A Model of Individual BMI Trajectories.

A risk factor model of body mass index (BMI) is an important building block of health simulations aimed at estimating government policy effects with regard to overweight and obesity. We created a model that generates representative population level distributions and that also mimics realistic BMI trajectories at an individual level so that policies aimed at individuals can be simulated. The model is constructed by combining several datasets. First, the population level distribution is extracted from a large, cross-sectional dataset. The trend in this distribution is estimated from historical data. In addition, longitudinal data are used to model how individuals move along typical trajectories over time. The model faithfully describes the population level distribution of BMI, stratified by sex, level of education and age. It is able to generate life course trajectories for individuals which seem plausible, but it does not capture extreme fluctuations, such as rapid weight loss.

Relative validity of habitual sugar and low/no-calorie sweetener consumption assessed by food frequency questionnaire, multiple 24-h dietary recalls and urinary biomarkers: an observational study within the SWEET project.

BACKGROUND: Studies investigating associations between sweeteners and health yield inconsistent results, possibly due to subjective self-report dietary assessment methods. OBJECTIVES: We compared the performance of a food frequency questionnaire (FFQ), multiple 24-h dietary recalls (24hRs), and urinary biomarkers to estimate intake of sugars and low/no-calorie sweeteners (LNCSs). METHODS: Participants (n = 848, age 54 ± 12 y) from a 2-y observational study completed 1 semiquantitative FFQ and ≥ 3 nonconsecutive 24hRs. Both methods assessed intake of sugars (mono- and disaccharides, sucrose, fructose, free and added sugars) and sweetened foods and beverages (sugary foods, fruit juice, and sugar or LNCS-containing beverages [sugar-sweetened beverages and low/no-calorie sweetened beverages (LNCSBs)]); 24hRs also included LNCS-containing foods and tabletop sweeteners (low/no-calorie sweetened foods [LNCSFs]). Urinary excretion of sugars (fructose+sucrose) and LNCSs (acesulfame K+sucralose+steviol glucuronide+cyclamate+saccharin) were simultaneously assessed using ultrapressure liquid chromatography coupled to tandem mass spectrometry in 288 participants with 3 annual 24-h urine samples. Methods were compared using, amongst others, validity coefficients (correlations corrected for measurement error). RESULTS: Median (interquartile range) FFQ intakes ranged from 0 (0-7) g/d for LNCSBs to 94 (73-117) g/d for mono- and disaccharides. LNCSB use was reported by 32% of participants. Median LNCSB+LNCSF intake using 24hRs was 1 (0-50) g/d and reported by 58%. Total sugar excretions were detected in 100% of samples [56 (37-85) mg/d] and LNCSs in 99% of urine samples [3 (1-10) mg/d]. Comparing FFQ against 24hRs showed VCs ranging from 0.38 (fruit juice) to 0.74 (LNCSB). VCs for comparing FFQ with urinary excretions were 0.25 to 0.29 for sugars and 0.39 for LNCSBs; for 24hR they amounted to 0.31-0.38 for sugars, 0.37 for LNCSBs, and 0.45 for LNCSFs. CONCLUSIONS: The validity of the FFQ against 24hRs for the assessment of sugars and LNCSBs ranged from moderate to good. Comparing self-reports and urine excretions showed moderate agreement but highlighted an important underestimation of LNCS exposure using self-reports.

Constructing synthetic populations in the age of big data.

BACKGROUND: To develop public health intervention models using micro-simulations, extensive personal information about inhabitants is needed, such as socio-demographic, economic and health figures. Confidentiality is an essential characteristic of such data, while the data should reflect realistic scenarios. Collection of such data is possible only in secured environments and not directly available for open-source micro-simulation models. The aim of this paper is to illustrate a method of construction of synthetic data by predicting individual features through models based on confidential data on health and socio-economic determinants of the entire Dutch population. METHODS: Administrative records and health registry data were linked to socio-economic characteristics and self-reported lifestyle factors. For the entire Dutch population (n = 16,778,708), all socio-demographic information except lifestyle factors was available. Lifestyle factors were available from the 2012 Dutch Health Monitor (n = 370,835). Regression model was used to sequentially predict individual features. RESULTS: The synthetic population resembles the original confidential population. Features predicted in the first stages of the sequential procedure are virtually similar to those in the original population, while those predicted in later stages of the sequential procedure carry the accumulation of limitations furthered by data quality and previously modelled features. CONCLUSIONS: By combining socio-demographic, economic, health and lifestyle related data at individual level on a large scale, our method provides us with a powerful tool to construct a synthetic population of good quality and with no confidentiality issues.

