The use of cluster analysis to derive dietary patterns: methodological considerations, reproducibility, validity and the effect of energy mis-reporting

Dietary Patterns Among US Children: A Cluster Analysis

BACKGROUND

Most children in the United States consume low-quality diets. Identifying children's dietary patterns and their association with sociodemographic characteristics is important for designing tailored youth dietary interventions.

OBJECTIVE

This study's objective was to use cluster analysis to investigate children's dietary patterns and these patterns' associations with sociodemographic characteristics.

DESIGN

Data from two cycles (2015-2016 and 2017-2018) of the National Health and Nutrition Examination Survey were evaluated to examine dietary patterns.

PARTICIPANTS AND SETTING

Participants included 3,044 US youth aged 2 to 11 years who completed at least 1 valid 24-hour diet recall.

MAIN OUTCOME MEASURES

The main outcome measures were Healthy Eating Index (HEI) 2015 component and composite scores.

STATISTICAL ANALYSES PERFORMED

A cluster analysis was performed on standardized scores of 11 components of the HEI-2015 to identify dietary patterns. One logistic analysis combined the two higher-HEI score clusters and the 2 lower-HEI score clusters to form a 3-category variable of higher-, medium-, and lower-HEI score clusters. Another logistic analysis contrasted 2 higher- and then the 2 lower-HEI clusters with each other to examine sociodemographic factors contributing to cluster membership.

RESULTS

Five clusters were identified, each displaying a distinct dietary pattern. Older, non-Hispanic Black, and overweight children had higher odds of being in the higher-HEI clusters than the medium-HEI cluster. Being older and non-Hispanic Black were linked to higher odds of being in the lower-HEI clusters than the medium-HEI cluster. Conversely, being Mexican American and living with a college-educated reference person were associated with lower odds of being in the lower-HEI clusters compared with the medium-HEI cluster. Among the higher-HEI clusters, Mexican American and Asian American children had higher odds of being in the Pescatarian cluster. Among the lower-HEI clusters, children who were racially or ethnically minoritized had lower odds of being in the Excess Sugar cluster.

CONCLUSIONS

Children in this study displayed different dietary patterns, with key sociodemographic variation.

A Dietary Pattern With More Fruits and Vegetables in Children of Mothers Who Immigrated to the United States From Latin America Is Associated With Healthful Nutrient Intake and Weight Status

BACKGROUND

Children from families who immigrated to the United States may consume a lower-quality diet compared with their US-born peers. However, specific features of their dietary patterns, which could be a focus for improving diet quality, are not well-studied.

OBJECTIVE

The aim of this study was to examine dietary patterns that distinguish interpersonal variability in dietary intake and explore the association of dietary patterns with nutrient intake and weight status.

DESIGN

This study was a cross-sectional analysis of baseline data from the Live Well randomized controlled trial collected between 2009 and 2010.

PARTICIPANTS/SETTING

Study participants included 313 children (3-12 years) whose mothers immigrated to the United States from Latin America and resided in Somerville, Massachusetts.

MAIN OUTCOME MEASURES

Dietary patterns (derived using principal component analysis); nutrient intake (derived from dietary data collected using the Block Food Screener); and weight status (categorized using body mass index z score based on measured height and weight).

STATISTICAL ANALYSES PERFORMED

Nutrient intake levels were compared across quartiles for dietary patterns using analysis of covariance. Multivariable logistic or linear regression models were used to determine the associations between dietary pattern scores and odds of overweight or obesity or z scores.

RESULTS

Two dietary patterns emerged: "fruits and vegetables" and "meat and sweets." Highest adherence to the fruits and vegetables pattern was associated with more healthful nutrient intake and lower odds of having overweight or obesity (odds ratio 0.37; 95% CI 0.16 to 0.98), but not body mass index z score (β = -.07; 95% CI -.51 to 0.36) compared with the lowest adherence. Adherence to the meat and sweets pattern was associated with less healthful nutrient intake but not with the odds of experiencing overweight or obesity (odds ratio 0.48; 95% CI 0.16 to 1.46).

CONCLUSIONS

A healthful dietary pattern in children of families who immigrated to the United States from Latin America may include a variety of fruits and vegetables. Longitudinal studies should be conducted to further assess the role of dietary intake patterns on the health of these children.

The Brazilian Cardioprotective Nutritional (BALANCE) Program improves diet quality in patients with established cardiovascular disease: Results from a multicenter randomized controlled trial

Dietary modifications are essential strategies for cardiovascular disease prevention. However, studies are needed to investigate the diet quality of individuals undergoing secondary prevention in cardiology and who received dietary intervention based on cardiovascular disease management. We prospectively evaluated the diet quality in the Brazilian Cardioprotective Nutritional Program Trial (BALANCE Program Trial). We hypothesized that the BALANCE Program could improve patients' dietary pattern according to different indices of diet quality such as the Dietary Inflammatory Index (DII); the dietary total antioxidant capacity; overall, healthful, and unhealthful Plant-Based Diet Index (PDI, hPDI, and uPDI, respectively); and modified Alternative Healthy Eating Index (mAHEI). This multicenter randomized, controlled trial included patients aged ≥45 years randomly assigned to either the experimental or control group. Data from 2185 participants at baseline and after 12, 24, 36, and 48 months showed that the intervention group (n = 1077) had lower mean values of DII and higher dietary total antioxidant capacity, PDI, hPDI, and mAHEI than the control group. The results also showed differences between the follow-up times for DII, hPDI, and uPDI (48 months vs baseline) and for PDI and mAHEI (24 months vs baseline), regardless of group. The interaction analysis demonstrated that the intervention group showed better results than the control group at 12, 24, 36, and 48 months for the DII and at months 12, 36, and 48 for the mAHEI. Our results provide prospective evidence that the BALANCE Program improved the diet quality in those in secondary cardiovascular prevention according to different indices, with the intervention group showing better results than the control group.

Comparing the Associations of Dietary Patterns Identified through Principal Component Analysis and Cluster Analysis with Colorectal Cancer Risk: A Large Case-Control Study in China

Examining the association between dietary patterns and colorectal cancer (CRC) risk can provide valuable insights beyond the assessment of individual foods or nutrients. However, there is a lack of in-depth analysis of dietary patterns and CRC risk in Chinese populations, and few studies have compared dietary patterns derived from different posteriori methods with the aim of predicting disease risk. The aim of this study was to derive dietary patterns using both principal component analysis (PCA) and cluster analysis (CA) and to assess their respective associations with CRC risk. A large-scale case-control study was conducted in Guangdong Province, China, including 2799 incident colorectal cancer cases and an equal number of frequency-matched controls. Dietary intake information was gathered through the use of a validated food frequency questionnaire. PCA and CA were used to derive dietary patterns. A multivariable logistic regression model was used to calculate the adjusted odds ratio (aOR) and 95% confidence interval (CI). Four major dietary patterns were identified by PCA. CA identified two dietary patterns, referred to as the "Balanced dietary pattern" and the "Refined grain dietary pattern". Notably, there were significant inverse associations between the milk-egg-nut-soy dietary pattern (aOR, 0.51; 95% CI, 0.42, 0.62), the vegetable-fruit dietary pattern (aOR, 0.61; 95%CI, 0.51, 0.74), and the poultry-fish dietary pattern (aOR, 0.81; 95%CI, 0.68, 0.97) and CRC risk. However, the red meat-preserved food dietary pattern was associated with an increased risk of CRC (aOR, 2.99; 95%CI, 2.43, 3.67). When compared with the Refined grain dietary pattern, the Balanced dietary pattern showed a decreased risk of CRC (aOR, 0.59; 95%CI, 0.52, 0.66). The results from the comparison of the two methods indicate that both CA and PCA derived remarkably similar patterns. The combined use of PCA and CA identified consistent underlying patterns, showing comparable associations with CRC risk. These findings suggest that individuals who prefer dietary patterns characterized by a high intake of red meat, preserved food, and refined grains should be cautious about their increased CRC risk. Conversely, dietary patterns rich in fruits, vegetables, and high-quality protein sources are advisable for the prevention of CRC in the Chinese population.

Clusters of carbohydrate-rich foods and associations with type 2 diabetes incidence: a prospective cohort study

BACKGROUND

About one in ten adults are living with diabetes worldwide. Intake of carbohydrates and carbohydrate-rich foods are often identified as modifiable risk factors for incident type 2 diabetes. However, strong correlation between food variables can make it difficult to identify true associations. The purpose of this study was to identify clusters of carbohydrate-rich foods and analyse their associations with type 2 diabetes incidence in the Malmö Diet and Cancer Study cohort in southern Sweden.

METHODS

Dietary intake of 26 622 participants was assessed using a validated three-part diet history method: a 7-day food diary, a 168-item food frequency questionnaire, and a 60-minute interview. K-means clustering analysis identified five clusters from 21 food variables. The Cox proportional hazard regression model was applied to calculate hazard ratios (HR) and 95% confidence intervals (CI) of the association between clusters and incident type 2 diabetes.

RESULTS

The cluster analysis resulted in five clusters; high vegetables/low added sugar, high sugar-sweetened beverages, high juice, high fruit, and high refined carbohydrates/low fruit & vegetables (reference). During mean follow-up of 18 years, 4046 type 2 diabetes cases were identified. After adjustment for potential confounding (including lifestyle, body mass index, and diet), a high fruit cluster (HR 0.86; 95% CI 0.78, 0.94) was inversely associated with type 2 diabetes compared to the reference cluster. No other significant associations were identified.

