjusted prevalence rate ratio (PRR) showed a significant positive correlation between RBC aggregation and hyperlipidemia (p-trend < 0.05). RBC aggregation and iron-related dietary pattern scores (high consumption of noodles and deep-fried foods and low intake of steamed, boiled, and raw food, dairy products, orange, red, and purple vegetables, white and light-green vegetables, seafood, and rice) were also significantly associated with hyperlipidemia (p-trend < 0.05) and MetS (p-trend = 0.01) after adjusting for age, sex, and log-transformed BMI. Our results may help dieticians develop dietary strategies for preventing dyslipidemia and MetS.

Exploratory dietary patterns: a systematic review of methods applied in pan-European studies and of validation studies.

Besides a priori approaches, using previous knowledge about food characteristics, exploratory dietary pattern (DP) methods, using data at hand, are commonly applied. This systematic literature review aimed to identify exploratory methods on DP in pan-European studies and to inform the development of the DEterminants of DIet and Physical ACtivity (DEDIPAC) toolbox of methods suitable for use in future European studies. The search was conducted in three databases on prospective studies in healthy, free-living people across the whole life span. To identify validated DP methods, an additional search without regional restrictions was conducted. Studies including at least two European countries were retained. The search resulted in six pan-European studies applying principal component/factor analysis (PC/FA) (n 5) or cluster analysis (n 2). The criteria to retain PC/factors ranged from the application of the eigenvalue>1 criterion, the scree plot and/or the interpretability criterion. Furthermore, seven validation studies were identified: DP, derived by PC/FA (n 6) or reduced rank regression (RRR) (n 1) were compared using dietary information from FFQ (n 6) or dietary history (n 1) as study instrument and dietary records (n 6) or 24-h dietary recalls (n 1) as reference. The correlation coefficients for the derived DP ranged from modest to high. To conclude, PC/FA was predominantly applied using the eigenvalue criterion and scree plot to retain DP, but a better description of the applied criteria is highly recommended to enable a standardised application of the method. Research gaps were identified for the methods cluster analysis and RRR, as well as for validation studies on DP.

Leveraging mixed and incomplete outcomes via reduced-rank modeling

Multivariate outcomes with multivariate features of possibly high dimension are routinely produced in various fields. In many real-world problems, the collected outcomes are of mixed types, including continuous measurements, binary indicators and counts, and a substantial proportion of values may also be missing. Regardless of their types, these mixed outcomes are often interrelated, representing diverse reflections or views of the same underlying data generation mechanism. As such, an integrative multivariate model can be beneficial. We develop a mixed-outcome reduced-rank regression, which effectively enables information sharing among different prediction tasks. Our approach integrates mixed and partially observed outcomes belonging to the exponential dispersion family, by assuming that all the outcomes are associated through a shared low-dimensional subspace spanned by the features. A general singular value regularized criterion is proposed, and we establish a non-asymptotic performance bound for the proposed estimators in the context of supervised learning with mixed outcomes from an exponential family and under a general sampling scheme of missing data. An iterative singular value thresholding algorithm is developed for optimization with convergence guarantee. The effectiveness of our approach is demonstrated by simulation studies and an application on predicting health-related outcomes in longitudinal studies of aging.

Biomaker patterns of fatty acids and other fat-soluble biocompounds in blood to indicate nuts intake in the European Prospective Investigation into Cancer and Nutrition (EPIC) study