Predicting self-perceived general health status using machine learning: an external exposome study.

BACKGROUND: Self-perceived general health (SPGH) is a general health indicator commonly used in epidemiological research and is associated with a wide range of exposures from different domains. However, most studies on SPGH only investigated a limited set of exposures and did not take the entire external exposome into account. We aimed to develop predictive models for SPGH based on exposome datasets using machine learning techniques and identify the most important predictors of poor SPGH status. METHODS: Random forest (RF) was used on two datasets based on personal characteristics from the 2012 and 2016 editions of the Dutch national health survey, enriched with environmental and neighborhood characteristics. Model performance was determined using the area under the curve (AUC) score. The most important predictors were identified using a variable importance procedure and individual effects of exposures using partial dependence and accumulated local effect plots. The final 2012 dataset contained information on 199,840 individuals and 81 variables, whereas the final 2016 dataset had 244,557 individuals with 91 variables. RESULTS: Our RF models had overall good predictive performance (2012: AUC = 0.864 (CI: 0.852-0.876); 2016: AUC = 0.890 (CI: 0.883-0.896)) and the most important predictors were "Control of own life", "Physical activity", "Loneliness" and "Making ends meet". Subjects who felt insufficiently in control of their own life, scored high on the De Jong-Gierveld loneliness scale or had difficulty in making ends meet were more likely to have poor SPGH status, whereas increased physical activity per week reduced the probability of poor SPGH. We observed associations between some neighborhood and environmental characteristics, but these variables did not contribute to the overall predictive strength of the models. CONCLUSIONS: This study identified that within an external exposome dataset, the most important predictors for SPGH status are related to mental wellbeing, physical exercise, loneliness, and financial status.

Validation of the smartphone-based dietary assessment tool "Traqq" for assessing actual dietary intake by repeated 2-h recalls in adults: comparison with 24-h recalls and urinary biomarkers.

BACKGROUND: Conventional dietary assessment methods are affected by measurement errors. We developed a smartphone-based 2-h recall (2hR) methodology to reduce participant burden and memory-related bias. OBJECTIVE: Assessing the validity of the 2hR method against traditional 24-h recalls (24hRs) and objective biomarkers. METHODS: Dietary intake was assessed in 215 Dutch adults on 6 randomly selected nonconsecutive days (i.e., 3 2hR-days and 3 24hRs) during a 4-wk period. Sixty-three participants provided 4 24-h urine samples, to assess urinary nitrogen and potassium concentrations. RESULTS: Intake estimates of energy (2052±503 kcal vs. 1976±483 kcal) and nutrients (e.g., protein: 78±23 g vs. 71±19 g; fat: 84±30 g vs. 79±26 g; carbohydrates: 220±60 g vs. 216±60 g) were slightly higher with 2hR-days than with 24hRs. Comparing self-reported protein and potassium intake to urinary nitrogen and potassium concentrations indicated a slightly higher accuracy of 2hR-days than 24hRs (protein: -14% vs. -18%; potassium: -11% vs. -16%). Correlation coefficients between methods ranged from 0.41 to 0.75 for energy and macronutrients and from 0.41 to 0.62 for micronutrients. Generally, regularly consumed food groups showed small differences in intake (<10%) and good correlations (>0.60). Intake of energy, nutrients, and food groups showed similar reproducibility (intraclass correlation coefficient) for 2hR-days and 24hRs. CONCLUSIONS: Comparing 2hR-days with 24hRs showed a relatively similar group-level bias for energy, most nutrients, and food groups. Differences were mostly due to higher intake estimates by 2hR-days. Biomarker comparisons showed less underestimation by 2hR-days as compared with 24hRs, suggesting that 2hR-days are a valid approach to assess the intake of energy, nutrients, and food groups. This trial was registered at the Dutch Central Committee on Research Involving Human Subjects (CCMO) registry as ABR. No. NL69065.081.19.