CONCLUSIONS

A dietary pattern defined by a high intake of fruits was associated with a lower incidence of type 2 diabetes. The findings provide additional evidence of a potential protective effect from fruit intake in reducing type 2 diabetes risk. Future studies are needed to explore this association further.

Population food intake clusters and cardiovascular disease incidence: a Bayesian quantifying of a prospective population-based cohort study in a low and middle-income country

Aims

This study was designed to explore the relationship between cardiovascular disease incidence and population clusters, which were established based on daily food intake.

Methods

The current study examined 5,396 Iranian adults (2,627 males and 2,769 females) aged 35 years and older, who participated in a 10-year longitudinal population-based study that began in 2001. The frequency of food group consumption over the preceding year (daily, weekly, or monthly) was assessed using a 49-item qualitative food frequency questionnaire (FFQ) administered via a face-to-face interview conducted by an expert dietitian. Participants were clustered based on their dietary intake by applying the semi-parametric Bayesian approach of the Dirichlet Process. In this approach, individuals with the same multivariate distribution based on dietary intake were assigned to the same cluster. The association between the extracted population clusters and the incidence of cardiovascular diseases was examined using Cox proportional hazard models.

Results

In the 10-year follow-up, 741 participants (401 men and 340 women) were diagnosed with cardiovascular diseases. Individuals were categorized into three primary dietary clusters: healthy, unhealthy, and mixed. After adjusting for potential confounders, subjects in the unhealthy cluster exhibited a higher risk for cardiovascular diseases [Hazard Ratio (HR): 2.059; 95% CI: 1.013, 4.184] compared to those in the healthy cluster. In the unadjusted model, individuals in the mixed cluster demonstrated a higher risk for cardiovascular disease than those in the healthy cluster (HR: 1.515; 95% CI: 1.097, 2.092). However, this association was attenuated after adjusting for potential confounders (HR: 1.145; 95% CI: 0.769, 1.706).

Conclusion

The results have shown that individuals within an unhealthy cluster have a risk that is twice as high for the incidence of cardiovascular diseases. However, these associations need to be confirmed through further prospective investigations.

Post Hoc Analysis of a Randomized Controlled Trial on Fasting and Plant-Based Diet in Rheumatoid Arthritis (NutriFast): Nutritional Supply and Impact on Dietary Behavior

This study aimed at comparing the nutrient supply and dietary behaviors during a plant-based diet (PBD) combined with time-restricted eating (TRE) to standard dietary recommendations in rheumatoid arthritis patients. In this open-label, randomized, controlled clinical trial, patients were assigned to either a 7-day fast followed by an 11-week PBD including TRE (A) or a 12-week anti-inflammatory diet following official German guidelines (German Nutrition Society, DGE) (B). Dietary habits were assessed by 3-day food records at weeks -1, 4 and 9 and food frequency questionnaires. 41 out of 53 participants were included in a post-hoc per protocol analysis. Both groups had similar energy, carbohydrate, sugar, fiber and protein intake at week 4. Group A consumed significantly less total saturated fat than group B (15.9 ± 7.7 vs. 23.2 ± 10.3 g/day; p = 0.02). Regarding micronutrients, group B consumed more vitamin A, B₁₂, D, riboflavin and calcium (each p ≤ 0.02). Zinc and calcium were below recommended intakes in both groups. Cluster analysis did not show clear group allocation after three months. Hence, dietary counselling for a PBD combined with TRE compared to a standard anti-inflammatory diet does not seem to lead to two different dietary clusters, i.e., actual different dietary behaviors as expected. Larger confirmatory studies are warranted to further define dietary recommendations for RA.

Dietary patterns and indicators of cardiometabolic risk among rural adolescents: A cross-sectional study at 15-year follow-up of the MINIMat cohort

Background

Diet being a modifiable factor, its relationship with cardiometabolic risk is of public health interest. The vast majority of studies on associations of dietary patterns with cardiometabolic risk indicators among adolescents are from high-income countries and urban settings. We sought to describe dietary patterns and examine their associations with selected cardiometabolic risk indicators-waist circumference (WC), systolic blood pressure, fasting lipid profile and insulin resistance-along with its gender stratification among adolescents in a low-income, rural setting.

Methods

This cross-sectional study utilized data from the 15-year follow-up of the Maternal and Infant Nutrition Interventions in Matlab (MINIMat) cohort in southeast Bangladesh. The children who were born as singletons to the mothers randomized in the MINIMat trial and had valid birth anthropometrics were eligible for the follow-up. We employed a single, qualitative 24-hour recall to assess diet. Dietary patterns were derived from simple K-means cluster analysis, and calculation of dietary diversity score (DDS) using a validated instrument. Anthropometric parameters and systolic blood pressure were recorded. Fasting plasma triglyceride, total cholesterol, low- and high-density lipoproteins, insulin and glucose levels were measured. We calculated insulin resistance using the Homeostasis Model Assessment equation (HOMA-IR). Three right-skewed outcome variables were natural log (Ln) transformed: WC, triglyceride and HOMA-IR. Omnibus and gender-specific multiple linear regression models were fitted.

Results

Among 2,253 adolescents (52.1% girls, 7.1% overweight/obese), we identified four diet clusters: Traditional, Fish-dominant, Meat-dominant, and High-variety. No significant associations were found between the clusters and indicators. On gender-stratification, triglyceride levels were lower among boys in the Fish-dominant (Ln-triglyceride β_adjusted: -0.09; 95% confidence interval (CI): -0.15, -0.02) and Meat-dominant (Ln-triglyceride β_adjusted: -0.08; 95% CI: -0.15, -0.004) clusters than among boys in the Traditional cluster. Compared to boys in the bottom quartile of DDS, boys in the top quartile had 2.1 mm of Hg (95% CI: 0.5, 3.6) higher systolic blood pressure and 1.9% (95% CI: 0.01-3.8%) higher WC.

Conclusion

While statistically significant, the gender-specific differences in triglyceride, systolic blood pressure, and waist circumference across dietary patterns were small. Associations between dietary patterns and cardiometabolic risk indicators may require a time lag beyond mid-adolescence to manifest in a rural setting. Prospective studies are warranted to delineate the magnitude and direction of those associations.

Dietary Patterns in Adolescent Obesity as Predictors of Long-Term Success Following an Intensive Inpatient Lifestyle Programme

(1) Background: Lifestyle interventions for adolescents with obesity show minor long-term effects on anthropometric parameters. The persistence of dietary changes after obesity inpatient rehabilitation has not been sufficiently investigated. (2) Objectives: To analyse dietary patterns in German adolescents with obesity as predictors of long-term success following an intensive inpatient lifestyle programme regarding food choices as well as body weight and comorbidities. (3) Methods: Food consumption data of 137 German adolescents with obesity aged 10-17 years were collected by a nutrition interview. Cluster analysis was used to group the participants according to their food consumption. Dietary patterns, changes in body weight and insulin resistance were compared over a 2-year-period. (4) Results: Three dietary patterns were identified. Big Eaters (n = 32) consume high amounts of total sugar and meat, Moderate Eaters (n = 66) have a diet comparable to the national average, and Snackers (n = 39) have a particularly high consumption of total sugar. Big Eaters and Snackers significantly reduced the consumption of total sugar. Among Moderate Eaters, no persistent changes were observed. (5) Conclusion: Weight reduction interventions can induce long-lasting changes in the diet of adolescents with obesity. Therefore, the success of a weight reduction intervention should not be determined by weight reduction only.

Dietary patterns and birth outcomes of healthy Lebanese pregnant women

Background

The aim of this study was to define the dietary patterns (DPs) of a sample of Lebanese pregnant women and to establish their correlation with maternal and neonatal outcomes.

Methods

A cross-sectional study was conducted among 358 Lebanese pregnant women. Maternal socio-demographic variables, anthropometric measurements, gestational weight gain and neonatal outcomes such as weight, length, head circumference and Apgar score were collected by qualified dietitians. Dietary intake was assessed by a validated food frequency questionnaire and three 24-h dietary recalls. DPs were determined, a posteriori, by a factor analysis to distinguish the inter-correlations between the food groups and a cluster analysis method to assemble the participants into groupings based on similarities in food consumption.

Results

The identified DPs were not exclusively composed of specific food groups, since some components were overlapping in the DPs. The first one was characterized by a high consumption of starchy vegetables, unsaturated fats and unhealthy foods, the second was rich in fruits and vegetables, seeds, rice and pasta, and in fried local meals and the third was mainly based on protein-rich foods like poultry, fish, eggs and dairy products. They were named respectively as “Westernized,” “Mixed” and “Neo-Mediterranean” by the research team. Women having the lowest pre-gestational body mass index (BMI) and higher gestational age followed mainly the “Neo-Mediterranean” eating pattern.

Conclusion

The three identified DPs among Lebanese pregnant women were correlated with the pre-gestational BMI, and some maternal variables. However, neither the maternal nor the neonatal outcomes were correlated with the DP adopted by the mothers.

