An analysis of protein language model embeddings for fold prediction.

The identification of the protein fold class is a challenging problem in structural biology. Recent computational methods for fold prediction leverage deep learning techniques to extract protein fold-representative embeddings mainly using evolutionary information in the form of multiple sequence alignment (MSA) as input source. In contrast, protein language models (LM) have reshaped the field thanks to their ability to learn efficient protein representations (protein-LM embeddings) from purely sequential information in a self-supervised manner. In this paper, we analyze a framework for protein fold prediction using pre-trained protein-LM embeddings as input to several fine-tuning neural network models, which are supervisedly trained with fold labels. In particular, we compare the performance of six protein-LM embeddings: the long short-term memory-based UniRep and SeqVec, and the transformer-based ESM-1b, ESM-MSA, ProtBERT and ProtT5; as well as three neural networks: Multi-Layer Perceptron, ResCNN-BGRU (RBG) and Light-Attention (LAT). We separately evaluated the pairwise fold recognition (PFR) and direct fold classification (DFC) tasks on well-known benchmark datasets. The results indicate that the combination of transformer-based embeddings, particularly those obtained at amino acid level, with the RBG and LAT fine-tuning models performs remarkably well in both tasks. To further increase prediction accuracy, we propose several ensemble strategies for PFR and DFC, which provide a significant performance boost over the current state-of-the-art results. All this suggests that moving from traditional protein representations to protein-LM embeddings is a very promising approach to protein fold-related tasks.

Coronavirus disease 2019 is associated with long-term depressive symptoms in Spanish older adults with overweight/obesity and metabolic syndrome.

BACKGROUND: The coronavirus disease 2019 (COVID-19) has serious physiological and psychological consequences. The long-term (>12 weeks post-infection) impact of COVID-19 on mental health, specifically in older adults, is unclear. We longitudinally assessed the association of COVID-19 with depression symptomatology in community-dwelling older adults with metabolic syndrome within the framework of the PREDIMED-Plus cohort. METHODS: Participants (n = 5486) aged 55-75 years were included in this longitudinal cohort. COVID-19 status (positive/negative) determined by tests (e.g. polymerase chain reaction severe acute respiratory syndrome coronavirus 2, IgG) was confirmed via event adjudication (410 cases). Pre- and post-COVID-19 depressive symptomatology was ascertained from annual assessments conducted using a validated 21-item Spanish Beck Depression Inventory-II (BDI-II). Multivariable linear and logistic regression models assessed the association between COVID-19 and depression symptomatology. RESULTS: COVID-19 in older adults was associated with higher post-COVID-19 BDI-II scores measured at a median (interquartile range) of 29 (15-40) weeks post-infection [fully adjusted ß = 0.65 points, 95% confidence interval (CI) 0.15-1.15; p = 0.011]. This association was particularly prominent in women (ß = 1.38 points, 95% CI 0.44-2.33, p = 0.004). COVID-19 was associated with 62% increased odds of elevated depression risk (BDI-II ≥ 14) post-COVID-19 when adjusted for confounders (odds ratio; 95% CI 1.13-2.30, p = 0.008). CONCLUSIONS: COVID-19 was associated with long-term depression risk in older adults with overweight/obesity and metabolic syndrome, particularly in women. Thus, long-term evaluations of the impact of COVID-19 on mental health and preventive public health initiatives are warranted in older adults.

Association of adiposity and its changes over time with COVID-19 risk in older adults with overweight/obesity and metabolic syndrome: a longitudinal evaluation in the PREDIMED-Plus cohort.

BACKGROUND: Cross-sectionally, older age and obesity are associated with increased coronavirus disease-2019 (COVID-19) risk. We assessed the longitudinal associations of baseline and changes in adiposity parameters with COVID-19 incidence in older adults at high cardiovascular risk. METHODS: This analysis included 6874 men and women (aged 55-75 years) with overweight/obesity and metabolic syndrome in the PREDIMED-Plus lifestyle intervention trial for cardiovascular risk reduction. Body weight, body-mass-index (BMI), waist circumference, waist-to-height ratio (WHtR), and a body shape index (ABSI) were measured at baseline and annual follow-up visits. COVID-19 was ascertained by an independent Event Committee until 31 December 2021. Cox regression models were fitted to evaluate the risk of COVID-19 incidence based on baseline adiposity parameters measured 5-6 years before the pandemic and their changes at the visit prior to censoring. RESULTS: At the time of censoring, 653 incident COVID-19 cases occurred. Higher baseline body weight, BMI, waist circumference, and WHtR were associated with increased COVID-19 risk. During the follow-up, every unit increase in body weight (HRadj (95%CI): 1.01 (1.00, 1.03)) and BMI (HRadj: 1.04 (1.003, 1.08)) was associated with increased COVID-19 risk. CONCLUSIONS: In older adults with overweight/obesity, clinically significant weight loss may protect against COVID-19. TRIAL REGISTRATION: This study is registered at the International Standard Randomized Controlled Trial (ISRCT; http://www.isrctn.com/ISRCTN89898870 ).