Pitfalls in the statistical analysis of microbiome amplicon sequencing data.

Microbiome data are characterized by several aspects that make them challenging to analyse statistically: they are compositional, high dimensional and rich in zeros. A large array of statistical methods exist to analyse these data. Some are borrowed from other fields, such as ecology or RNA-sequencing, while others are custom-made for microbiome data. The large range of available methods, and which is continuously expanding, means that researchers have to invest considerable effort in choosing what method(s) to apply. In this paper we list 14 statistical methods or approaches that we think should be generally avoided. In several cases this is because we believe the assumptions behind the method are unlikely to be met for microbiome data. In other cases we see methods that are used in ways they are not intended to be used. We believe researchers would be helped by more critical evaluations of existing methods, as not all methods in use are suitable or have been sufficiently reviewed. We hope this paper contributes to a critical discussion on what methods are appropriate to use in the analysis of microbiome data.

Diverging metabolic effects of 2 energy-restricted diets differing in nutrient quality: a 12-week randomized controlled trial in subjects with abdominal obesity.

BACKGROUND: Despite the established relation between energy restriction (ER) and metabolic health, the most beneficial nutrient composition of a weight-loss diet is still a subject of debate. OBJECTIVES: The aim of the study was to examine the additional effects of nutrient quality on top of ER. METHODS: A parallel-designed, 12-week 25% ER dietary intervention study was conducted (clinicaltrials.gov: NCT02194504). Participants aged 40-70 years with abdominal obesity were randomized over 3 groups: a 25% ER high-nutrient-quality diet (n = 40); a 25% ER low-nutrient-quality diet (n = 40); or a habitual diet (n = 30). Both ER diets were nutritionally adequate, and the high-nutrient-quality ER diet was enriched in MUFAs, n-3 PUFAs, fiber, and plant protein and reduced in fructose. Before and after the intervention, intrahepatic lipids, body fat distribution, fasting and postprandial responses to a mixed-meal shake challenge test of cardiometabolic risk factors, lipoproteins, vascular measurements, and adipose tissue transcriptome were assessed. RESULTS: The high-nutrient-quality ER diet (-8.4 ± 3.2) induced 2.1 kg more weight loss (P = 0.007) than the low-nutrient-quality ER diet (-6.3 ± 3.9), reduced fasting serum total cholesterol (P = 0.014) and plasma triglycerides (P < 0.001), promoted an antiatherogenic lipoprotein profile, and induced a more pronounced decrease in adipose tissue gene expression of energy metabolism pathways than the low-quality ER diet. Explorative analyses showed that the difference in weight loss between the two ER diets was specifically present in insulin-sensitive subjects (HOMA-IR ≤ 2.5), in whom the high-nutrient-quality diet induced 3.9 kg more weight loss than the low-nutrient-quality diet. CONCLUSIONS: A high-nutrient-quality 25% ER diet is more beneficial for cardiometabolic health than a low-nutrient-quality 25% ER diet. Overweight, insulin-sensitive subjects may benefit more from a high- than a low-nutrient-quality ER diet with respect to weight loss, due to potential attenuation of glucose-induced lipid synthesis in adipose tissue.

Mapping chronic disease prevalence based on medication use and socio-demographic variables: an application of LASSO on administrative data sources in healthcare in the Netherlands.

BACKGROUND: Policymakers generally lack sufficiently detailed health information to develop localized health policy plans. Chronic disease prevalence mapping is difficult as accurate direct sources are often lacking. Improvement is possible by adding extra information such as medication use and demographic information to identify disease. The aim of the current study was to obtain small geographic area prevalence estimates for four common chronic diseases by modelling based on medication use and socio-economic variables and next to investigate regional patterns of disease. METHODS: Administrative hospital records and general practitioner registry data were linked to medication use and socio-economic characteristics. The training set (n = 707,021) contained GP diagnosis and/or hospital admission diagnosis as the standard for disease prevalence. For the entire Dutch population (n = 16,777,888), all information except GP diagnosis and hospital admission was available. LASSO regression models for binary outcomes were used to select variables strongly associated with disease. Dutch municipality (non-)standardized prevalence estimates for stroke, CHD, COPD and diabetes were then based on averages of predicted probabilities for each individual inhabitant. RESULTS: Adding medication use data as a predictor substantially improved model performance. Estimates at the municipality level performed best for diabetes with a weighted percentage error (WPE) of 6.8%, and worst for COPD (WPE 14.5%)Disease prevalence showed clear regional patterns, also after standardization for age. CONCLUSION: Adding medication use as an indicator of disease prevalence next to socio-economic variables substantially improved estimates at the municipality level. The resulting individual disease probabilities could be aggregated into any desired regional level and provide a useful tool to identify regional patterns and inform local policy.