Consumers attitudes toward animal welfare friendly produce: An island of Ireland study

Public concern has arisen between increased productivity in animal production systems and farm animal welfare (FAW). Consumers demand ethical production systems and tend to purchase products that reflect their animal welfare concerns. They are becoming increasingly judicious, with growing expectations around animal produce. Nonetheless, positive attitudes concerning FAW do not always correspond to purchasing behavior, where higher concern is not related to higher purchasing frequency. The study seeks to determine if individuals behave differently in their dual roles as citizens and consumers, expressing preferences for FAW but ultimately failing to act on such preferences when making purchasing decisions. Theoretically, the research uses elements of the COM-B model to investigate the importance of capability, opportunity, and motivation for consumers. The study is conducted on an island of Ireland basis and investigates the perceptions and purchasing frequency of Irish and Northern Irish consumer’s (n = 972). Firstly, attitudes towards purchasing animal welfare products are elicited using Categorical Principal Component Analysis (CATPCA). Cluster analysis subsequently defines three distinct typologies based upon the final cluster centers attained from CATPCA. The largest cluster is characterized as being Indifferent (n = 672) and differentiated by relatively low concern for animal welfare. Despite this, their purchasing frequency of animal welfare products is monthly. The other two clusters, The Engaged and The Struggling, are similar in size and share the same concern for animal welfare but are polarized in their motivation, capability, and opportunity to purchase animal welfare friendly products. The Struggling typology act as citizens rather than consumers; while they are highly concerned for welfare they have significantly lower purchasing frequency even compared to The Indifferent. Finally, Non-Linear Canonical Correlation Analysis (NLCCA) is used to reveal and visually demonstrate the structure of the whole dataset. The analysis demonstrates that concern for animal welfare is not associated with purchase frequency of animal welfare friendly produce when contextual factors are not conducive to doing so. The findings are relevant to a broad audience including industry, policymakers, and academics.

The association between the dietary pattern in abdominal obesity based on visceral fat index and dyslipidaemia in the Henan Rural Cohort Study

The present study aimed to explore the association between dietary patterns in abdominal obesity obtained by reduced-rank regression (RRR) with visceral fat index (VFI) as a dependent variable and dyslipidemia in rural adults in Henan, China. A total of 29538 people aged 18-79 were selected from the Henan Rural Cohort Study. RRR analysis was used to identify dietary patterns; logistic regression analysis and restricted cubic spline regression models were applied to analyze the association between dietary patterns in abdominal obesity and dyslipidemia. VFI was used as a mediator to estimate the mediation effect. The dietary pattern in abdominal obesity was characterized by high carbohydrate and red meat intake and low consumption of fresh fruits, vegetables, milk, etc. After full adjustment, the highest quartile of dietary pattern scores was significantly associated with an increased risk of dyslipidemia (OR: 1·33, 95 % CI 1·23-1·44, P_trend < 0·001), there was a non-linear dose-response relationship between them (P_{overall-association} < 0·001, P_{non-lin-association} = 0·022). The result was similar in dose-response between the dietary pattern scores and VFI. The indirect effect partially mediated by VFI was significant (OR: 1·07, 95 % CI 1·06-1·08). VIF explained approximately 53·3 % of odds of dyslipidemia related to the dietary pattern. Abdominal obesity dietary pattern scores positively affected VFI and dyslipidemia; there was a dose-response in both relationships. Dyslipidemia progression increased with higher abdominal obesity dietary pattern scores. In addition, VFI played a partial mediating role in relationship between abdominal obesity dietary pattern and dyslipidemia.

The impacts of measurement errors on a dietary pattern analyses:a simulation study based on dietary data from the China Multi-Ethnic Cohort (CMEC) study

BACKGROUND

Measurement error is a significant challenge in nutritional epidemiology research. Compared with traditional, isolated-nutrient research, dietary-pattern studies provide a more comprehensive approach to chronic disease prevention and have become popular in recent years. However, few studies have examined the impacts of measurement errors on dietary pattern analyses.

OBJECTIVES

We investigated the impacts of measurement errors on the 2 most commonly used dietary pattern derivation methods: principal component factor analysis (PCFA) and K-means cluster analysis (KCA).

METHODS

We conducted a simulation study by taking the dietary data collected in the China Multi-Ethnic Cohort study as the "true values" and adding linear measurement errors for each food group to consider both systematic and random errors. We investigated the impacts of measurement errors from 2 aspects: distortion of the derived dietary patterns and estimated associations between the dietary patterns and disease.

RESULTS

For both systematic and random errors, larger measurement errors caused more serious distortion of dietary patterns derived by PCFA and KCA, with consistency rates ranging from 67.5% to 100% and from 13.4% to 88.4%, respectively. In addition, for both systematic and random errors, larger measurement errors caused greater attenuation effects on the association coefficients. For a beneficial association (coefficient, -0.5), the estimated coefficients ranged from -0.287 to -0.450 and from -0.231 to -0.394 in the PCFA and KCA, respectively. For a harmful association (coefficient, 0.5), the estimated coefficients ranged from 0.295 to 0.449 and from -0.003 to 0.373 in the PCFA and KCA, respectively. Dietary patterns derived by PCFA with factor loadings of low discrepancies and dietary patterns derived by KCA with small cluster sample sizes are more vulnerable to measurement error.

CONCLUSIONS

Measurement errors could distort dietary patterns and attenuate the dietary pattern-disease association. The stability of dietary patterns under measurement errors differs when using PCFA and KCA.

Clustering Analysis and Machine Learning algorithms in the prediction of dietary patterns: Cross-sectional results of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil)

BACKGROUND

Machine learning investigates how computers can automatically learn. This study aimed to predict dietary patterns and compare algorithm performance in making predictions of dietary patterns.

METHODS

We analyzed the data of public employees (n=12,667) participating in the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil). The K-means clustering algorithm and six other classifiers (support vector machines, naïve Bayes, K-nearest neighbors, decision tree, random forest, and xgboost) were used to predict the dietary patterns.

RESULTS

K-means clustering identified two dietary patterns. Cluster 1, labeled the Western pattern, was characterized by a higher energy intake and consumption of refined cereals, beans and other legumes, tubers, pasta, processed and red meats, high-fat milk and dairy products, and sugary beverages; Cluster 2, labeled the Prudent pattern, was characterized by higher intakes of fruit, vegetables, whole cereals, white meats, and milk and reduced-fat milk derivatives. The most important predictors were age, sex, per capita income, education level, and physical activity. The accuracy of the models varied from moderate to good (69-72%).

CONCLUSIONS

The algorithms' performance in dietary pattern prediction was similar, and the models presented may provide support in screener tasks and guide health professionals in the analysis of dietary data. This article is protected by copyright. All rights reserved.

Dietary patterns and associated factors of children under two years of age born prematurely

Objective:

To identify the dietary patterns and associated factors of children aged between 6 and 23 months, born prematurely and assisted at a University Hospital in Curitiba, state of Paraná, Brazil.

Methods:

The parents or guardians of the 135 children were asked about their children’s eating habits and the family’s socioeconomic and demographic conditions. Information regarding birth and health history were obtained from medical records. Data on food consumption were subjected to exploratory factor analysis and the principal component analysis method was used to estimate the factor loads. Multiple linear regression was performed to verify possible associations.

Results:

Two dietary patterns were observed: “unhealthy” and “healthy.” The “unhealthy” pattern was significantly associated with maternal age, the child’s corrected age, and gestational age at birth. The “healthy pattern” was associated with the child’s corrected age. Maternal age and child’s corrected age remained significant after multiple regression analyses. For the “unhealthy” pattern, a positive effect was observed, suggesting that the consumption of this pattern is higher as the child’s age increases and less intense for children with mothers aged 30 years or older. For the “healthy” dietary pattern, the same two variables showed statistical significance. The authors observed a direct proportion between the age and consumption of food groups in both patterns.

Conclusions:

These results indicate the importance of nutritional education for younger mothers regarding their children’s eating practices, especially as the child grows.

Dietary networks identified by Gaussian graphical model and general and abdominal obesity in adults

BACKGROUND

Gaussian graphical model (GGM) has been introduced as a new approach to identify patterns of dietary intake. We aimed to investigate the link between dietary networks derived through GGM and obesity in Iranian adults.

METHOD

A cross-sectional study was conducted on 850 men and women (age range: 20-59 years) who attended the local health centers in Tehran. Dietary intake was evaluated by using a validated food frequency questionnaire. GGM was applied to identify dietary networks. The odds ratios (ORs) and 95% confidence intervals (CIs) of general and abdominal adiposity across tertiles of dietary network scores were estimated using logistic regression analysis controlling for age, sex, physical activity, smoking status, marital status, education, energy intake and menopausal status.

RESULTS

GGM identified three dietary networks, where 30 foods were grouped into six communities. The identified networks were healthy, unhealthy and saturated fats networks, wherein cooked vegetables, processed meat and butter were, respectively, central to the networks. Being in the top tertile of saturated fats network score was associated with a higher likelihood of central obesity by waist-to-hip ratio (OR: 1.56, 95%CI: 1.08, 2.25; P for trend: 0.01). There was also a marginally significant positive association between higher unhealthy network score and odds of central obesity by waist circumference (OR: 1.37, 95%CI: 0.94, 2.37; P for trend: 0.09). Healthy network was not associated with central adiposity. There was no association between dietary network scores and general obesity.