Summary Introduction Nuts are nutrient dense foods, rich in unsaturated fatty acids and other fat-soluble biocompounds. Epidemiological and intervention studies have shown that a high intake of nuts is associated with a reduced risk of coronary heart disease; however, inaccurate estimation of nut intake using self-reported methods might limit detection of other health benefits such as reduced risk of cancer, where potential strengths of associations are likely smaller. This study aimed to identify biomarker patterns including fatty acids (FAs) and other fat-soluble compounds (carotenoids and tocopherols) in blood to indicate nut intake by applying a statistical algorithm combining dimension reduction and variable selection methods. Methods The study included 2324 subjects (44% men) aged 35–70 years from 8 European countries in EPIC cross-sectional study. Recent and habitual nut intakes were assessed with 24-h dietary recalls (24-HDR) and dietary questionnaire (DQ), respectively. Potential biomarkers related to nut intake included 22 plasma FAs and serum fat-soluble compounds (7 carotenoids and 2 tocopherols). They were measured by gas chromatography and high-performance liquid chromatography, respectively. In order to identify patterns of these biomarkers, maximizing the explained variability in nut intake, reduced rank regression (RRR) models were used with optimal subsets of biomarkers selected by two different variable selection methods; RRR-based variable importance in projection (RRR-VIP) and least absolute shrinkage and selection operator (LASSO). Prior to the main analysis, nut intake and biomarker levels were log-transformed and adjusted for energy intake (nut intake only) and country taking residuals in linear models. The performance of RRR models was evaluated by Pearson correlation coefficients of biomarker pattern scores with recent and habitual nut intakes through internal two-fold cross-validation. Results Higher performance was observed in biomarker patterns on habitual nut intake reported in the DQ. The selected subsets of biomarkers slightly differed depending on which selection method was used, but arachidic acid, linoleic acid and tocopherols, known as abundant compounds in nuts, were commonly selected as biomarkers to explain habitual nut intake in both methods. Pearson correlation coefficients of biomarker pattern scores with habitual nut intake (r = 0.27) were two to three times stronger than those of single biomarkers (linoleic acid r = 0.08; arachidic acid r = 0.11). Neither biomarker patterns nor single biomarkers performed well in explaining recent nut intake reported in the 24-HDR. Conclusions Biomarker patterns consisting of fatty acids or fat-soluble biocompounds better indicated nut intakes compared to any of these single biomarkers evaluated, especially for habitual nut intake. Whether these identified biomarker patterns improve assessment of nut intake in diet-disease studies needs to be evaluated in future studies.

A dietary pattern derived using B-vitamins and its relationship with vascular markers over the life course

SummaryBackgroundDiet may influence vascular function through elevated homocysteine (Hcy) concentrations. However the relationship between dietary patterns (DP), characterised by Hcy and its associated nutrients is unknown.ObjectiveTo identify a DP characterised by plasma Hcy, dietary folate and dietary vitamin B12, and examine its associations with two markers of vascular function: carotid intima-media thickness (cIMT) and pulse wave velocity (PWV).Methods1562 participants of the MRC National Survey of Health and Development (NSHD), a British birth cohort, with dietary data measured at least once between 36 and 60–64 years, and cIMT or PWV measured at 60–64 years were included. DPs were derived using reduced rank regression with three intermediate variables: 1) plasma Hcy (μmol/L) 2) folate intake (μg/1000 kcal) 3) vitamin B12 intake (μg/1000 kcal). Multiple regression models assessed associations between the derived DP z-scores and vascular function adjusting for dietary misreporting, socioeconomic position, BMI, smoking, physical activity and diabetes.ResultsA DP explaining the highest amount of shared variation (4.5%) in plasma Hcy, dietary folate and dietary vitamin B12 highly correlated with folate (r = 0.96), moderately correlated with vitamin B12 (r = 0.27), and weakly correlated with Hcy (r = 0.10). This “high B-vitamin” DP (including folate) was characterised by high intakes of vegetables, fruit and low fibre breakfast cereal, and low intakes of processed meat, white bread, sugar and preserves. No associations were observed between DP z-scores and vascular function at any time point following adjustment for covariates.ConclusionThis study explored a specific hypothesised pathway linking diet to vascular function. Although we found no consistent evidence for an association between a high B-vitamin DP and vascular function, we did observe an association with CRP and triglycerides in secondary analyses. Further analyses using strongly correlated and biologically relevant intermediate variables are required to refine investigations into diet and CVD in longitudinal cohort data.