Dietary intake of polychlorinated dibenzo-p-dioxins and furans, adiposity and obesity status.

INTRODUCTION: The principal source of exposure to Polychlorinated dibenzo-p-dioxins and polychlorinated dibenzo-p-furans (PCDD/Fs) in humans comes from food intake. PCDD/Fs, are a family of potential endocrine disruptors and have been associated with different chronic diseases such as diabetes and hypertension. However, studies assessing the relationship between dietary exposure to PCDD/Fs and adiposity or obesity status in a middle-aged population are limited. OBJECTIVE: To assess cross-sectionally and longitudinally the associations between estimated dietary intake (DI) of PCDD/Fs and body mass index (BMI), waist circumference, and the prevalence/incidence of obesity and abdominal obesity in a middle-aged population. METHODS: In 5899 participants aged 55-75 years (48% women) living with overweight/obesity from the PREDIMED-plus cohort, PCDD/Fs DI was estimated using a 143-item validated food-frequency questionnaire, and the levels of food PCDD/F expressed as Toxic Equivalents (TEQ). Consequently, cross-sectional and prospective associations between baseline PCDD/Fs DI (in pgTEQ/week) and adiposity or obesity status were assessed at baseline and after 1-year follow-up using multivariable cox, logistic or linear regression models. RESULTS: Compared to participants in the first PCDD/F DI tertile, those in the highest tertile presented a higher BMI (ß-coefficient [confidence interval]) (0.43kg/m2 [0.22; 0.64]; P-trend <0.001), a higher waist circumference (1.11 cm [0.55; 1.66]; P-trend <0.001), and a higher prevalence of obesity and abdominal obesity (1.05 [1.01; 1.09] and 1.02 [1.00; 1.03]; P-trend = 0.09 and 0.027, respectively). In the prospective analysis, participants in the top PCDD/F DI baseline tertile showed an increase in waist circumference compared with those in the first tertile after 1-year of follow-up (ß-coefficient 0.37 cm [0.06; 0.70]; P-trend = 0.015). CONCLUSION: Higher DI of PCDD/Fs was positively associated with adiposity parameters and obesity status at baseline and with changes in waist circumference after 1-year of follow-up in subjects living with overweight/obesity. Further large prospective studies using a different population with longer follow-up periods are warranted in the future to strengthen our results.

Carbon dioxide (CO2) emissions and adherence to Mediterranean diet in an adult population: the Mediterranean diet index as a pollution level index.

BACKGROUND: Research related to sustainable diets is is highly relevant to provide better understanding of the impact of dietary intake on the health and the environment. AIM: To assess the association between the adherence to an energy-restricted Mediterranean diet and the amount of CO2 emitted in an older adult population. DESIGN AND POPULATION: Using a cross-sectional design, the association between the adherence to an energy-reduced Mediterranean Diet (erMedDiet) score and dietary CO2 emissions in 6646 participants was assessed. METHODS: Food intake and adherence to the erMedDiet was assessed using validated food frequency questionnaire and 17-item Mediterranean questionnaire. Sociodemographic characteristics were documented. Environmental impact was calculated through greenhouse gas emissions estimations, specifically CO2 emissions of each participant diet per day, using a European database. Participants were distributed in quartiles according to their estimated CO2 emissions expressed in kg/day: Q1 (≤2.01 kg CO2), Q2 (2.02-2.34 kg CO2), Q3 (2.35-2.79 kg CO2) and Q4 (≥2.80 kg CO2). RESULTS: More men than women induced higher dietary levels of CO2 emissions. Participants reporting higher consumption of vegetables, fruits, legumes, nuts, whole cereals, preferring white meat, and having less consumption of red meat were mostly emitting less kg of CO2 through diet. Participants with higher adherence to the Mediterranean Diet showed lower odds for dietary CO2 emissions: Q2 (OR 0.87; 95%CI: 0.76-1.00), Q3 (OR 0.69; 95%CI: 0.69-0.79) and Q4 (OR 0.48; 95%CI: 0.42-0.55) vs Q1 (reference). CONCLUSIONS: The Mediterranean diet can be environmentally protective since the higher the adherence to the Mediterranean diet, the lower total dietary CO2 emissions. Mediterranean Diet index may be used as a pollution level index.