CONCLUSIONS

Unhealthy and saturated fat dietary networks were associated with abdominal adiposity in adults. GGM-derived dietary networks represent dietary patterns and can be used to investigate diet-disease associations.

A posteriori dietary patterns better explain variations of the gut microbiome than individual markers in the American Gut Project

BACKGROUND

Individual diet components and specific dietary regimens have been shown to impact the gut microbiome.

OBJECTIVES

Here, we explored the contribution of long-term diet by searching for dietary patterns that would best associate with the gut microbiome in a population-based cohort.

METHODS

Using a priori and a posteriori approaches, we constructed dietary patterns from an FFQ completed by 1800 adults in the American Gut Project. Dietary patterns were defined as groups of participants or combinations of food variables (factors) driven by criteria ranging from individual nutrients to overall diet. We associated these patterns with 16S ribosomal RNA-based gut microbiome data for a subset of 744 participants.

RESULTS

Compared to individual features (e.g., fiber and protein), or to factors representing a reduced number of dietary features, 5 a posteriori dietary patterns based on food groups were best associated with gut microbiome beta diversity (P ≤ 0.0002). Two patterns followed Prudent-like diets-Plant-Based and Flexitarian-and exhibited the highest Healthy Eating Index 2010 (HEI-2010) scores. Two other patterns presented Western-like diets with a gradient in HEI-2010 scores. A fifth pattern consisted mostly of participants following an Exclusion diet (e.g., low carbohydrate). Notably, gut microbiome alpha diversity was significantly lower in the most Western pattern compared to the Flexitarian pattern (P ≤ 0.009), and the Exclusion diet pattern was associated with low relative abundance of Bifidobacterium (P ≤ 1.2 × 10-7), which was better explained by diet than health status.

CONCLUSIONS

We demonstrated that global-diet a posteriori patterns were more associated with gut microbiome variations than individual dietary features among adults in the United States. These results confirm that evaluating diet as a whole is important when studying the gut microbiome. It will also facilitate the design of more personalized dietary strategies in general populations.

Knowledge, attitudes and behaviors on antibiotic use and resistance among healthcare workers in Italy, 2019: investigation by a clustering method

BACKGROUND

Identifying healthcare workers (HCW) who have less awareness and knowledge on antibiotic use and resistance represents a challenge for public health, since it might help the development of novel educational and training initiatives tailored on specific subgroups of professionals. This work aims to compare knowledge, attitudes and behaviors on antibiotic use and resistance across different groups of Italian HCW.

METHODS

We used data from the multi-country and multi-professional survey launched by the European Centre for Disease Prevention and Control between 28 January to 4 March 2019 to assess knowledge, attitude and behaviors of HCW on antibiotics, antibiotic use and resistance. We distinguished three clusters of HCW using the Two-Step Cluster analysis, based on their personal and professional characteristics (i.e. profession, role, activity as prescriber, setting, and activity as antibiotic use advisor).

RESULTS

In general, cluster 1 consisted mostly of allied healthcare workers, while clusters 2 and 3 were made up almost completely of pharmacists and medical doctors, respectively. Interestingly, healthcare workers in cluster 3 had the highest knowledge on antibiotic use and resistance. Workers in cluster 1, instead, were those reporting the highest awareness of the importance and role of hand hygiene as an infection prevention and control measure. However, HCW in cluster 2 were those who recognized more their role of advisors on prudent antibiotic use.

CONCLUSIONS

Italian HCW exhibited different knowledge, attitudes, and behaviors on antibiotic use and resistance. These findings raised the need for educational and training interventions targeting specific professional groups.

Data-Driven Clustering Approach to Derive Taste Perception Profiles from Sweet, Salt, Sour, Bitter, and Umami Perception Scores: An Illustration among Older Adults with Metabolic Syndrome

BACKGROUND

Current approaches to studying relations between taste perception and diet quality typically consider each taste-sweet, salt, sour, bitter, umami-separately or aggregately, as total taste scores. Consistent with studying dietary patterns rather than single foods or total energy, an additional approach may be to study all 5 tastes collectively as "taste perception profiles."

OBJECTIVE

We developed a data-driven clustering approach to derive taste perception profiles from taste perception scores and examined whether profiles outperformed total taste scores for capturing individual variability in taste perception.

METHODS

The cohort included 367 community-dwelling adults [55-75 y; 55% female; BMI (kg/m2): 32.2 ± 3.6] with metabolic syndrome from PREDIMED-Plus, Valencia. Cluster analysis identified subgroups of individuals with similar patterns in taste perception (taste perception profiles); quantitative criteria were used to select the cluster algorithm, determine the optimal number of clusters, and assess the profiles' validity and stability. Goodness-of-fit parameters from adjusted linear regression evaluated the individual variability captured by each approach.

RESULTS

A k-means algorithm with 6 clusters best fit the data and identified the following taste perception profiles: Low All, High Bitter, High Umami, Low Bitter & Umami, High All But Bitter and High All But Umami. All profiles were valid and stable. Compared with total taste scores, taste perception profiles explained more variability in bitter and umami perception (adjusted R2: 0.19 vs. 0.63, respectively; 0.40 vs. 0.65, respectively) and were comparable for sweet, salt, and sour. In addition, taste perception profiles captured differential perceptions of each taste within individuals, whereas these patterns were lost with total taste scores.

CONCLUSIONS

Among older adults with metabolic syndrome, taste perception profiles derived via data-driven clustering may provide a valuable approach to capture individual variability in perception of all 5 tastes and their collective influence on diet quality. This trial was registered at https://www.isrctn.com/ as ISRCTN89898870.

Association of Diet, Physical Activity Guidelines and Cardiometabolic Risk Markers in Children

The aim was to identify different dietary and physical activity (PA) patterns in 5- to 14-year-old children with a high prevalence of overweight and obesity using cluster analysis based on their adherence to the Spanish Society of Community Nutrition dietary guidelines and levels of PA, and to determine their associations with age, sex, body composition, and cardiometabolic risk markers. In 549 children, hierarchical cluster analysis was used to identify subgroups with similar adherence to dietary recommendations and level of PA. Three clusters were identified: Cluster 1, with the lowest level of vigorous PA and adherence to dietary recommendations; Cluster 2, with the lowest levels of moderate and vigorous PA and the highest adherence to dietary recommendations; and Cluster 3, with the highest level of PA, especially vigorous PA and a medium level adherence to dietary recommendations. Cluster 3 had lower total body fat and higher lean body mass percentages than Cluster 2. Cluster 2 had lower high-density lipoprotein cholesterol and higher low-density lipoprotein cholesterol levels than Cluster 1. The results from our study suggest that it is important to consider adherence to PA recommendations together with adherence to dietary guidelines to understand patterns of obesogenic habits in pediatric populations with high prevalence of overweight and obesity.

Associated factors with dietary patterns among children under 2 years of age: a study in childcare centres and homes of South Brazil

Identifying dietary patterns in different environments attended by children is relevant to guide public politics. The aim of this study was to analyse the factors associated with dietary patterns of children under 2 years of age in childcare centres and at homes. This transversal study was enrolled in municipal childcare centres of Guaratuba, Paraná, Brazil. Food consumption data from 256 children were obtained by the food record method. From the consumption data, four dietary patterns were identified by factor analysis using the principal component method in each environment. ‘Traditional’ patterns were observed in both environments; the ‘less healthy’ pattern was found only at homes. Other patterns identified in childcare centres include ‘snacks’, ‘nutritive’ and ‘pasta and meats’; at homes, it was possible to identify patterns such as ‘milk and cereals’ and ‘mixed’. Children over 12 months presented higher scores for all the patterns in both environments. Obese children had lower scores for all the patterns in childcare centres. There was an association between maternal age below 21 years and higher adherence to a ‘less healthy’ pattern, maternal level of education less than 8 years and lower adherence to the ‘mixed’ pattern and lower familiar income per capita and higher adherence to the ‘snacks’ and ‘traditional’ patterns in childcare centres. In conclusion, adherence to dietary patterns was associated with socio-economic, demographic variables and nutritional status and further studies are needed, especially those with a longitudinal design, enabling the monitoring of dietary patterns.

The effect of different methods to identify, and scenarios used to address energy intake misestimation on dietary patterns derived by cluster analysis

Background

All self-reported dietary intake data are characterized by measurement error, and validation studies indicate that the estimation of energy intake (EI) is particularly affected.

Methods

Using self-reported food frequency and physical activity data from Alberta’s Tomorrow Project participants (n = 9847 men 16,241 women), we compared the revised-Goldberg and the predicted total energy expenditure methods in their ability to identify misreporters of EI. We also compared dietary patterns derived by k-means clustering under different scenarios where misreporters are included in the cluster analysis (Inclusion); excluded prior to completing the cluster analysis (ExBefore); excluded after completing the cluster analysis (ExAfter); and finally, excluded before the cluster analysis but added to the ExBefore cluster solution using the nearest neighbor method (InclusionNN).

Results

The predicted total energy expenditure method identified a significantly higher proportion of participants as EI misreporters compared to the revised-Goldberg method (50% vs. 47%, p < 0.0001). k-means cluster analysis identified 3 dietary patterns: Healthy, Meats/Pizza and Sweets/Dairy. Among both men and women, participants assigned to dietary patterns changed substantially between ExBefore and ExAfter and also between the Inclusion and InclusionNN scenarios (Hubert and Arabie’s adjusted Rand Index, Kappa and Cramer’s V statistics < 0.8).