Development of an Inflammatory Food Index and Association with Inflammatory Markers and Diabetes—Longitudinal Study of Adult Health (ELSA-Brasil)

Introduction: Diet plays a central role in the regulation of inflammation and is closely related to the development of chronic diseases. Dietary patterns portray eating habits more comprehensively than nutrients or foods alone, and provide better understanding and predictive power in relation to causal processes. The objective was to develop an Index of Food Inflammation (IFI) derived from Reduced Rank Regression (RRR) based on its relation with inflammatory markers in the ELSA-Brasil population, and to evaluate its association with weight gain and incidence of diabetes. Methods: Food consumption data were obtained from a food frequency questionnaire of 9778 participants, after exclusions. The RRR methodology initially involved 42 food groups. Spearman’s correlation coefficients were used to evaluate the associations between quantitative variables, and, for the IFI, linear regression to evaluate its association with weight gain, logistic regression with large gains, and Cox regression with incident diabetes. Results: The inflammatory dietary pattern was composed of 17 pro-inflammatory food groups, including processed or red meat, pork, sugary soft drinks, and hotdogs, and seven anti-inflammatory groups, notably fruits, nuts and wine. The IFI was associated with weight gain, risk of a large weight gain (quartile 4 vs. 1: adjusted OR=1.42; 95% CI 1.15-1.76) and risk of developing diabetes (quartile 4 vs. 1: adjusted HR=1.29; 95% CI 1.01-1.65). Conclusion: The associations, although small, suggest that subclinical inflammation could be one of the ways by which diet causes certain chronic diseases. Ultra-processing of foods could be an important determinant in the pro-inflammatory aspect of some current dietary patterns. Disclosure B.P. Riboldi: None. V.C. Luft: None. M.I. Schmidt: Research Support; Self; Eli Lilly and Company. Research Support; Spouse/Partner; Eli Lilly and Company. B.B. Duncan: Research Support; Self; Eli Lilly and Company. Research Support; Spouse/Partner; Eli Lilly and Company.

A Robust Reduced Rank Graph Regression Method for Neuroimaging Genetic Analysis.

To characterize associations between genetic and neuroimaging data, a variety of analytic methods have been proposed in neuroimaging genetic studies. These methods have achieved promising performance by taking into account inherent correlation in either the neuroimaging data or the genetic data alone. In this study, we propose a novel robust reduced rank graph regression based method in a linear regression framework by considering correlations inherent in neuroimaging data and genetic data jointly. Particularly, we model the association analysis problem in a reduced rank regression framework with the genetic data as a feature matrix and the neuroimaging data as a response matrix by jointly considering correlations among the neuroimaging data as well as correlations between the genetic data and the neuroimaging data. A new graph representation of genetic data is adopted to exploit their inherent correlations, in addition to robust loss functions for both the regression and the data representation tasks, and a square-root-operator applied to the robust loss functions for achieving adaptive sample weighting. The resulting optimization problem is solved using an iterative optimization method whose convergence has been theoretically proved. Experimental results on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset have demonstrated that our method could achieve competitive performance in terms of regression performance between brain structural measures and the Single Nucleotide Polymorphisms (SNPs), compared with state-of-the-art alternative methods.

Identification of sustainable dietary patterns by a multicriteria approach in the NutriNet-Santé cohort

In the context of the increasingly damaging impact of agro-food systems on human health and the environment, this study aimed to evaluate the sustainability of current dietary patterns using multi criteria analysis to characterize consumers with different dietary patterns. In a sample of 29,413 participants of the NutriNet-Sante Study, total and organic food intakes of 264 items gathered in 17 food groups were evaluated using a food frequency questionnaire. Eight sustainability indicators were assessed individually. To identify combinations of food groups explaining the maximum variation in the sustainability indicators, we applied reduced rank regression with food groups as predictors and sustainability indicators as response variables. Then, the first two RRR-derived factors were used to classify participants using an ascending hierarchical classification. Six clusters were identified, among then one cluster (N = 23.07%) presented a good compromise between all the sustainable dimensions. Indeed, their diets emitted on average 36% less greenhouses gas, included 30% more organic food and exhibited a close level of affordability compared to the average. The dietary composition in this group, were characterized by -58% of red meat, -37% of white meat, -25% of cheese but +15% of fruits, +15% of vegetable and a similar contribution of fish and starches, compared to the sample average. Finally, this study showed that in the current food system context, the observed diets rarely meet all sustainability criteria. Oppositions between healthy, eco-friendly and affordability often remain. However, we obServed some diets with high compatibility with all sustainable indicators selected for the study. Thus sustainable diets are emerging and should be promoted. However, it would be relevant to conduct additional research in other population to estimate acceptability.