Metabolic syndrome criteria and severity and carbon dioxide (CO2) emissions in an adult population.

BACKGROUND: Metabolic syndrome (MetS) has become a growing risk factor of some non-communicable diseases. Increase of greenhouse gas emissions affects the planet. AIMS: To assess the association between MetS severity and amount of carbon dioxide (CO2) emitted in an adult population. DESIGN: Cross-sectional study (n = 6646; 55-76-year-old-men; 60-75-year-old-women with MetS). METHODS: Dietary habits were assessed using a pre-validated semi quantitative 143-item food frequency questionnaire. The amount of CO2 emitted due to the production of food consumed by person and day was calculated using a European database, and the severity of the MetS was calculated with the MetS Severity Score. RESULTS: Higher glycaemia levels were found in people with higher CO2 emissions. The risk of having high severe MetS was related to high CO2 emissions. CONCLUSIONS: Low CO2 emissions diet would help to reduce MetS severity. Advantages for both health and the environment were found following a more sustainable diet. TRIAL REGISTRATION: ISRCTN, ISRCTN89898870 . Registered 05 September 2013.

Unsupervised protein embeddings outperform hand-crafted sequence and structure features at predicting molecular function.

MOTIVATION: Protein function prediction is a difficult bioinformatics problem. Many recent methods use deep neural networks to learn complex sequence representations and predict function from these. Deep supervised models require a lot of labeled training data which are not available for this task. However, a very large amount of protein sequences without functional labels is available. RESULTS: We applied an existing deep sequence model that had been pretrained in an unsupervised setting on the supervised task of protein molecular function prediction. We found that this complex feature representation is effective for this task, outperforming hand-crafted features such as one-hot encoding of amino acids, k-mer counts, secondary structure and backbone angles. Also, it partly negates the need for complex prediction models, as a two-layer perceptron was enough to achieve competitive performance in the third Critical Assessment of Functional Annotation benchmark. We also show that combining this sequence representation with protein 3D structure information does not lead to performance improvement, hinting that 3D structure is also potentially learned during the unsupervised pretraining. AVAILABILITY AND IMPLEMENTATION: Implementations of all used models can be found at https://github.com/stamakro/GCN-for-Structure-and-Function. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Integrative development of a short screening questionnaire of highly processed food consumption (sQ-HPF).

BACKGROUND: Recent lifestyle changes include increased consumption of highly processed foods (HPF), which has been associated with an increased risk of non-communicable diseases (NCDs). However, nutritional information relies on the estimation of HPF consumption from food-frequency questionnaires (FFQ) that are not explicitly developed for this purpose. We aimed to develop a short screening questionnaire of HPF consumption (sQ-HPF) that integrates criteria from the existing food classification systems. METHODS: Data from 4400 participants (48.1% female and 51.9% male, 64.9 ± 4.9 years) of the Spanish PREDIMED-Plus ("PREvention with MEDiterranean DIet") trial were used for this analysis. Items from the FFQ were classified according to four main food processing-based classification systems (NOVA, IARC, IFIC and UNC). Participants were classified into tertiles of HPF consumption according to each system. Using binomial logistic regression, food groups associated with agreement in the highest tertile for at least two classification systems were chosen as items for the questionnaire. ROC analysis was used to determine cut-off points for the frequency of consumption of each item, from which a score was calculated. Internal consistency of the questionnaire was assessed through exploratory factor analysis (EFA) and Cronbach's analysis, and agreement with the four classifications was assessed with weighted kappa coefficients. RESULTS: Regression analysis identified 14 food groups (items) associated with high HPF consumption for at least two classification systems. EFA showed that items were representative contributors of a single underlying factor, the "HPF dietary pattern" (factor loadings around 0.2). We constructed a questionnaire asking about the frequency of consumption of those items. The threshold frequency of consumption was selected using ROC analysis. Comparison of the four classification systems and the sQ-HPF showed a fair to high agreement. Significant changes in lifestyle characteristics were detected across tertiles of the sQ-HPF score. Longitudinal changes in HPF consumption were also detected by the sQ-HPF, concordantly with existing classification systems. CONCLUSIONS: We developed a practical tool to measure HPF consumption, the sQ-HPF. This may be a valuable instrument to study its relationship with NCDs. TRIAL REGISTRATION: Retrospectively registered at the International Standard Randomized Controlled Trial Registry ( ISRCTN89898870 ) on July 24, 2014.