Identification of Lifestyle Behaviors Associated with Recurrence and Survival in Colorectal Cancer Patients Using Random Survival Forests.

Current lifestyle recommendations for cancer survivors are the same as those for the general public to decrease their risk of cancer. However, it is unclear which lifestyle behaviors are most important for prognosis. We aimed to identify which lifestyle behaviors were most important regarding colorectal cancer (CRC) recurrence and all-cause mortality with a data-driven method. The study consisted of 1180 newly diagnosed stage I-III CRC patients from a prospective cohort study. Lifestyle behaviors included in the current recommendations, as well as additional lifestyle behaviors related to diet, physical activity, adiposity, alcohol use, and smoking were assessed six months after diagnosis. These behaviors were simultaneously analyzed as potential predictors of recurrence or all-cause mortality with Random Survival Forests (RSFs). We observed 148 recurrences during 2.6-year median follow-up and 152 deaths during 4.8-year median follow-up. Higher intakes of sugary drinks were associated with increased recurrence risk. For all-cause mortality, fruit and vegetable, liquid fat and oil, and animal protein intake were identified as the most important lifestyle behaviors. These behaviors showed non-linear associations with all-cause mortality. Our exploratory RSF findings give new ideas on potential associations between certain lifestyle behaviors and CRC prognosis that still need to be confirmed in other cohorts of CRC survivors.

Lifestyle after colorectal cancer diagnosis in relation to recurrence and all-cause mortality.

BACKGROUND: An unhealthy lifestyle is associated with the risk of colorectal cancer (CRC), but it is unclear whether overall lifestyle after a CRC diagnosis is associated with risks of recurrence and mortality. OBJECTIVES: To examine associations between postdiagnosis lifestyle and changes in lifestyle after a CRC diagnosis with risks of CRC recurrence and all-cause mortality. METHODS: The study population included 1425 newly diagnosed, stage I-III CRC patients from 2 prospective cohort studies enrolled between 2010 and 2016. Lifestyle, including BMI, physical activity, diet, and alcohol intake, was assessed at diagnosis and at 6 months postdiagnosis. We assigned lifestyle scores based on concordance with 2 sets of cancer prevention guidelines-from the World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) and the American Cancer Society (ACS)-and national disease prevention guidelines. Higher scores indicate healthier lifestyles. We computed adjusted HRs and 95% CIs using Cox regression. RESULTS: We observed 164 recurrences during a 2.8-year median follow-up and 171 deaths during a 4.4-year median follow-up. No associations were observed for CRC recurrence. A lifestyle more consistent with the ACS recommendations was associated with a lower all-cause mortality risk (HR per +1 SD, 0.85; 95% CI: 0.73-0.995). The same tendency was observed for higher WCRF/AICR (HR, 0.92; 95% CI: 0.78-1.08) and national (HR, 0.90; 95% CI: 0.77-1.05) lifestyle scores, although these associations were statistically nonsignificant. Generally, no statistically significant associations were observed for BMI, physical activity, diet, or alcohol. Improving one's lifestyle after diagnosis (+1 SD) was associated with a lower all-cause mortality risk for the ACS (HR, 0.80; 95% CI: 0.67-0.96) and national (HR, 0.84; 95% CI: 0.70-0.999) scores, yet was statistically nonsignificant for the WCRF/AICR score (HR, 0.94; 95% CI: 0.78-1.13). CONCLUSIONS: A healthy lifestyle after CRC diagnosis and improvements therein were not associated with the risk of CRC recurrence, but were associated with a decreased all-cause mortality risk.