Conclusions

Different scenarios used to account for EI misreporters influenced cluster analysis and hence the composition of the dietary patterns. Continued efforts are needed to explore and validate methods and their ability to identify and mitigate the impact of EI misestimation in nutritional epidemiology.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12937-021-00696-3.

A review of statistical methods for dietary pattern analysis

Background

Dietary pattern analysis is a promising approach to understanding the complex relationship between diet and health. While many statistical methods exist, the literature predominantly focuses on classical methods such as dietary quality scores, principal component analysis, factor analysis, clustering analysis, and reduced rank regression. There are some emerging methods that have rarely or never been reviewed or discussed adequately.

Methods

This paper presents a landscape review of the existing statistical methods used to derive dietary patterns, especially the finite mixture model, treelet transform, data mining, least absolute shrinkage and selection operator and compositional data analysis, in terms of their underlying concepts, advantages and disadvantages, and available software and packages for implementation.

Results

While all statistical methods for dietary pattern analysis have unique features and serve distinct purposes, emerging methods warrant more attention. However, future research is needed to evaluate these emerging methods’ performance in terms of reproducibility, validity, and ability to predict different outcomes.

Conclusion

Selection of the most appropriate method mainly depends on the research questions. As an evolving subject, there is always scope for deriving dietary patterns through new analytic methodologies.

Data-Driven Dietary Patterns, Nutrient Intake and Body Weight Status in a Cross-Section of Singaporean Children Aged 6-12 Years

Pattern analysis of children’s diet may provide insights into chronic disease risk in adolescence and adulthood. This study aimed to assess dietary patterns of young Singaporean children using cluster analysis. An existing dataset included 15,820 items consumed by 561 participants (aged 6–12 years) over 2 days of dietary recall. Thirty-seven food groups were defined and expressed as a percentage contribution of total energy. Dietary patterns were identified using k-means cluster analysis. Three clusters were identified, “Western”, “Convenience” and “Local/hawker”, none of which were defined by more prudent dietary choices. The “Convenience” cluster group had the lowest total energy intake (mean 85.8 ± SD 25.3% of Average Requirement for Energy) compared to the other groups (95.4 ± 25.9% for “Western” and 93.4 ± 25.3% for “Local/hawker”, p < 0.001) but also had the lowest calcium intake (66.3 ± 34.7% of Recommended Dietary Allowance), similar to intake in the “Local/hawker” group (69.5 ± 38.9%) but less than the “Western” group (82.8 ± 36.1%, p < 0.001). These findings highlight the need for longitudinal analysis of dietary habit in younger Singaporeans in order to better define public health messaging targeted at reducing risk of major noncommunicable disease.

Metabolomic Biomarkers of Healthy Dietary Patterns and Cardiovascular Outcomes

PURPOSE OF REVIEW

Healthy dietary patterns are recommended for prevention of CVD. Recently, metabolomics has been used to identify biomarkers of healthy dietary patterns and elucidate mechanisms underlying diet-disease associations. This review provides an overview of approaches to define healthy dietary patterns, discusses important issues related to using metabolomics to describe healthy dietary patterns, and summarizes studies identifying blood metabolites associated with hypothesis-driven healthy dietary patterns and cardiovascular risk factors and incident CVD.

RECENT FINDINGS

We identified 17 studies which reported on blood metabolomic signatures of 5 healthy dietary patterns (Healthy Eating Index, Alternative Healthy Eating Index, the Dietary Approaches to Stop Hypertension diet, Mediterranean diet, vegetarian diet). Four of these studies evaluated associations between diet-related metabolites and cardiovascular outcomes. Many metabolites replicated across different healthy dietary patterns, which suggest that they may represent biomarkers of generally healthy diets. Unsaturated lipids positively associated with healthy dietary patterns were inversely associated with incident CVD, suggesting that they may be a pathway through which diet is associated with a lower risk of CVD. Although many metabolites replicated across cross-sectional studies, few metabolites identified as candidate biomarkers of healthy diets in feeding studies replicated in observational studies. Additionally, limited evidence exists on the ability of diet-related metabolites to predict cardiovascular outcomes. Replication of candidate biomarkers of dietary patterns in different study designs and more studies evaluating the associations between diet-related metabolites and cardiovascular outcomes are needed.

Analysis of dietary patterns and nutritional adequacy in lactating women: a multicentre European cohort (ATLAS study)

Eating habits of lactating women can influence the nutrient composition of human milk, which in turn influences nutrient intake of breastfed infants. The aim of the present study was to identify food patterns and nutritional adequacy among lactating women in Europe. Data from a multicentre European longitudinal cohort (ATLAS study) were analysed to identify dietary patterns using cluster analysis. Dietary information from 180 lactating women was obtained using 3-d food diaries over the first 4 months of lactation. Four dietary patterns were identified: 'vege-oils', 'fish-poultry', 'confectionery-salads' and 'mixed dishes'. Nutrition adequacy was not significantly different between clusters, but the 'vege-oils' cluster tended to yield the highest nutrition adequacy measured by Mean Adequacy Ratio. Compared with European dietary reference values (DRVs) for lactating women, women in all clusters had inadequate intakes of energy, pantothenic acid, folate, vitamin C, vitamin A, vitamin D, zinc, iodine, potassium and linoleic acid. Adequate intake for fibre and α-linolenic acid was only achieved in the 'vege-oils' cluster. Overall, fat intake was above DRVs. The present study showed that various dietary patterns do not adequately supply all nutrients, indicating a need to promote overall healthy dietary habits for European lactating women.

Dietary Patterns are Differentially Associated with Atypical and Melancholic Subtypes of Depression

Diet has been associated with the risk of depression, whereas different subtypes of depression have been linked with different cardiovascular risk factors (CVRFs). In this study, our aims were to (1) identify dietary patterns with exploratory factor analysis, (2) assess cross-sectional associations between dietary patterns and depression subtypes, and (3) examine the potentially mediating effect of dietary patterns in the associations between CVRFs and depression subtypes. In the first follow-up of the population-based CoLaus|PsyCoLaus study (2009–2013, 3554 participants, 45.6% men, mean age 57.5 years), a food frequency questionnaire assessed dietary intake and a semi-structured interview allowed to characterize major depressive disorder into current or remitted atypical, melancholic, and unspecified subtypes. Three dietary patterns were identified: Western, Mediterranean, and Sweet-Dairy. Western diet was positively associated with current atypical depression, but negatively associated with current and remitted melancholic depression. Sweet-Dairy was positively associated with current melancholic depression. However, these dietary patterns did not mediate the associations between CVRFs and depression subtypes. Hence, although we could show that people with different subtypes of depression make different choices regarding their diet, it is unlikely that these differential dietary choices account for the well-established associations between depression subtypes and CVRFs.

Dietary Patterns and Their Association with Body Composition and Cardiometabolic Markers in Children and Adolescents: Genobox Cohort

Diet is a key factor for obesity development; however, limited data are available on dietary cluster analysis in children with obesity. We aimed to assess the associations between dietary patterns and obesity and several cardiometabolic markers. Anthropometry, bioelectrical impedance, blood pressure and plasma biomarkers of oxidative stress, inflammation and endothelial damage were determined in 674 Caucasian children, aged 5-16, with normal or excess weight. Using a food frequency questionnaire and cluster analysis, two consistent dietary patterns were shown, labeled as health conscious (HC) and sweet and processed (SP). The HC pattern included a greater proportion of participants with overweight/obesity than the SP cluster (80.1% vs. 63.8%). However, children with obesity within the HC cluster, showed less abdominal fat, through waist to hip (0.93 vs. 0.94) and waist to height (0.61 vs. 0.63) indexes (p < 0.01). Univariate general models showed several additional differences in cardiometabolic risk biomarkers in the global and stratified analyses, with a healthier profile being observed mainly in the HC cluster. However, multivariate models questioned these findings and pointed out the need for further studies in this field. Anyhow, our findings support the benefits of a healthy diet and highlight the importance of dietary patterns in the cardiometabolic risk assessment of children with overweight/obesity, beyond weight control.

Cluster analysis identifies patients at risk of catheter-associated urinary tract infections in intensive care units: findings from the SPIN-UTI Network

BACKGROUND

Although preventive strategies have been proposed against catheter-associated urinary tract infections (CAUTIs) in intensive care units (ICUs), more efforts are needed to control the incidence rate.

AIM

To distinguish patients according to their characteristics at ICU admission, and to identify clusters of patients at higher risk for CAUTIs.

METHODS

A two-step cluster analysis was conducted on 9656 patients from the Italian Nosocomial Infections Surveillance in Intensive Care Units project.