Association between the Prime Diet Quality Score and depressive symptoms in a Mediterranean population with metabolic syndrome. Cross-sectional and 2-year follow-up assessment from PREDIMED-PLUS study.

The burden of depression is increasing worldwide, specifically in older adults. Unhealthy dietary patterns may partly explain this phenomenon. In the Spanish PREDIMED-Plus study, we explored (1) the cross-sectional association between the adherence to the Prime Diet Quality Score (PDQS), an a priori-defined high-quality food pattern, and the prevalence of depressive symptoms at baseline (cross-sectional analysis) and (2) the prospective association of baseline PDQS with changes in depressive symptomatology after 2 years of follow-up. After exclusions, we assessed 6612 participants in the cross-sectional analysis and 5523 participants in the prospective analysis. An energy-adjusted high-quality dietary score (PDQS) was assessed using a validated FFQ. The cross-sectional association between PDQS and the prevalence of depression or presence of depressive symptoms and the prospective changes in depressive symptoms were evaluated through multivariable regression models (logistic and linear models and mixed linear-effects models). PDQS was inversely associated with depressive status in the cross-sectional analysis. Participants in the highest quintile of PDQS (Q5) showed a significantly reduced odds of depression prevalence as compared to participants in the lowest quartile of PDQS (Q1) (OR (95 %) CI = 0·82 (0·68, 0·98))). The baseline prevalence of depression decreased across PDQS quintiles (Pfor trend = 0·015). A statistically significant association between PDQS and changes in depressive symptoms after 2-years follow-up was found (ß (95 %) CI = -0·67 z-score (-1·17, -0·18). A higher PDQS was cross-sectionally related to a lower depressive status. Nevertheless, the null finding in our prospective analysis raises the possibility of reverse causality. Further prospective investigation is required to ascertain the association between PDQS and changes in depressive symptoms along time.

Prospective associations between a priori dietary patterns adherence and kidney function in an elderly Mediterranean population at high cardiovascular risk.

PURPOSE: To assess the association between three different a priori dietary patterns adherence (17-item energy reduced-Mediterranean Diet (MedDiet), Trichopoulou-MedDiet and Dietary Approach to Stop Hypertension (DASH)), as well as the Protein Diet Score and kidney function decline after one year of follow-up in elderly individuals with overweight/obesity and metabolic syndrome (MetS). METHODS: We prospectively analyzed 5675 participants (55-75 years) from the PREDIMED-Plus study. At baseline and at one year, we evaluated the creatinine-based estimated glomerular filtration rate (eGFR) and food-frequency questionnaires-derived dietary scores. Associations between four categories (decrease/maintenance and tertiles of increase) of each dietary pattern and changes in eGFR (ml/min/1.73m2) or ≥ 10% eGFR decline were assessed by fitting multivariable linear or logistic regression models, as appropriate. RESULTS: Participants in the highest tertile of increase in 17-item erMedDiet Score showed higher upward changes in eGFR (ß: 1.87 ml/min/1.73m2; 95% CI: 1.00-2.73) and had lower odds of ≥ 10% eGFR decline (OR: 0.62; 95% CI: 0.47-0.82) compared to individuals in the decrease/maintenance category, while Trichopoulou-MedDiet and DASH Scores were not associated with any renal outcomes. Those in the highest tertile of increase in Protein Diet Score had greater downward changes in eGFR (ß: - 0.87 ml/min/1.73m2; 95% CI: - 1.73 to - 0.01) and 32% higher odds of eGFR decline (OR: 1.32; 95% CI: 1.00-1.75). CONCLUSIONS: Among elderly individuals with overweight/obesity and MetS, only higher upward change in the 17-item erMedDiet score adherence was associated with better kidney function after one year. However, increasing Protein Diet Score appeared to have an adverse impact on kidney health. TRIAL REGISTRATION NUMBER: ISRCTN89898870 (Data of registration: 2014).

Factors associated with successful dietary changes in an energy-reduced Mediterranean diet intervention: a longitudinal analysis in the PREDIMED-Plus trial.

