Associations Between Meal Planning and Food Ordering Features of a Digital Precision Nutrition Platform and Improvements in Diet Quality

Abstract Objectives Foodsmart is a digital precision nutrition platform that seeks to address barriers to healthy eating with features including (but not limited to) personalized dietary recommendations based on food frequency questionnaire (FFQ) responses, a meal planner, grocery list, and food ordering capabilities. We investigated the association between engagement with these features and changes in diet quality over time. Methods We included 9939 obese adults ages 18–80 living in the U.S. who enrolled in Foodsmart. Usual dietary intake, age, weight, height, and sex were reported through a 53-item FFQ, modified from the National Cancer Institue's Diet History Questionnaire. A healthy diet score was calculated, which contains 7 components: fruits, vegetables, protein ratio (white meat/vegetarian protein: red/processed meat), carbohydrate ratio (fiber: carbohydrate), fat ratio (polyunsaturated: saturated/trans fats), sodium, and hydration (% of daily fluid goal). Change in diet quality was calculated by subtracting users’ first score from their last score. Multivariate logistic regression models were used to estimate odds ratios (OR) and 95% confidence intervals (CI) of improvement in diet quality per engagement with each feature, adjusted for gender, age, baseline BMI, baseline diet score, and enrollment time. Results Of the 9,939 adults included in the analysis, 57% were between 40 and 59 years old and 21% were male. In multivariate regression models, each additional engagement with personalized dietary recommendations was associated with a 27% higher likelihood of improving diet (OR: 1.27, 95% CI: 1.16–1.39, P < 0.001). Each additional engagement with the meal planner, grocery list, and food ordering was associated, respectively, with a 12% (OR: 1.12, 95% CI: 1.02–1.22, >P = 0.01), 15% (OR: 1.15, 95% CI: 1.05–1.26, P < 0.001), and 18% (OR: 1.18, 95% CI: 1.00–1.38, P = 0.05) higher likelihood of improving diet quality. Conclusions In our prospective study of obese users of a digital precision nutrition platform, we found that more frequent engagement with features related to meal planning and food ordering was associated with diet quality improvement. Funding Sources This study was funded by Foodsmart.

Download Full-text

Influence of Employment Conditions and Length of Residence on Adherence to Dietary Recommendations in Immigrant Workers in Spain

International Journal of Environmental Research and Public Health ◽

10.3390/ijerph15112488 ◽

2018 ◽

Vol 15 (11) ◽

pp. 2488 ◽

Cited By ~ 3

Author(s):

Ikram Benazizi ◽

Elena Ronda-Pérez ◽

Rocío Ortíz-Moncada ◽

José Martínez-Martínez

Keyword(s):

Logistic Regression ◽

Food Consumption ◽

Dairy Products ◽

Regression Models ◽

Immigrant Workers ◽

Sociodemographic Characteristics ◽

Dietary Recommendations ◽

Logistic Regression Models ◽

Employment Conditions ◽

Sugary Drinks

The objective of this article is to analyze the influence of employment conditions on adherence to dietary recommendations among those born in Spain and immigrants by their time of residence. Data were used from the Platform of Longitudinal Studies of Immigrant Families (PELFI) cohort (n = 215) to compare Spaniards and immigrants with <14 and >14 years of residence. The questionnaire on frequency of food consumption (15 items) was used to measure adherence to dietary recommendations. Logistic regression models were used, adjusting for sociodemographic characteristics and employment conditions. Adherence to dietary recommendations was greater among Spaniards, followed by immigrants with >14 years of residence and <14 years of residence. The greatest adherence among Spaniards was for eggs (immigrants ≥ 14 years: 1/ORa = 2.89, <14 years: 1/ORa = 3.92), fish (immigrants ≥ 14 immigrants: 1/ORa = 2.33, <14 years: 1/ORa = 4.72), vegetables (immigrants ≥ 14 years: 1/ORa = 3.26, <14 years: 1/ORa = 4.87), dairy products (immigrants ≥ 14 years: 1/ORa = 14.34, <14 years: 1/ORa = 26.78), and sugary drinks (immigrants ≥14 years: 1/ORa = 2.12, <14 years: 1/ORa = 3.48), and the lowest adherence was for the consumption of sausages and cold cuts (immigrants ≥ 14 years: Ora = 7.62, <14 years: ORa = 24.65). Adjusting for sociodemographic and employment conditions variables did not result in variation in the observed differences between Spaniards, immigrants with <14 years of residence, and immigrants with >14 years of residence.