Association between blood pressure related dietary patterns and identified cognitive performance in the elderly Chinese-a study by reduced rank regression method

Objective: To study the association between blood pressure related dietary pattern and cognitive impairment in the elderly. Methods: In 2015, all participants who were aged ≥60 and participated in the Nutrition and Chronic disease family cohort were involved in the study. Information on demographic variables, lifestyle and health status was collected. Cognitive performance was assessed by the Mini Mental State Evaluation (MMSE) scale. Blood pressure, height and weight were measured by trained medical personnel and fasting venous blood samples were collected for testing on serum level of triglycerides and total cholesterol. Both SBP and DBP were used as response variables when dietary patterns were identified by reduced rank regression method. Logistic regression models were fit to explore the associations of scores on blood pressure-related dietary pattern and cognitive impairment. Results: Two blood related dietary patterns were identified. The first one was characterized by high consumption of vegetables and less meat, eggs and dessert (Pattern 1), while the second one was with high consumption of meat, soy products, wine and fried foods and less intake of dairy (Pattern 2). Data showed that the Pattern 1 was associated with the risk of cognitive impairment. Comparing with the lowest quartile of score of this dietary pattern, the risk of cognitive impairment in the highest quartile group showed a significant (P<0.01) increase, with OR=1.94 (1.21-3.11) and showing significant (P=0.002) linear trend. However, no significant association was observed (P>0.05) with cognitive impairment in the second dietary pattern. Conclusion: Blood pressure-related dietary pattern was positively associated with cognitive impairment.

Inflammatory Dietary Pattern, IL-17F Genetic Variant, and the Risk of Colorectal Cancer.

A proinflammatory diet may increase the risk of colorectal cancer, but its role may differ according to individuals’ genetic variants. We aimed to examine whether a specific dietary pattern reflecting inflammation was associated with a risk of colorectal cancer and whether IL-17F genetic variant altered this association. In a study of 695 colorectal cancer cases and 1846 controls, we derived a reduced rank regression dietary pattern using 32 food groups as predictors and the plasma C-reactive protein (CRP) concentration as the response. High CRP levels were associated with a high risk of colorectal cancer (OR (95% CI) = 3.58 (2.65–4.82) for the highest quartile vs. lowest quartile). After adjusting for potential confounding factors, high pattern scores were associated with a high risk of colorectal cancer (OR (95% CI) = 9.98 (6.81–14.62) for the highest quartile vs. lowest quartile). When stratified by the IL-17F rs763780 genotype, this association was stronger for individuals carrying the C allele (p for interaction = 0.034), particularly for individuals with rectal cancer (p for interaction = 0.011). In conclusion, a dietary pattern reflecting inflammation was significantly associated with colorectal cancer risk. Moreover, this association could be modified according to the IL-17F rs763780 genotype and anatomic site.

Handling Outlier in Two-Ways Table Data: The Robustness of Row-Column Interaction Model

As part of our recent statistical research on modelling of the two-ways table data, here we will to investigate of the robustness of Row Column Interaction Model (RCIM). Row Column Interaction Model is a Reduced-Rank Vector Generalized Linear Models (RR-VGLM) class of modelling with the first linear predictor is modelled by the sum of the column effect, row effect, and interaction effect. The interaction effect was shown as a reduced-rank regression. We focused on outlying observations in the two ways data table. Outliers known as sample points that have unique characteristics, they differ from the majority of the whole sample. But there are some outliers that are difficult to identify due to the location and size of the data. Our previous proposed of handling outlier in Additive Main Effect and Multiplicative (AMMI) modelling by applying Robust Alternating Regression in Factor Analytics model. The two models will be compared in analysing two-ways table data that containing some outliers. In this research, two-ways table data are generated randomly follows normal distribution on Additive Main Effect and Multiplicative Interaction model by first two principal components (or AMMI2 for short), with two different types of outlier’s placement. The RCIM model seem provide a better result in fitting the data than Robust factor model, the RCIM model have smaller error, even for Normal distribution or Asymmetric Laplace Distribution (ALD).