PURPOSE: Long-term nutrition trials may fail to respond to their original hypotheses if participants do not comply with the intended dietary intervention. We aimed to identify baseline factors associated with successful dietary changes towards an energy-reduced Mediterranean diet (MedDiet) in the PREDIMED-Plus randomized trial. METHODS: Longitudinal analysis of 2985 participants (Spanish overweight/obese older adults with metabolic syndrome) randomized to the active intervention arm of the PREDIMED-Plus trial. Dietary changes were assessed with a 17-item energy-reduced MedDiet questionnaire after 6 and 12 months of follow-up. Successful compliance was defined as dietary changes from baseline of ≥ 5 points for participants with baseline scores < 13 points or any increase if baseline score was ≥ 13 points. We conducted crude and adjusted multivariable logistic regression models to identify baseline factors related to compliance. RESULTS: Consistent factors independently associated with successful dietary change at both 6 and 12 months were high baseline perceived self-efficacy in modifying diet (OR6-month: 1.51, 95% CI 1.25-1.83; OR12-month: 1.66, 95% CI 1.37-2.01), higher baseline fiber intake (OR6-month: 1.62, 95% CI 1.07-2.46; OR12-month: 1.62, 95% CI 1.07-2.45), having > 3 chronic conditions (OR6-month: 0.65, 95% CI 0.53-0.79; OR12-month: 0.76, 95% CI 0.62-0.93), and suffering depression (OR6-month: 0.80, 95% CI 0.64-0.99; OR12-month: 0.71, 95% CI 0.57-0.88). CONCLUSION: Our results suggested that recruitment of individuals with high perceived self-efficacy to dietary change, and those who initially follow diets relatively richer in fiber may lead to greater changes in nutritional recommendations. Participants with multiple chronic conditions, specifically depression, should receive specific tailored interventions. TRIAL REGISTRATION: ISRCTN registry 89898870, 24th July 2014 retrospectively registered http://www.isrctn.com/ISRCTN89898870 .

Pro-vegetarian food patterns and cardiometabolic risk in the PREDIMED-Plus study: a cross-sectional baseline analysis.

PURPOSE: We explored the cross-sectional association between the adherence to three different provegetarian (PVG) food patterns defined as general (gPVG), healthful (hPVG) and unhealthful (uPVG), and the cardiometabolic risk in adults with metabolic syndrome (MetS) of the PREDIMED-Plus randomized intervention study. METHODS: We performed a cross-sectional analysis of baseline data from 6439 participants of the PREDIMED-Plus randomized intervention study. The gPVG food pattern was built by positively scoring plant foods (vegetables/fruits/legumes/grains/potatoes/nuts/olive oil) and negatively scoring, animal foods (meat and meat products/animal fats/eggs/fish and seafood/dairy products). The hPVG and uPVG were generated from the gPVG by adding four new food groups (tea and coffee/fruit juices/sugar-sweetened beverages/sweets and desserts), splitting grains and potatoes and scoring them differently. Multivariable-adjusted robust linear regression using MM-type estimator was used to assess the association between PVG food patterns and the standardized Metabolic Syndrome score (MetS z-score), a composed index that has been previously used to ascertain the cardiometabolic risk, adjusting for potential confounders. RESULTS: A higher adherence to the gPVG and hPVG was associated with lower cardiometabolic risk in multivariable models. The regression coefficients for 5th vs. 1st quintile were - 0.16 (95% CI: - 0.33 to 0.01) for gPVG (p trend: 0.015), and - 0.23 (95% CI: - 0.41 to - 0.05) for hPVG (p trend: 0.016). In contrast, a higher adherence to the uPVG was associated with higher cardiometabolic risk, 0.21 (95% CI: 0.04 to 0.38) (p trend: 0.019). CONCLUSION: Higher adherence to gPVG and hPVG food patterns was generally associated with lower cardiovascular risk, whereas higher adherence to uPVG was associated to higher cardiovascular risk.

Vitamin K dietary intake is associated with cognitive function in an older adult Mediterranean population.