Download Full-text

Diet Quality Measured by Healthy Eating Index-2015 is Inversely Associated with Depression: NHANES 2011–2014

Current Developments in Nutrition ◽

10.1093/cdn/nzaa046_068 ◽

2020 ◽

Vol 4 (Supplement_2) ◽

pp. 568-568 ◽

Cited By ~ 1

Author(s):

St John Katherine ◽

David Gee

Keyword(s):

Logistic Regression ◽

Depressive Symptoms ◽

Healthy Eating ◽

Diet Quality ◽

Total Protein ◽

Regression Models ◽

Healthy Eating Index ◽

Added Sugar ◽

Logistic Regression Models ◽

Sugar Intake

Abstract Objectives The purpose of this study was to assess the relationship between diet quality and self-reported depressive symptoms in a nationally representative sample of the U.S. population. Methods A total of 8448 adults over the age of 20 were included in this study. Dietary data were based on 24-hour diet recalls and diet quality was measured using the USDA Healthy Eating Index (HEI)-2015. Symptoms of depression were assessed by trained interviewers using the Patient Health Questionnaire-9 (PHQ-9). Logistic regression models determined whether HEI-2015 scores differed between depressed and non-depressed adults after controlling for age, gender, poverty income ratio, and body mass index. Additional logistic regression models examined the association of HEI-2015 diet components and risk of depression. Results The overall prevalence of depression in this sample was 9.3%. Rates of depression were higher among women, middle-aged adults, obese subjects, and those of lower socioeconomic status. Compared to non-depressed adults (n = 7665), depressed adults (n = 783) had a lower total HEI-2015 score. HEI-2015 score was significantly (P < 0.001) and inversely associated with depressive symptoms, both before and after adjusting for covariates. Greater intake of whole fruit, total protein, and lower added sugar intake was associated with a reduced risk of depression. Conclusions Dietitians should be aware that depressed patients may be consuming less-optimal diets than the non-depressed population and should assess diet quality. Lower intakes of whole fruit, total protein, and higher added sugar intake in particular may be prevalent in those with depressive symptoms. All health professionals should encourage a better balanced diet, and a diet pattern that adheres to the 2015–2020 DGA should be promoted for overall health. Funding Sources Central Washington University.

Download Full-text

Consumption of Ultra-Processed Foods and Diet Quality Among U.S. Adults and Children

Current Developments in Nutrition ◽

10.1093/cdn/nzaa046_043 ◽

2020 ◽

Vol 4 (Supplement_2) ◽

pp. 543-543

Author(s):

Junxiu Liu ◽

Euridice Steele ◽

Dimitra Karageorgou ◽

Renata Micha ◽

Carlos Monteiro ◽

...

Keyword(s):

Diet Quality ◽

The United States ◽

Healthy Eating Index ◽

Processed Meat ◽

Processed Foods ◽

The Third ◽

Poor Diet ◽

Adults And Children ◽

Diet Score ◽

Score Range

Abstract Objectives Consumption of ultra-processed foods (UPF) has been linked to higher added sugar, sodium, and unhealthful fats, but the association of UPF with overall diet quality is unclear. We investigated how UPF consumption relates to overall diet quality in the U.S. Methods Dietary data were derived from the two most recent cycles of NHANES (2013–2016) including 10,064 adults and 5920 children with one or more 24-hr dietary recalls. UPFs were identified by the degree of processing using the NOVA classification, with total UPF intake (% energy) assessed in quartiles. Diet quality was assessed using the validated American Heart Association (AHA) 2020 continuous primary diet score (range 0–50; based on total fruits/vegetables, whole grains, fish/shellfish, sugar-sweetened beverages, and sodium), AHA secondary diet score (range 0–80; adding nuts/seeds/legumes, processed meat, and saturated fat), and Healthy Eating Index (HEI)-2015 score (range 0–100). Poor diet was defined as < 40% adherence to the AHA primary score. Linear and logistic regressions estimated the relationships between UPF intake and diet quality, accounting for NHANES complex survey design and adjusting for age, sex, race/ethnicity, income, and education. Results Compared to the lowest quartile of UPF consumption (<42.3%E), the multivariable-adjusted AHA primary score was −7.60% (−1.82 points, P < 0.001) lower in the second quartile, −17.9% (−4.31 points, P < 0.001) lower in the third quartile, and −29.3% (−7.04 points, P < 0.001) lower in the highest quartile (>67.8%E). Corresponding values in children were −8.01% (−1.73, P < 0.001), −15.2% (−3.28, P < 0.001), and −29.8% (−6.43, P < 0.001). Compared to the lowest quartile of UPF, adults in the second quartile had 63% higher risk of poor diet quality (multivariable-adjusted OR 1.63; 95% CI 1.35–1.98); in the third quartile, 2.5-fold higher risk (OR, 2.64; 2.13–3.26); and in the highest quartile, nearly 5-fold higher risk (OR 4.85; 3.93–5.97). Corresponding ORs for poor diet quality in children were 1.52 (1.22–1.89), 2.44 (1.94–3.06), and 5.37 (4.16–6.94) across quartiles of UPF, respectively. Findings were similar using the AHA secondary score and HEI-2015 score. Conclusions Higher consumption of UPF associates with substantially lower diet quality in the United States among both adults and children. Funding Sources NIH.