FINDINGS

Three clusters of patients were identified. Type of admission, patient origin and administration of antibiotics had the greatest weight on the clustering model. Cluster 1 comprised more patients with a medical type of ICU admission who came from the community. Cluster 2 comprised patients who were more likely to come from other wards/hospitals, and to report administration of antibiotics 48 h before or after ICU admission. Cluster 3 was similar to Cluster 2 but was characterized by a lower percentage of patients with administration of antibiotics 48 h before or after ICU admission. Patients in Clusters 1 and 2 had a longer duration of urinary catheterization [median 7 days, interquartile range (IQR) 12 days for Cluster 1; median 7 days, IQR 11 days for Cluster 2] than patients in Cluster 3 (median 6 days, IQR 8 days; P<0.001). Interestingly, patients in Cluster 1 had a higher incidence of CAUTIs (3.5 per 100 patients) compared with patients in the other two clusters (2.5 per 100 patients in both clusters; P=0.033).

CONCLUSION

To the authors' knowledge, this is the first study to use cluster analysis to identify patients at higher risk of CAUTIs who could gain greater benefit from preventive strategies.

Dietary Inflammatory Index is associated with Healthy Eating Index, Alternative Healthy Eating Index, and dietary patterns among Iranian adults

Background

Recent investigations have evaluated the effect of the inflammatory potential of diet in several populations by calculating the Dietary Inflammatory Index (DII) score. We aimed to evaluate the association of the DII with the Healthy Eating Index (HEI), the Alternative Healthy Eating Index (AHEI), and dietary pattern (DP) among healthy Iranian adults.

Methods

A cross‐sectional study was conducted among 4365 middle‐aged adults. Major DPs and DII score were identified using a validated semi‐quantitative food frequency questionnaire (FFQ). Poisson regression was used to evaluate the association of DPs, HEI, and AHEI across tertiles of DII.

Results

After adjustment for confounding variables, a low HEI (HEI < 55) and AHEI (AHEI < 56.5) were more prevalent among the participants in the highest tertile of DII compared to the first tertile (PR: 1.13, P‐value <.05; PR: 1.10, P‐value <.05; respectively). Adherence to a balanced healthy dietary pattern was significantly lower in subjects with a diet that was more pro‐inflammatory compared to those with anti‐inflammatory diet (PR: 0.85, P‐value P < .01). No significant association was found between the DII and a western DP. High levels of HDL and hip and waist circumference were observed in the highest tertile of DII, and high levels of dietary intake of protein and fiber, minerals, fasting blood glucose, and monounsaturated fat were reported in the lowest tertile of DII.

Conclusion

The highest tertile of the DII (a pro‐inflammatory diet) was associated with a lower HEI, AHEI, and lower adherence to balanced DP in a representative sample of adults in Iran.

Dietary Patterns and Health Outcomes among African American Maintenance Hemodialysis Patients

The association between dietary patterns and health outcomes, such as quality of life (QOL), in maintenance hemodialysis (MHD) patients with certain racial backgrounds has not been studied in detail. QOL is a powerful outcome measure in which dietary patterns could be a modifying factor. This study is a secondary analysis examining the association between dietary patterns and health outcomes in 101 African American (AA) maintenance hemodialysis (MHD) patients participating in the Palm Tocotrienols in Chronic Hemodialysis (PATCH) study. Quality of life (QOL) was assessed using the Kidney Disease Quality of Life 36-item survey (KDQOL-36™). Blood samples were analyzed for lipids, lipoprotein subfractions, and inflammatory markers. Food intake was measured using six non-consecutive 24-h dietary recalls over 15 months. Implausible energy intake reports were screened out by comparing reported energy intake (rEI) with predicted total energy expenditure (pTEE). Cluster analysis, using the k-means algorithm, identified two distinct dietary patterns in the study population: a high “sugar sweetened beverage” pattern (hiSSB) and a low “sugar sweetened beverage pattern” (loSSB). In the hiSSB group, consumption of SSB accounted for ~28% of energy intake, while SSB represented only 9% of energy intake in the loSSB group. The hiSSB group was characterized by a higher intake of total calories, sugar and percentage of kilocalories from carbohydrates, whereas the percentage of kilocalories from protein and fat was lower. While additional micronutrient intakes differed between groups (vitamin C, zinc, chromium), these were significantly lower than recommended values in the entire cohort. Patients in the hiSSB group presented with lower high-density lipoprotein cholesterol (HDL-C), lower large HDL particles and smaller low density lipoprotein (LDL) particle diameters. Antidepressant usage was significantly higher in the hiSSB group. Patients in the hiSSB group scored lower across all five KDQOL domains and scored significantly lower in the mental composite domain. MHD patients following a hiSSB dietary pattern had smaller dense LDL particles, lower HDL-C, and a lower QOL. Suboptimal intakes of fruits, vegetables, and grains as well as key micronutrients were evident in both patterns.

Education and lifestyle predict change in dietary patterns and diet quality of adults 55 years and over

Background

Diet is a key risk factor for chronic disease, and an increasing concern among older adults. We aim to examine the changes in dietary patterns using principal component analysis and a diet quality index among older adults and examine the predictors of dietary change over a 4 year period.

Methods

Data was obtained via a postal survey in a prospective cohort, the Wellbeing Eating and Exercise for a Long Life (WELL) study. Australian adults aged 55 years and over (n = 1005 men and n = 1106 women) completed a food frequency at three time points and provided self-reported personal characteristics. Principal component analysis was used to assess dietary patterns and diet quality was assessed using the 2013 Revised Dietary Guideline Index. The relationships between predictors and change in dietary patterns were assessed by multiple linear regression.

Results

Two dietary patterns were consistently identified in men and women at three time points over 4 years. One was characterised by vegetables, fruit and white meat, and the other was characterised by red and processed meat and processed foods. Reduced consumption of key food groups within the principal component analysis-determined dietary patterns was observed. An increase in diet quality over 4 years was observed in men only. Reported higher education levels and favourable lifestyle characteristics, including not smoking and physical activity, at baseline predicted an increase in healthier dietary patterns over 4 years.

Conclusions

There was stability in the main dietary patterns identified over time, however participants reported an overall decrease in the frequency of consumption of key food groups. Compliance with the Australian Dietary Guidelines remained poor and therefore targeting this population in nutritional initiatives is important. Design of nutrition promotion for older adults need to consider those with lower socioeconomic status, as having a lower level of education was a predictor of poorer dietary patterns. It is important to consider how nutrition behaviours can be targeted alongside other lifestyle behaviours, such as smoking and inadequate physical activity to improve health.

Multidimensional Analysis of Food Consumption Reveals a Unique Dietary Profile Associated with Overweight and Obesity in Adolescents

There is a significant increase in overweight and obesity in adolescents worldwide. Here, we performed a cross-sectional study to examine the potential association between food consumption profiles and overweight in a large number of adolescents from Brazil. Sampling by clusters and conglomerates was carried out in students of public schools in Salvador, Brazil, between June and December 2009 and 1496 adolescents were evaluated. Data on socio-epidemiological data, anthropometric status and food consumption were captured. Multivariate analyses, such as hierarchical clustering and correlation networks, were used to perform a detailed description of food consumption profiles. There were differences in age and anthropometric status related to sex. Four clusters of food groups were identified based on the intake profile in the study population. No disparities in food intake were observed in individuals stratified by sex or anthropometric status. Furthermore, network analysis revealed that overweight or obesity were hallmarked by a selectivity in the ingestion of food groups that resulted in the appearance of inverse correlations of consumption, which was not present in eutrophic adolescents. Thus, overweight and obesity are associated with preferential choices of ingestion of specific food groups, which result in the appearance of inverse correlations of consumption. Such knowledge may serve as basis for future targeted nutritional interventions in adolescents.

Mediterranean food pattern vs. Mediterranean diet: a necessary approach?

"Food pattern" and "diet" sometimes are referred with the same meaning. "Food pattern" can be defined by a priori and a posteriori approaches and refers to the characterisation of foods and frequency of consumption within a population combining with various characteristics. "Diet" may be represented as an individual way of life and considers the food consumption and others like culture, history, society, economy, sustainability, culinary activities, conviviality, physical activity and rest. Mediterranean diet fits on these two concepts. The question is if we are measuring Mediterranean diet as a whole, or whether we are only measuring one of its parts, neglecting the rest of components. Can this compromise the way we monitor its adherence and evolution, with the possibility of losing the perception of certain aspects? How can we preserve and promote the concept if we do not monitor all the pieces of the puzzle?

Dietary patterns associated with inflammatory biomarkers in a Northern German population

PURPOSE

The aim of the present study was to derive overall and sex-specific dietary patterns associated with inflammatory biomarkers in a general population sample from Northern Germany.

METHODS

The present analysis included 1158 participants (477 men, 681 women, mean age: 53.1 years; mean body mass index: 26.2 kg/m²) of the Food Chain Plus (FoCus) cohort in Kiel, Germany. Participants completed a semi-quantitative food frequency questionnaire and provided blood samples. Reduced rank regression with C-reactive protein (CRP) and Interleukin 6 (IL-6) as response variables was used to derive dietary patterns. After a mean follow-up of 1.7 years, a second blood sample was obtained in a subsample of 112 individuals. Multiple regression models were used to examine the association between dietary patterns at baseline and inflammatory biomarkers at follow-up.