BACKGROUND: In the last years, evidence that dietary vitamin K could have a role in the cognitive domain has increased. However, data from large trials are limited. The objective of this study was to assess the association of 2 year changes in the dietary intake of vitamin K with cognitive function measured through neuropsychological performance tests. METHODS: In 5,533 participants of the multicentre PREDIMED-Plus study (48.1% women, age 65.1 ± 4.9 years with overweight/obesity and metabolic syndrome), we assessed the adjusted odds ratios of cognitive function decline according to 2 year changes in vitamin K intake. Participants answered a battery of cognitive function tests and Food Frequency Questionnaires (FFQs) in order to estimate the vitamin K dietary intake. RESULTS: After adjusting for potential cofounders, the highest tertile of change of dietary vitamin K intake (median [IQR]; 194.4 µg/d [120.9, 373.1]) was inversely associated with a Mini-Mental State Examination (MMSE) score ≤24 (OR [95% CI]; 0.53 [0.35, 0.79] P for trend = 0.002) compared with a decrease in the intake of vitamin K (median [IQR]; -97.8 µg/d [-292.8, -51.5]). A significant positive association between changes in dietary vitamin K intake and the semantic verbal fluency test scores (OR [95% CI]; 0.69 [0.51, 0.94] P for trend = 0.019) was found. CONCLUSIONS: An increase of the intake of dietary vitamin K was associated with better cognitive function scores, independently of recognised risk factors for cognitive decline, in an older adult Mediterranean population with high cardiovascular risk.

Dietary diversity and depression: cross-sectional and longitudinal analyses in Spanish adult population with metabolic syndrome. Findings from PREDIMED-Plus trial.

OBJECTIVE: To examine the cross-sectional and longitudinal (2-year follow-up) associations between dietary diversity (DD) and depressive symptoms. DESIGN: An energy-adjusted dietary diversity score (DDS) was assessed using a validated FFQ and was categorised into quartiles (Q). The variety in each food group was classified into four categories of diversity (C). Depressive symptoms were assessed with Beck Depression Inventory-II (Beck II) questionnaire and depression cases defined as physician-diagnosed or Beck II >= 18. Linear and logistic regression models were used. SETTING: Spanish older adults with metabolic syndrome (MetS). PARTICIPANTS: A total of 6625 adults aged 55-75 years from the PREDIMED-Plus study with overweight or obesity and MetS. RESULTS: Total DDS was inversely and statistically significantly associated with depression in the cross-sectional analysis conducted; OR Q4 v. Q1 = 0·76 (95 % CI (0·64, 0·90)). This was driven by high diversity compared to low diversity (C3 v. C1) of vegetables (OR = 0·75, 95 % CI (0·57, 0·93)), cereals (OR = 0·72 (95 % CI (0·56, 0·94)) and proteins (OR = 0·27, 95 % CI (0·11, 0·62)). In the longitudinal analysis, there was no significant association between the baseline DDS and changes in depressive symptoms after 2 years of follow-up, except for DD in vegetables C4 v. C1 = (ß = 0·70, 95 % CI (0·05, 1·35)). CONCLUSIONS: According to our results, DD is inversely associated with depressive symptoms, but eating more diverse does not seem to reduce the risk of future depression. Additional longitudinal studies (with longer follow-up) are needed to confirm these findings.

FoldHSphere: deep hyperspherical embeddings for protein fold recognition.

BACKGROUND: Current state-of-the-art deep learning approaches for protein fold recognition learn protein embeddings that improve prediction performance at the fold level. However, there still exists aperformance gap at the fold level and the (relatively easier) family level, suggesting that it might be possible to learn an embedding space that better represents the protein folds. RESULTS: In this paper, we propose the FoldHSphere method to learn a better fold embedding space through a two-stage training procedure. We first obtain prototype vectors for each fold class that are maximally separated in hyperspherical space. We then train a neural network by minimizing the angular large margin cosine loss to learn protein embeddings clustered around the corresponding hyperspherical fold prototypes. Our network architectures, ResCNN-GRU and ResCNN-BGRU, process the input protein sequences by applying several residual-convolutional blocks followed by a gated recurrent unit-based recurrent layer. Evaluation results on the LINDAHL dataset indicate that the use of our hyperspherical embeddings effectively bridges the performance gap at the family and fold levels. Furthermore, our FoldHSpherePro ensemble method yields an accuracy of 81.3% at the fold level, outperforming all the state-of-the-art methods. CONCLUSIONS: Our methodology is efficient in learning discriminative and fold-representative embeddings for the protein domains. The proposed hyperspherical embeddings are effective at identifying the protein fold class by pairwise comparison, even when amino acid sequence similarities are low.

Effect of an Intensive Weight-Loss Lifestyle Intervention on Kidney Function: A Randomized Controlled Trial.