Validity Coefficient of Repeated Measurements of Urinary Marker of Sugar Intake Is Comparable to Urinary Nitrogen as Marker of Protein Intake in Free-living Subjects.

BACKGROUND: Studies do not show consistent relationships between self-reported intake of sugar and outcome of disease. To overcome the drawbacks of self-reported intake methods, we investigated whether there is an agreement in ranking of individuals between their self-reported sugar intake and urinary sucrose and fructose. METHODS: We used data of 198 Dutch adults (106 women) from the DUPLO study. Sugar intake of all foods and drinks consumed over 24-hour period was estimated by collecting duplicate portions (DP) and 24-hour recalls (24hR), telephone (24hRT) and Web-based (24hRW), while sugar excretion was based on 24-hour urine samples. Sugar content of 24hR was calculated using a newly developed sugar database and sugar content of DPs and urine samples was calculated using high-performance liquid chromatography-atomic emission spectrometry and LC/MS-MS, respectively. Measurement error models assessed validity coefficients (VC) and attenuation factors (AF). Coefficients were compared with those of protein biomarker. RESULTS: The VC for the marker, using DP as reference, showed comparability with substantially better ranking of participants (0.72 for women and 0.93 for men), than 24hRT (0.57 and 0.78) or 24hRW (0.70 and 0.78) as reference in the sucrose models. The VC of the sucrose models was within 10% of the protein models, except for the model with 24hRT as reference, among women. The AF started at higher values and increased by a greater factor compared with the VC. CONCLUSIONS: Repeated measurements of urinary sucrose and fructose as a marker of daily sucrose intake had a ranking performance comparable to urinary nitrogen as marker of protein intake in free-living Dutch adults. IMPACT: The validation of the sugar biomarker in a free-living population with three different dietary assessment methods and its comparable ranking ability with a good recovery biomarker (i.e., protein biomarker) have important research applications. The biomarker may be used for validating dietary assessment methods, for monitoring compliance in human feeding studies, for monitoring the effect of public health interventions, and as a surrogate for ranking subjects according to sucrose intake when information on sucrose in food composition databases is lacking.

Low Meat Consumption in the Netherlands Is Associated With Higher Intake of Fish, Nuts, Seeds, Cheese, Sweets, and Snacks: Results From a Two-Part Model.

Studies on sustainable diets show a need for replacement of animal-based foods by plant-based foods, which is also called "the protein transition." To gain insight into the acceptability of such diet shifts, this study evaluated which current food sources people consume at varying amounts of meat consumption. The study population consisted of 4,313 participants aged 1-79 years of the Dutch National Food Consumption Survey 2012-2016, which assessed diet using two nonconsecutive 24-h dietary recalls. A two-part statistical model was used that accounts for both repeated measures and the correlation between probability and amount of consumption. Results are presented for quartiles of low to high meat consumption, by age and sex. Depending on age and sex, a higher consumption of fish (>100%), nuts and seeds (73-156%), cheese (34-111%), and sweets and snacks (28-81%) is observed in the lowest quartile of meat consumption compared to the highest. For fish, nuts, seeds, and cheese, this increase is mainly due to probability of consumption (>100%, 61-93%, and 16-64%, respectively). For sweets and snacks, the increase is mainly due to the amount of consumption (26-72%). Probability of potato consumption is 29-51% lower at low meat consumption. Vegetable consumption is lower mainly due to amount of consumption (6-29%). The results from the two-part model suggest that shifting away from a traditional Dutch high meat-vegetable-potatoes pattern is associated with higher probability of consuming fish, nuts and seeds, and cheese, but also increased amounts of sweets and snacks. This illustrates that analyzing the probability and amount part separately in relation to behavioral or physiological determinants extends our understanding of the diet according to meat consumption. These insights are important when developing realistic and acceptable food-based dietary guidelines for meat reduction.

Associations of Abdominal Skeletal Muscle Mass, Fat Mass, and Mortality among Men and Women with Stage I-III Colorectal Cancer.