RESULTS

The overall pattern characterised by high intakes of soft drinks, meat, potatoes and sauce, and low intakes of other cereals (except pasta/rice), wine, nuts, seeds, vegetarian dishes, vegetable oil, and fish was positively associated with CRP (OR 2.20; 95% CI 1.12, 4.35) and IL-6 (OR 3.14; 95% CI 1.26, 7.87) at follow-up. In men, the dietary pattern was higher in soft drinks, processed meat and low in cereals and plant-based fats. In women, the pattern was characterised by soft drinks, meat, vegetables and low in other cereals, wine, nuts, and seeds. The association between sex-specific patterns with inflammatory biomarkers was weaker for CRP.

CONCLUSION

We identified dietary patterns positively associated with established biomarkers of chronic low-grade inflammation.

Using Cluster Analysis to Identify Subgroups of College Students at Increased Risk for Cardiovascular Disease

BACKGROUND AND PURPOSE

 To examine the co-occurrence of cardiovascular risk factors and cluster subgroups of college students for cardiovascular risks.

METHODS

A cross sectional descriptive study was conducted using co-occurrence patterns and hierarchical clustering analysis in 158 college students.

RESULTS

The top co-occurring cardiovascular risk factors were overweight/obese and hypertension (10.8%, n = 17). Of the total 34 risk factors that co-occurred, 30 of them involved being overweight/obese. A six-cluster-solution was obtained, two clusters displayed elevated levels of lifetime and 30-year cardiovascular disease risks.

CONCLUSIONS

The hierarchical cluster analysis identified that single White males with a family history of heart disease, overweight/obese, hypertensive or diabetes, and occasionally (weekly) consumed red meat, take antihypertensive medication, and hyperlipidemia were considered the higher risk group compared to other subgroups.

Dietary patterns of elderly persons from the city of São Paulo: evidence from the SABE (Health, Wellbeing and Aging) survey

Abstract Objective: to determine the dietary patterns of elderly persons in terms of sociodemographic, lifestyle and clinical aspects. Method: a cross-sectional study was conducted using the cohorts of the SABE (Health, Wellbeing and Aging) study. An epidemiological, home-based study representative of the city of São Paulo was carried out. The population of this study included 1,304 elderly persons (≥60 years), of both genders, selected by probabilistic stratified sampling, interviewed in 2010. Food intake data were obtained through a qualitative food frequency questionnaire. Dietary patterns were determined by exploratory factorial analysis by principal components. The Wald test was used for complex sampling. Results: four dietary patterns were identified: inadequate patterns, consisting of fried and canned food, sausages, sweets, tubers, industrialized sauces and eggs; modified pattern, skimmed milk, whole grain breads and cereals, light/diet/zero foods; beneficial pattern, fruits, vegetables and tubers; and traditional Brazilian pattern, vegetable oils, rice, refined cereals and white bread, meats and legumes (beans). The sociodemographic and lifestyle profile of the elderly persons who adopted each dietary pattern was different. Conclusion: healthier dietary patterns were associated with the female gender, older elderly persons, two or more chronic diseases, higher levels of schooling, a better lifestyle and a better self-perception of health. It is important to consider all these aspects as they are determinant in the type of diet adopted by this population.

"Traditional" and "Healthy" Dietary Patterns Are Associated with Low Cardiometabolic Risk in Brazilian Subjects

This study aimed at determining the dietary patterns and investigating their association with cardiometabolic risk markers in a brazilian population at risk. This transversal study was carried out with data of 265 patients (n = 123 M/172 W, age 42 ± 16 years) of the Cardiovascular Health Care Program-PROCARDIO-UFV, Brazil-who had their first appointment between 2012 and 2017. A 24-hour recall was applied. The dietary patterns were determined by Principal Component Analysis. Anthropometric, clinical-metabolic, sociodemographic, and lifestyle data were collected through medical record analysis. Five patterns were identified: "Traditional", "Caloric", "Unhealthy", "Healthy," and "Healthy Snacks". In bivariate analysis, the "Healthy" pattern was negatively associated with WC (waist circunference), BMI (body mass index), WHR (waist-to-hip ratio), SBP (systolic blood pressure), fasting glucose, TG/HDL, LDL/HDL, and TG/HDL values and positively to HDL. The "Traditional" pattern was positively associated with adiposity indicators (WC, BMI, and WHR) and negatively associated with body fat, TyG (triglyceride-glucose index), HDL, and LDL (P < 0.05). However, in adjusted models of Poisson regression, individuals with positive factor score (higher adherence) in the "Traditional" and "Healthy" patterns had less occurrence of abdominal obesity (PR 0.85; 95% CI 0.74-0.99/PR 0.88; 95% CI 0.02-0.76), as well as dyslipidemia (PR 0.06; 95% CI 0.02-0.51/PR 0.03; 95% CI 0.01-0.27), diabetes (PR 0.05; 95% CI 0.01-0.45/PR 0.02; 95% CI 0.01-021), and hypertension (PR 0.06; 95% CI 0.02-0.50/PR 0.02; 95% CI 0.01-0.21). A greater adherence to the "Healthy" pattern was associated with lower values to cardiometabolic risk markers and less occurrence of chronic diseases, while the "Traditional" pattern presented contradictory results.

Application of a new dietary pattern analysis method in nutritional epidemiology

Background

Diet plays an important role in chronic disease, and the use of dietary pattern analysis has grown rapidly as a way of deconstructing the complexity of nutritional intake and its relation to health. Pattern analysis methods, such as principal component analysis (PCA), have been used to investigate various dimensions of diet. Existing analytic methods, however, do not fully utilize the predictive potential of dietary assessment data. In particular, these methods are often suboptimal at predicting clinically important variables.

Methods

We propose a new dietary pattern analysis method using the advanced LASSO (Least Absolute Shrinkage and Selection Operator) model to improve the prediction of disease-related risk factors. Despite the potential advantages of LASSO, this is the first time that the model has been adapted for dietary pattern analysis. Hence, the systematic evaluation of the LASSO model as applied to dietary data and health outcomes is highly innovative and novel. Using Food Frequency Questionnaire data from NHANES 2005–2006, we apply PCA and LASSO to identify dietary patterns related to cardiovascular disease risk factors in healthy US adults (n = 2609) after controlling for confounding variables (e.g., age and BMI). Both analyses account for the sampling weights. Model performance in terms of prediction accuracy is evaluated using an independent test set.

Results

PCA yields 10 principal components (PCs) that together account for 65% of the variation in the data set and represent distinct dietary patterns. These PCs are then used as predictors in a regression model to predict cardiovascular disease risk factors. We find that LASSO better predicts levels of triglycerides, LDL cholesterol, HDL cholesterol, and total cholesterol (adjusted R² = 0.861, 0.899, 0.890, and 0.935 respectively) than does the traditional, linear-regression-based, dietary pattern analysis method (adjusted R² = 0.163, 0.005, 0.235, and 0.024 respectively) when the latter is applied to components derived from PCA.

Conclusions

The proposed method is shown to be an appropriate and promising statistical means of deriving dietary patterns predictive of cardiovascular disease risk. Future studies, involving different diseases and risk factors, will be necessary before LASSO’s broader usefulness in nutritional epidemiology can be established.

Association between eating time interval and frequency with ideal cardiovascular health: Results from a random sample Czech urban population

BACKGROUND AND AIMS

The frequency and timing of meals may affect cardiovascular health (CVH) outcomes, but large-scale epidemiological studies are lacking. The aim of this study was to understand the relationship between eating time interval and frequency, and measures of ideal CVH in the Kardiovize Brno cohort study, a random urban sample population in Central Europe.

METHODS AND RESULTS

1659 members of the Kardiovize Brno 2030 cohort were included in a cross-sectional study (mean age = 46.86 years; 44.6% male). Exposure variables were eating time interval and frequency, and skipping meals. Primary outcomes were indices of CVH, including body mass index, diet, physical activity, smoking, blood pressure, glucose and cholesterol, and the composite CVH score. Cluster analysis and binary logistic regression analysis were used to evaluate eating habits and the association between variables. After adjustment for well-known risk factors, subjects who skipped breakfast or the afternoon snack had a higher risk of poor CVH (OR = 1.613; 95%CI = 1.121-2.320; p = 0.010; OR = 1.409; 95%CI = 1.110-1.788; p = 0.005, respectively). Moreover, we identified three clusters of individuals based on eating habits; from cluster 1 to cluster 3, eating time interval and frequency increased and this was associated with increases in CVH score from 8.70 (SEM = 0.10) in cluster 1, and 9.06 (SEM = 0.08) in cluster 2 to 9.42 (SEM = 0.09) in cluster 3 (p-trend = 0.019).

CONCLUSIONS

Our findings suggest that skipping breakfast or the afternoon snack are risk factors for poor CVH, while higher eating time interval and frequency may promote ideal CVH.

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.

Fluid intake patterns of children and adolescents: results of six Liq.In7 national cross-sectional surveys

PURPOSE

This study aimed to identify and characterize patterns of fluid intake in children and adolescents from six countries: Argentina, Brazil, China, Indonesia, Mexico and Uruguay.

METHODS

Data on fluid intake volume and type amongst children (4-9 years; N = 1400) and adolescents (10-17 years; N = 1781) were collected using the validated 7-day fluid-specific record (Liq.In7 record). To identify relatively distinct clusters of subjects based on eight fluid types (water, milk and its derivatives, hot beverages, sugar-sweetened beverages (SSB), 100% fruit juices, artificial/non-nutritive sweetened beverages, alcoholic beverages, other beverages), a cluster analysis (partitioning around k-medoids algorithm) was used. Clusters were then characterized according to their socio-demographics and lifestyle indicators.