INTRODUCTION: Large randomized trials testing the effect of a multifactorial weight-loss lifestyle intervention including Mediterranean diet (MedDiet) on renal function are lacking. Here, we evaluated the 1-year efficacy of an intensive weight-loss intervention with an energy-reduced MedDiet (erMedDiet) plus increased physical activity (PA) on renal function. METHODS: Randomized controlled "PREvención con DIeta MEDiterránea-Plus" (PREDIMED-Plus) trial is conducted in 23 Spanish centers comprising 208 primary care clinics. Overweight/obese (n = 6,719) adults aged 55-75 years with metabolic syndrome were randomly assigned (1:1) to an intensive weight-loss lifestyle intervention with an erMedDiet, PA promotion, and behavioral support (intervention) or usual-care advice to adhere to an energy-unrestricted MedDiet (control) between September 2013 and December 2016. The primary outcome was 1-year change in estimated glomerular filtration rate (eGFR). Secondary outcomes were changes in urine albumin-to-creatinine ratio (UACR), incidence of moderately/severely impaired eGFR (<60 mL/min/1.73 m2) and micro- to macroalbuminuria (UACR ≥30 mg/g), and reversion of moderately (45 to <60 mL/min/1.73 m2) to mildly impaired GFR (60 to <90 mL/min/1.73 m2) or micro- to macroalbuminuria. RESULTS: After 1 year, eGFR declined by 0.66 and 1.25 mL/min/1.73 m2 in the intervention and control groups, respectively (mean difference, 0.58 mL/min/1.73 m2; 95% CI: 0.15-1.02). There were no between-group differences in mean UACR or micro- to macroalbuminuria changes. Moderately/severely impaired eGFR incidence and reversion of moderately to mildly impaired GFR were 40% lower (HR 0.60; 0.44-0.82) and 92% higher (HR 1.92; 1.35-2.73), respectively, in the intervention group. CONCLUSIONS: The PREDIMED-Plus lifestyle intervention approach may preserve renal function and delay CKD progression in overweight/obese adults.

Dietary vitamin D intake and colorectal cancer risk: a longitudinal approach within the PREDIMED study.

PURPOSE: We evaluated whether the intake of dietary vitamin D is associated with the incidence of both colorectal cancer (CRC) and colon cancer in the framework of the PREDIMED cohort of older adults at high cardiovascular risk. METHODS: We analyzed data from 7216 men and women (55-80 years) without CRC at baseline from the PREvención con DIeta MEDiterránea study. Baseline consumption of vitamin D was assessed using a validated 137-item food frequency questionnaire. Cox proportional hazards ratios (HRs) of CRC and colon cancer incidence were estimated for quartiles and per 1-SD of baseline vitamin D intake. RESULTS: During a median follow-up of 6 years, we documented 97 incident CRC cases after the exclusion of subjects with no baseline dietary data and/or outliers of energy intake. A non-significant HRs and 95% confidence intervals (CIs) of CRC for the comparison of extreme quartiles (4th vs 1st) of vitamin D intake were observed [0.55 (0.30-1.00), P for trend = 0.072], whereas it was significant for colon cancer incidence alone [0.44 (0.22-0.90), P for trend = 0.032]. However, this association became significant in CRC and colon cancer incidence, after excluding 391 subjects consuming baseline vitamin D and/or calcium medication or prescribed supplements [0.52 (0.28-0.96) and 0.41 (0.12-0.85), respectively]. CONCLUSION: A higher dietary intake of vitamin D was significantly associated with a reduced CRC risk in individuals at high cardiovascular risk.

Dietary folate intake and metabolic syndrome in participants of PREDIMED-Plus study: a cross-sectional study.

PURPOSE: We examined the association between dietary folate intake and a score of MetS (metabolic syndrome) and its components among older adults at higher cardiometabolic risk participating in the PREDIMED-Plus trial. METHODS: A cross-sectional analysis with 6633 with overweight/obesity participants with MetS was conducted. Folate intake (per 100 mcg/day and in quintiles) was estimated using a validated food frequency questionnaire. We calculated a MetS score using the standardized values as shown in the formula: [(body mass index + waist-to-height ratio)/2] + [(systolic blood pressure + diastolic blood pressure)/2] + plasma fasting glucose-HDL cholesterol + plasma triglycerides. The MetS score as continuous variable and its seven components were the outcome variables. Multiple robust linear regression using MM-type estimator was performed to evaluate the association adjusting for potential confounders. RESULTS: We observed that an increase in energy-adjusted folate intake was associated with a reduction of MetS score (ß for 100 mcg/day = - 0.12; 95% CI: - 0.19 to - 0.05), and plasma fasting glucose (ß = - 0.03; 95% CI: - 0.05 to - 0.02) independently of the adherence to Mediterranean diet and other potential confounders. We also found a positive association with HDL-cholesterol (ß = 0.07; 95% CI: 0.04-0.10). These associations were also observed when quintiles of energy-adjusted folate intake were used instead. CONCLUSION: This study suggests that a higher folate intake may be associated with a lower MetS score in older adults, a lower plasma fasting glucose, and a greater HDL cholesterol in high-risk cardio-metabolic subjects.