Download Full-text

Abstract P516: A Traditional Mexican Diet Score is Not Associated With Increased Risk of Hypertension and is Comparable to Other Diet Scores Among U.S. Adults of Mexican Heritage in the Hispanic Community Health Study/Study of Latinos Cohort

Circulation ◽

10.1161/circ.141.suppl_1.p516 ◽

2020 ◽

Vol 141 (Suppl_1) ◽

Author(s):

Martha Tamez ◽

Eric Rimm ◽

Bernard Rosner ◽

Daniela Sotres-Alvarez ◽

Martha L Daviglus ◽

...

Keyword(s):

Blood Pressure ◽

Diet Quality ◽

Regression Models ◽

Health Study ◽

Hypertension Risk ◽

Hispanic Community ◽

Mexican Heritage ◽

Unit Increase ◽

Diet Score

Introduction: Little is known about whether prevailing diet quality scores apply to Hispanics/Latinos when assessing hypertension risk or if a diet score applied to a traditional diet would be more relevant. Objective: To compare a traditional Mexican diet score (tMexS) against three prevailing diet quality scores (Mediterranean Diet Score [MeDS], Alternate Healthy Eating Index-2010 [AHEI-2010], and Dietary Approaches to Stop Hypertension [DASH]) in association with risk of hypertension and change in systolic and diastolic blood pressure (BP) in adults of Mexican heritage in the U.S. Methods: Data were analyzed from 3,542 U.S. adults of Mexican heritage in the multicenter population-based Hispanic Community Health Study/Study of Latinos (HCHS/SOL) cohort with baseline (2008-2011) and follow-up (2014-2017) visits, and without CVD and hypertension at baseline. Diet quality scores were calculated from the average of two 24-h recalls. Higher scores of MeDS, AHEI-2010, and DASH indicate a healthier diet, while a higher tMexS indicates a more traditional Mexican diet. Hypertension was defined as an average of ≥140 systolic or ≥90 mmHg diastolic BP from triplicate measurements, or self-reported use of antihypertensive medications. Logistic regression models tested the association of four dietary scores with incident hypertension, and linear regression models tested the association of dietary scores and changes in systolic and diastolic BP by hypertension status. Multivariable models with complex survey procedures were adjusted for confounders, and the final model included all dietary scores simultaneously. Results: After an average 6y of follow-up, there were 553 incident cases of hypertension. Mean (±SE) total score was 5.8 ± 0.05 for tMexS (range: 0-12), 5.3 ± 0.04 for MeDS (range: 0-9), 51.4 ± 0.19 for AHEI-2010 (range: 0-110), and 48.5 ± 0.22 for DASH (range: 0-90). The tMexS, MeDS, and DASH were not associated with risk of hypertension. A 10-unit increase in the AHEI-2010 was associated with a lower risk of hypertension (OR: 0.77; 95% CI: 0.60, 0.98). Among normotensive adults, diastolic BP decreased by 0.24 ± 0.11 mmHg per one-unit increase in tMexS (p-value=0.02). No significant changes were noted for systolic BP nor among adults with hypertension. Conclusions: In this prospective cohort, the AHEI-2010, but not DASH, MeDS, or tMexS, was associated with lower hypertension risk among U.S. adults of Mexican heritage, while the tMexS reduced diastolic BP among normotensive adults. To decrease hypertension risk, adults of Mexican heritage should adhere to AHEI-2010 dietary recommendations, while consuming traditional Mexican foods for modest benefits on blood pressure.