RESULTS

The six interpretable clusters identified were: low drinkers-SSB (n 523), low drinkers-water and milk (n 615), medium mixed drinkers (n 914), high drinkers-SSB (n 513), high drinkers-water (n 352) and very high drinkers-water (n 264). Country of residence was the dominant characteristic, followed by socioeconomic level, in all six patterns.

CONCLUSIONS

This analysis showed that consumption of water and SSB were the primary drivers of the clusters. In addition to country, socio-demographic and lifestyle factors played a role in determining the characteristics of each cluster. This information highlights the need to target interventions in particular populations aimed at changing fluid intake behavior and improving health in children and adolescents.

Dietary Pattern and Macronutrients Profile on the Variation of Inflammatory Biomarkers: Scientific Update

It is known that the dietary pattern and macronutrients profile may influence the expression and secretion of inflammatory biomarkers, and the low-grade inflammation is associated with the manifestation of noncommunicable chronic diseases. Therefore, this review aimed to present and discuss the role of dietary patterns and macronutrients on the variation of inflammatory markers related to NCD risk. Scientific evidences within the last five years based on clinical trials, case-controls, cohorts, and cross-sectional studies indicate that normocaloric, carbohydrate-moderated, low-glycemic index, protein-moderated, monounsaturated and polyunsaturated fatty acid-rich, omega-3, and low-saturated fat diets display positive effects on the inflammatory state, both in healthy individuals and in those with cardiovascular risk, although the second group seems to benefit more from changes in the dietary profile.

Overall Dietary Intake and Prognosis after Breast Cancer: A Systematic Review

Breast cancer survivors often seek information about how lifestyle factors, such as diet, may influence their prognosis. Previous studies have reviewed evidence around single nutrients, individual foods or food groups. We reviewed studies examining relationships between overall dietary intake and prognosis in breast cancer survivors. A systematic search was conducted to identify studies, published until June 2016, which assessed associations between overall dietary intake (i.e., quality; score; pattern) and mortality and/or recurrence in breast cancer survivors. We identified seven eligible studies. Studies were heterogeneous regarding diet assessment timing (before/after diagnosis); mean age and menopausal status; and dietary intake measure (statistically derived/a priori defined indices). Better overall dietary intake (i.e., better quality; healthy/prudent pattern; less inflammatory diet) was associated with decreased risk of overall and non-breast cancer mortality, in most studies. Insufficient evidence is available to draw conclusions regarding breast cancer-specific survival and disease recurrence. Following breast cancer diagnosis, better overall dietary intake may independently improve overall and non-breast cancer survival. Survivors may improve prognosis by adopting more healthful dietary patterns consistent with dietary guidelines and/or prudent diet. Future adequately powered studies should consider measuring dietary intake consistently to better understand the role of diet in disease-specific outcomes.

Different Dietary Patterns and New-Onset Diabetes Mellitus After Kidney Transplantation: A Cross-Sectional Study

OBJECTIVE

To verify the association of dietary patterns and dietary components with new-onset diabetes mellitus after transplantation (NODAT).

DESIGN

Cross-sectional study.

SUBJECTS

Adult kidney transplant recipients, without history of diabetes before transplantation, who received a kidney transplant and were followed up for at least 1 year. One hundred and sixteen subjects recruited between January 2013 and August 2014. Diagnosis of NODAT was established according to the American Diabetes Association criteria for type 2 diabetes.

METHODS

Demographic, clinical, and anthropometric data were collected. Dietary intake was assessed by food frequency questionnaire, administered by a registered dietitian. Dietary patterns were identified by cluster analysis. Chi-square test was used to verify the association between dietary patterns and NODAT. Total energy, fiber, and cholesterol intake were calculated. Consumption of macronutrients, carbohydrates, proteins, and fats (total fats and saturated, monounsaturated, polyunsaturated and trans fatty acids), were expressed in percentage of total energy intake.

RESULTS

Twenty-eight patients developed NODAT in the follow-up period. They presented higher body mass index and body fat percentage, as well as higher levels of triglycerides and urinary protein/creatinine ratio than the non-NODAT group. Two dietary patterns, I and II, were identified. Pattern II was characterized by higher intake of total, saturated, monounsaturated, and trans fats than pattern I. No association between the dietary patterns and NODAT was identified (P = .905), and there was no difference in the distribution of macronutrients, dietary fiber, and dietary cholesterol between the groups with and without NODAT.

CONCLUSION

Posttransplant dietary patterns were not different between patients with and without NODAT. Further larger and prospective studies are needed to evaluate a possible relationship between dietary components and NODAT incidence in kidney transplant recipients.

[Methods of a posteriori identification of food patterns in Brazilian children: a systematic review]

The objective of this study is to provide guidance for identifying dietary patterns using the a posteriori approach, and analyze the methodological aspects of the studies conducted in Brazil that identified the dietary patterns of children. Articles were selected from the Latin American and Caribbean Literature on Health Sciences, Scientific Electronic Library Online and Pubmed databases. The key words were: Dietary pattern; Food pattern; Principal Components Analysis; Factor analysis; Cluster analysis; Reduced rank regression. We included studies that identified dietary patterns of children using the a posteriori approach. Seven studies published between 2007 and 2014 were selected, six of which were cross-sectional and one cohort, Five studies used the food frequency questionnaire for dietary assessment; one used a 24-hour dietary recall and the other a food list. The method of exploratory approach used in most publications was principal components factor analysis, followed by cluster analysis. The sample size of the studies ranged from 232 to 4231, the values of the Kaiser-Meyer-Olkin test from 0.524 to 0.873, and Cronbach's alpha from 0.51 to 0.69. Few Brazilian studies identified dietary patterns of children using the a posteriori approach and principal components factor analysis was the technique most used.

Diet quality is associated with reduced incidence of cancer and self-reported chronic disease: Observations from Alberta's Tomorrow Project

The objective of this study was to assess diet quality using the Healthy Eating Index-2005 Canada (HEI-2005-Canada) and its association with risk of cancer and chronic disease in a sample of Alberta's Tomorrow Project (ATP) participants. Food frequency questionnaires completed by 25,169 participants (38% men; mean age 50.3 (9.2)) enrolled between 2000 and 2008 were used to calculate HEI-2005-Canada scores. Data from a subset of participants (n=10,735) who reported no chronic disease at enrollment were used to investigate the association between HEI-2005-Canada score and development of self-reported chronic disease at follow-up (2008). Participants were divided into HEI-2005-Canada score quartiles. Cox proportional hazards models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for cancer and chronic disease incidence. In this cohort, mean HEI-2005-Canada scores for men and women were 50.9 and 55.5 (maximum range 0-100), respectively. In men, higher HEI-2005-Canada score (Q4 vs. Q1) was associated with lower cancer risk (HR (95% CI) 0.63 (0.49-0.83)) over the course of follow-up (mean (SD)=10.4 (2.3) years); the same was not observed in women. In contrast, higher overall HEI-2005-Canada score (Q4 vs. Q1) was associated with lower risk of self-reported chronic disease (0.85 (0.75-0.97)) in both men and women over follow-up (4.2 (2.3) years). In conclusion, in this cohort better diet quality was associated with a lower risk of cancer in men and lower risk of chronic disease in both sexes. Future studies with longer follow-up and repeated measures of diet may be helpful to elucidate sex-specific associations between dietary quality and disease outcomes.

Household food group expenditure patterns are associated with child anthropometry at ages 5, 8 and 12 years in Ethiopia, India, Peru and Vietnam

Population-level analysis of dietary influences on nutritional status is challenging in part due to limitations in dietary intake data. Household expenditure surveys, covering recent household expenditures and including key food groups, are routinely conducted in low- and middle-income countries. These data may help identify patterns of food expenditure that relate to child growth.

OBJECTIVES

We investigated the relationship between household food expenditures and child growth using factor analysis.

METHODS

We used data on 6993 children from Ethiopia, India, Peru and Vietnam at ages 5, 8 and 12y from the Young Lives cohort. We compared associations between household food expenditures and child growth (height-for-age z scores, HAZ; body mass index-for-age z scores, BMI-Z) using total household food expenditures and the "household food group expenditure index" (HFGEI) extracted from household expenditures with factor analysis on the seven food groups in the child dietary diversity scale, controlling for total food expenditures, child dietary diversity, data collection round, rural/urban residence and child sex. We used the HFGEI to capture households' allocations of their finances across food groups in the context of local food pricing, availability and pReferences RESULTS: The HFGEI was associated with significant increases in child HAZ in Ethiopia (0.07), India (0.14), and Vietnam (0.07) after adjusting for all control variables. Total food expenditures remained significantly associated with increases in BMI-Z for India (0.15), Peru (0.11) and Vietnam (0.06) after adjusting for study round, HFGEI, dietary diversity, rural residence, and whether the child was female. Dietary diversity was inversely associated with BMI-Z in India and Peru. Mean dietary diversity increased from age 5y to 8y and decreased from age 8y to 12y in all countries.

CONCLUSION

Household food expenditure data provide insights into household food purchasing patterns that significantly predict HAZ and BMI-Z. Including food expenditure patterns data in analyses may yield important information about child nutritional status and linear growth.