Association between coffee consumption and total dietary caffeine intake with cognitive functioning: cross-sectional assessment in an elderly Mediterranean population.

PURPOSE: Coffee is rich in compounds such as polyphenols, caffeine, diterpenes, melanoidins and trigonelline, which can stimulate brain activity. Therefore, the possible association of coffee consumption with cognition is of considerable research interest. In this paper, we assess the association of coffee consumption and total dietary caffeine intake with the risk of poor cognitive functioning in a population of elderly overweight/obese adults with metabolic syndrome (MetS). METHODS: PREDIMED-plus study participants who completed the Mini-Mental State Examination test (MMSE) (n = 6427; mean age = 65 ± 5 years) or a battery of neuropsychological tests were included in this cross-sectional analysis. Coffee consumption and total dietary caffeine intake were assessed at baseline using a food frequency questionnaire. Logistic regression models were fitted to evaluate the association between total, caffeinated and decaffeinated coffee consumption or total dietary caffeine intake and cognitive impairment. RESULTS: Total coffee consumers and caffeinated coffee consumers had better cognitive functioning than non-consumers when measured by the MMSE and after adjusting for potential confounders (OR 0.63; 95% CI 0.44-0.90 and OR 0.56; 95% CI 0.38-0.83, respectively). Results were similar when cognitive performance was measured using the Clock Drawing Test (CDT) and Trail Making Test B (TMT-B). These associations were not observed for decaffeinated coffee consumption. Participants in the highest tertile of total dietary caffeine intake had lower odds of poor cognitive functioning than those in the reference tertile when screened by the MMSE (OR 0.64; 95% CI 0.47-0.87) or other neurophysiological tests evaluating a variety of cognitive domains (i.e., CDT and TMT-A). CONCLUSIONS: Coffee consumption and total dietary caffeine intake were associated with better cognitive functioning as measured by various neuropsychological tests in a Mediterranean cohort of elderly individuals with MetS. TRIAL REGISTRATION: ISRCTN89898870. Registration date: July 24, 2014.

Physical activity and metabolic syndrome severity among older adults at cardiovascular risk: 1-Year trends.

BACKGROUND AND AIMS: Modifiable lifestyle factors, such as physical activity (PA) and Mediterranean diet (MD), decrease metabolic syndrome (MetS). The aim was to assess 1-year changes of leisure-time physical activity (LTPA), sedentary behavior, and diet quality according to MetS severity in older population at high cardiovascular risk. METHODS AND RESULTS: Prospective analysis of 55-75-year-old 4359 overweight/obese participants with MetS (PREDIMED-Plus trial) categorized in tertiles according to 1-year changes of a validated MetS severity score (MetSSS). Anthropometrics, visceral adiposity index, triglycerides and glucose index, dietary nutrient intake, biochemical marker levels, dietary inflammatory index, and depression symptoms were measured. Diet quality was assessed by 17-item MD questionnaire. PAs were self-reported using the Minnesota-REGICOR Short Physical Activity Questionnaire and 30-s chair stand test. Sedentary behaviors were measured using the Spanish version of the Nurses' Health Study questionnaire. After 1-year follow-up, decreasing MetSSS was associated with an anti-inflammatory dietary pattern, high intake of vegetables, fruits, legumes, nuts, whole grain cereals, white fish, and bluefish and low intake of refined cereals, red and processed meat, cookies/sweets, and snacks/ready-to-eat-meals. It resulted in high intake of polyunsaturated fatty acids, omega-3 fatty acids, protein, fiber, vitamins B1, B6, B9, C, D, potassium, magnesium, and phosphorus and low glycemic index and saturated fatty acid, trans fatty acid, and carbohydrates intake. Regarding PA and sedentary behavior, decreasing MetSSS was associated with increased moderate-to-vigorous LTPA, chair stand test, and decreased sedentary and TV-viewing time. CONCLUSION: Decreasing MetSSS was associated with an anti-inflammatory dietary pattern, high LTPA, high MD adherence, low sedentary time, and low depression risk.