Download Full-text

Associations between motives for dish choice during home-meal preparation and diet quality in French adults: findings from the NutriNet-Santé study

British Journal Of Nutrition ◽

10.1017/s0007114517000666 ◽

2017 ◽

Vol 117 (6) ◽

pp. 851-861

Author(s):

Pauline Ducrot ◽

Caroline Méjean ◽

Philippine Fassier ◽

Benjamin Allès ◽

Serge Hercberg ◽

...

Keyword(s):

Diet Quality ◽

Regression Models ◽

Fruits And Vegetables ◽

Food Group ◽

Causal Link ◽

Healthy Diet ◽

Logistic Regression Models ◽

Meal Preparation ◽

Nutritional Guidelines ◽

The Difference

AbstractA number of motives such as constraints or pleasure have been suggested to influence dish choices during home-meal preparation. However, no study has evaluated how the importance conferred to these motives potentially influence diet quality. The present study aims at investigating the difference in diet quality according to the importance attached by individuals to various dish choice motives. The importance of twenty-seven criteria related to dish choices on weekdays was evaluated among 48 010 French adults from the NutriNet-Santé study. ANCOVA and logistic regression models, adjusted for sociodemographic and lifestyle factors, were used to evaluate the association between the importance attached to dish choice motives (yes v. no) and energy and food group intakes, as well as adherence to French nutritional guidelines (modified Programme National Nutrition Santé-Guideline Score (mPNNS-GS)). A higher adherence to nutritional guidelines was observed in individuals attaching importance to a healthy diet (mPNNS-GS score 7·87 (sd 0·09) v. 7·39 (sd 0·09)) and specific diets (mPNNS-GS score 7·73 (sd 0·09) v. 7·53 (sd 0·09)), compared with those who attached little/no importance (all P<0·0001). These individuals also exhibited higher intakes of fruits and vegetables, but a lower consumption of meat, milk and cheese, sugary products and convenience foods compared with their respective counterparts (all P<0·0001). For other motives, that is, constraints, pleasure and organisation, only small differences were observed. The main difference in diet quality was related to the importance placed on a healthy diet. Although a causal link should be demonstrated, our findings suggested that strategies aiming at enabling people to take into account diet quality during home-meal preparation might be effective levers to promote healthy eating.

Download Full-text

Positive Attitudes and Reasons for Consuming Legumes Are Associated with Legume Intake Among Adults in Puerto Rico

Current Developments in Nutrition ◽

10.1093/cdn/nzaa043_052 ◽

2020 ◽

Vol 4 (Supplement_2) ◽

pp. 201-201

Author(s):

Elena Hemler ◽

Martha Tamez ◽

Jose Rodriguez Orengo ◽

Josiemer Mattei

Keyword(s):

Puerto Rico ◽

Puerto Rican ◽

Urban Areas ◽

Regression Models ◽

Linear Regression Models ◽

Logistic Regression Models ◽

Funding Sources ◽

Positive Attitudes ◽

Adult Participants ◽

Adjusted Logistic Regression

Abstract Objectives Legumes are a healthy staple of the traditional Puerto Rican diet. Emphasizing legume intake could be a culturally acceptable strategy for chronic disease prevention in Puerto Rico. This study aimed to investigate attitudes, reasons, intentions, and expectations about legume consumption in Puerto Rico and associations with dietary intake. Methods This study includes adult participants (30–75y) from the Puerto Rico Assessment of Diet, Lifestyle and Diseases, who completed a validated food frequency questionnaire and a legumes questionnaire (n = 206). Attitudes and reasons were assessed with a Likert scale and assigned 0 points for disagree, 1 point for neutral, and 2 points for agree. Scores were computed by summing the points for 12 attitudes or 12 reasons; higher scores indicate more positive attitudes and stronger reasons for consuming legumes. Adjusted linear regression models examined associations between each score and bean intake, rice intake, and ratio of bean to rice intake. Adjusted logistic regression models tested the associations between specific intentions and expectations about legume consumption and bean intake. Results Most participants were female (70%), Puerto Rican (80%), and lived in urban areas (90%). Participants reported consuming a median of 0.47 servings/d of beans and 0.57 servings/d of rice. After adjusting for sociodemographic and lifestyle factors, each additional positive attitude was associated with a 0.08 (95% CI: 0.04, 0.1) servings/d higher bean intake and 0.66 (95% CI: 0.04, 1.28) higher ratio of bean to rice intake. Each additional reason to consume legumes was associated with a 0.04 (95% CI: 0.004, 0.06) servings/d higher bean intake. No other significant associations were observed for the attitudes or reasons scores, nor for expectations or intentions to consume legumes. Conclusions Having positive attitudes and more reasons for consuming legumes are associated with higher bean intake. Positive attitudes are also associated with a higher ratio of bean to rice intake. Interventions aiming to increase legume intake to prevent cardiometabolic disease in Puerto Rico should emphasize building positive attitudes and reasons for consuming legumes. Funding Sources Anonymous donations; Northarvest Bean Growers Association; FDI Clinical Research; and National Institutes of Health.