BACKGROUND: The associations of abdominal skeletal muscle mass index (SMI), visceral and subcutaneous adipose tissue (VAT and SAT, respectively), and mortality among patients with stage I-III colorectal cancer may differ for men and women, but only few studies stratified their data into men and women. We investigated associations of abdominal SMI, VAT, and SAT with overall mortality among men and among women with stage I-III colorectal cancer. METHODS: SMI, VAT, and SAT were assessed from abdominal CT images for 1,998 patients with stage I-III colorectal cancer diagnosed between 2006 and 2015. Restricted cubic splines (RCS) were used to investigate associations of SMI, VAT, and SAT with overall mortality. RESULTS: Average age of the participants was 67.9 ± 10.6 years and 58% were men. During a median follow-up of 4.3 years, 546 (27%) patients died. Among men, the association of SMI and mortality was statistically significant in a nonlinear way in the RCS analyses, with lower SMI levels associated with higher mortality. SMI was not associated with mortality among women. SAT was associated with mortality in a nonlinear way for men and for women, with lower SAT levels being associated with higher mortality. VAT was not significantly associated with mortality in men or women. CONCLUSION: Associations of abdominal skeletal muscle mass with mortality among patients with colorectal cancer were not the same for men and for women. IMPACT: This study stresses the importance for more attention on sex-related differences in body composition and cancer outcomes.

Evaluating the Validity of a Food Frequency Questionnaire in Comparison with a 7-Day Dietary Record for Measuring Dietary Intake in a Population of Survivors of Colorectal Cancer.

BACKGROUND: Food frequency questionnaires (FFQs) are a commonly used method to assess dietary intake in epidemiological studies. It is important to evaluate the validity of FFQs in the population of interest. OBJECTIVE: To evaluate the validity of an FFQ for measuring dietary intake in survivors of colorectal cancer (CRC), relative to a 7-day dietary record. DESIGN: Dietary intake was assessed 1 year after the end of CRC treatment. Participants first completed a 7-day dietary record and 2 weeks later a 253-item FFQ that measured intake in the preceding month. PARTICIPANTS/SETTING: Data were used from a subsample of participants (n=100) enrolled in an ongoing prospective study (EnCoRe study) in the Netherlands, from 2015 to 2018. MAIN OUTCOME MEASURES: Estimated intakes of total energy, 19 nutrients, and 20 food groups as well as scoring adherence to the dietary recommendations of the World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) were compared between both dietary assessment methods. STATISTICAL ANALYSES PERFORMED: Means and standard deviations, Spearman rank correlations corrected for within-person variation and total energy, and κ agreement between quintiles were assessed. RESULTS: The median Spearman correlation corrected for within-person variation for nutrients and total energy was 0.60. Correlations >0.50 were found for 15 of 19 nutrients, with highest agreement for vitamin B-12 (0.74), polysaccharides (0.75), and alcohol (0.91). On average, 73% (range=60% to 84%) of participants were classified into the exact same or adjacent nutrient quintile. The median Spearman correlation corrected for within-person variation for food groups was 0.62. Correlations >0.50 were found for 17 of 20 food groups, with highest agreement for cereals and cereal products (0.96), fish (0.96), and potatoes (0.99). The Spearman correlation between total scores of the WCRF/AICR dietary recommendations was 0.53. CONCLUSIONS: Relative to a 7-day dietary record, the validity of an FFQ for measuring dietary intake among survivors of CRC appeared moderate to good for most nutrients and food groups.

Potential impact of reduced tobacco use on life and health expectancies in Belgium.

OBJECTIVES: We investigated the potential impact of reduced tobacco use scenarios on total life expectancy and health expectancies, i.e., healthy life years and unhealthy life years. METHODS: Data from the Belgian Health Interview Survey 2013 were used to estimate smoking and disability prevalence. Disability was based on the Global Activity Limitation Indicator. We used DYNAMO-HIA to quantify the impacts of risk factor changes and to compare the "business-as-usual" with alternative scenarios. RESULTS: The "business-as-usual" scenario estimated that in 2028 the 15-year-old men/women would live additional 50/52 years without disability and 14/17 years with disability. The "smoking-free population" scenario added 3.4/2.8 healthy life years and reduced unhealthy life years by 0.79/1.9. Scenarios combining the prevention of smoking initiation with smoking cessation programs are the most effective, yielding the largest increase in healthy life years (1.9/1.7) and the largest decrease in unhealthy life years (- 0.80/- 1.47). CONCLUSIONS: Health impact assessment tools provide different scenarios for evidence-informed public health actions. New anti-smoking strategies or stricter enforcement of existing policies potentially gain more healthy life years and reduce unhealthy life years in Belgium.

Colorectal cancer survivors only marginally change their overall lifestyle in the first 2 years following diagnosis.

PURPOSE: A healthy lifestyle after colorectal cancer (CRC) diagnosis may improve prognosis. Data related to lifestyle change in CRC survivors are inconsistent and potential interrelated changes are unknown. METHODS: We assessed dietary intake, physical activity, body mass index (BMI), waist circumference, and smoking among 1072 patients diagnosed with stages I-III CRC at diagnosis, 6 months and 2 years post-diagnosis. An overall lifestyle score was constructed based on the 2018 World Cancer Research Fund/American Institute of Cancer Research recommendations (range 0-7). We used linear mixed models to analyze changes in lifestyle over time. RESULTS: Participants had a mean (± SD) age of 65 ± 9 years and 43% had stage III disease. In the 2 years following CRC diagnosis, largest changes were noted for sugary drinks (- 45 g/day) and red and processed meat intake (- 62 g/week). BMI (+ 0.4 kg/m2), waist circumference (+ 2 cm), and dietary fiber intake (- 1 g/day) changed slightly. CRC survivors did not statistically significant change their mean intake of fruits and vegetables, alcohol, or ultra-processed foods nor did they change their physical activity or smoking behavior. Half of participants made simultaneous changes that resulted in improved concordance with one component as well as deteriorated concordance with another component of the lifestyle score. Overall lifestyle score changed from a mean 3.4 ± 0.9 at diagnosis to 3.5 ± 0.9 2 years post-diagnosis. CONCLUSIONS: CRC survivors hardly improve their overall lifestyle after diagnosis. IMPLICATIONS FOR CANCER SURVIVORS: Given the importance of a healthy lifestyle, strategies to effectively support behavior changes in CRC survivors need to be identified.

Using enhanced regression calibration to combine dietary intake estimates from 24 h recall and FFQ reduces bias in diet-disease associations.

OBJECTIVE: To illustrate the impact of combining 24 h recall (24hR) and FFQ estimates using regression calibration (RC) and enhanced regression calibration (ERC) on diet-disease associations. SETTING: Wageningen area, the Netherlands, 2011-2013. DESIGN: Five approaches for obtaining self-reported dietary intake estimates of protein and K were compared: (i) uncorrected FFQ intakes (FFQ); (ii) uncorrected average of two 24hR ( $\overline {\rm R}$ ); (iii) average of FFQ and $\overline {\rm R}$ ( ${\overline {\rm F}}\,\overline {\rm R}}$ ); (iv) RC from regression of 24hR v. FFQ; and (v) ERC by adding individual random effects to the RC approach. Empirical attenuation factors (AF) were derived by regression of urinary biomarker measurements v. the resulting intake estimates. PARTICIPANTS: Data of 236 individuals collected within the National Dietary Assessment Reference Database. RESULTS: Both FFQ and 24hR dietary intake estimates were measured with substantial error. Using statistical techniques to correct for measurement error (i.e. RC and ERC) reduced bias in diet-disease associations as indicated by their AF approaching 1 (RC 1·14, ERC 0·95 for protein; RC 1·28, ERC 1·34 for K). The larger sd and narrower 95% CI of AF obtained with ERC compared with RC indicated that using ERC has more power than using RC. However, the difference in AF between RC and ERC was not statistically significant, indicating no significantly better de-attenuation by using ERC compared with RC. AF larger than 1, observed for the ERC for K, indicated possible overcorrection. CONCLUSIONS: Our study highlights the potential of combining FFQ and 24hR data. Using RC and ERC resulted in less biased associations for protein and K.

Importance of details in food descriptions in estimating population nutrient intake distributions.

BACKGROUND: National food consumption surveys are important policy instruments that could monitor food consumption of a certain population. To be used for multiple purposes, this type of survey usually collects comprehensive food information using dietary assessment methods like 24-h dietary recalls (24HRs). However, the collection and handling of such detailed information require tremendous efforts. We aimed to improve the efficiency of data collection and handling in 24HRs, by identifying less important characteristics of food descriptions (facets) and assessing the impact of disregarding them on energy and nutrient intake distributions. METHODS: In the Dutch National Food Consumption Survey 2007-2010, food consumption data were collected through interviewer-administered 24HRs using GloboDiet software in 3819 persons. Interviewers asked participants about the characteristics of each food item according to applicable facets. Food consumption data were subsequently linked to the food composition database. The importance of facets for predicting energy and each of the 33 nutrients was estimated using the random forest algorithm. Then a simulation study was performed to determine the influence of deleting less important facets on population nutrient intake distributions. RESULTS: We identified 35% facets as unimportant and deleted them from the total food consumption database. The majority (79.4%) of the percent difference between percentile estimates of the population nutrient intake distributions before and after facet deletion ranged from 0 to 1%, while 20% cases ranged from 1 to 5% and 0.6% cases more than 10%. CONCLUSION: We concluded that our procedure was successful in identifying less important food descriptions in estimating population nutrient intake distributions. The reduction in food descriptions has the potential to reduce the time needed for conducting interviews and data handling while maintaining the data quality of the survey.

Potential gains in health expectancy by improving lifestyle: an application for European regions.

BACKGROUND: Prevention aiming at smoking, alcohol consumption, and BMI could potentially bring large gains in life expectancy (LE) and health expectancy measures such as Healthy Life Years (HLY) and Life Expectancy in Good Perceived Health (LEGPH) in the European Union. However, the potential gains might differ by region. METHODS: A Sullivan life table model was applied for 27 European countries to calculate the impact of alternative scenarios of lifestyle behavior on life and health expectancy. Results were then pooled over countries to present the potential gains in HLY and LEGPH for four European regions. RESULTS: Simulations show that up to 4 years of extra health expectancy can be gained by getting all countries to the healthiest levels of lifestyle observed in EU countries. This is more than the 2 years to be gained in life expectancy. Generally, Eastern Europe has the lowest LE, HLY, and LEGPH. Even though the largest gains in LEPGH and HLY can also be made in Eastern Europe, the gap in LE, HLY, and LEGPH can only in a small part be closed by changing smoking, alcohol consumption, and BMI. CONCLUSION: Based on the current data, up to 4 years of good health could be gained by adopting lifestyle as seen in the best-performing countries. Only a part of the lagging health expectancy of Eastern Europe can potentially be solved by improvements in lifestyle involving smoking and BMI. Before it is definitely concluded that lifestyle policy for alcohol use is of relatively little importance compared to smoking or BMI, as our findings suggest, better data should be gathered in all European countries concerning alcohol use and the odds ratios of overconsumption of alcohol.

Estimating disease prevalence from drug utilization data using the Random Forest algorithm.

BACKGROUND: Aggregated claims data on medication are often used as a proxy for the prevalence of diseases, especially chronic diseases. However, linkage between medication and diagnosis tend to be theory based and not very precise. Modelling disease probability at an individual level using individual level data may yield more accurate results. METHODS: Individual probabilities of having a certain chronic disease were estimated using the Random Forest (RF) algorithm. A training set was created from a general practitioners database of 276 723 cases that included diagnosis and claims data on medication. Model performance for 29 chronic diseases was evaluated using Receiver-Operator Curves, by measuring the Area Under the Curve (AUC). RESULTS: The diseases for which model performance was best were Parkinson's disease (AUC = .89, 95% CI = .77-1.00), diabetes (AUC = .87, 95% CI = .85-.90), osteoporosis (AUC = .87, 95% CI = .81-.92) and heart failure (AUC = .81, 95% CI = .74-.88). Five other diseases had an AUC >.75: asthma, chronic enteritis, COPD, epilepsy and HIV/AIDS. For 16 of 17 diseases tested, the medication categories used in theory-based algorithms were also identified by our method, however the RF models included a broader range of medications as important predictors. CONCLUSION: Data on medication use can be a useful predictor when estimating the prevalence of several chronic diseases. To improve the estimates, for a broader range of chronic diseases, research should use better training data, include more details concerning dosages and duration of prescriptions, and add related predictors like hospitalizations.