Download Full-text

Discrepancy Between Perceived Diet Quality and Actual Diet Quality Among US Adult Cancer Survivors (OR14-05-19)

Current Developments in Nutrition ◽

10.1093/cdn/nzz030.or14-05-19 ◽

2019 ◽

Vol 3 (Supplement_1) ◽

Author(s):

Hong Xue ◽

Jin Liu

Keyword(s):

Logistic Regression ◽

Cancer Survivors ◽

Diet Quality ◽

Regression Models ◽

Vegetable Intake ◽

Multinomial Logistic Regression ◽

Calorie Intake ◽

Fruit Intake ◽

Logistic Regression Models

Abstract Objectives Postdiagnosis diet and food choices are essential for survivorship management. However, diet quality of cancer survivors is generally low. This study aimed to examine he temporal trends of diet quality of cancer survivors, assess the discrepancy between perceived diet quality and actual diet quality among cancer survivors, and its impact on their actual diet quality. Methods The National Health and Nutrition Examination Survey (NHANES) data 2005 -2014 were used in this study. Logistic regression models were fit to examine the influence of misperception of eating healthiness on diet quality measured by healthy eating index (HEI). Multinomial logistic regression models were fit to examine the factors that may influence the likelihood of diet quality misperception. Results Between 2005 and 2014, there was an increasing trend in HEI in cancer survivors, with a small but significant increase of 0.7 per year. Non-Hispanic white survivors had a consistent higher total HEI score compared to other race/ethnic groups during this period. There were 11.66%, 29.67% and 38.02% cancer survivors who rated their overall diet quality as excellent, very good, or good, while 16.96% and 3.69% perceived their overall diet as fair and poor respectively. Kappa statistic indicated a low agreement between the self-perceived diet quality and the actual diet quality measured by HEI among cancer survivors (0.06, 95% CI: 0.02, 0.09). with adjustment of age, gender, race/ethnicity, and SES status, over-rate misperception was associated with a 5.39 lower total HEI score P < 0.0001), 1.00 lower HEI score of empty calorie intake (P = 0.0028), 0.15 lower score of vegetable intake (P = 0.108) and 0.29 lower score of fruit intake. On the other hand, under-rate misperception was associated with a 7.12 higher total HEI score P < 0.0001), 2.57 higher HEI score of empty calorie intake (P < 0.0001), 0.02 higher score of vegetable intake (P = 0.904) and 0.84 higher score of fruit intake (P = 0.001). Multinomial logistic regression suggested that higher income was also associated with higher odds of being an over-rater. Individuals with college education or above were more likely to over-rate their diet quality compared to those with high school or under education (OR: 1.32, 95% CI: 1.005 1.732). Moreover, Hispanics were more likely than Non-Hispanic Whites to over-rate their diet quality ( OR: 1.792, 95% CI: 1.062, 3.024). Conclusions The divergence between self-assessed eating health and the HEI measured diet quality was an important factor that may have influenced cancer survivors’ diet behavior and diet quality. Funding Sources NA.

Download Full-text

The Association Between Peanut and Peanut Butter Consumption and Cognitive Function

Current Developments in Nutrition ◽

10.1093/cdn/nzab033_026 ◽

2021 ◽

Vol 5 (Supplement_2) ◽

pp. 26-26

Author(s):

Elizabeth Katzman ◽

Samara Nielsen

Keyword(s):

Cognitive Function ◽

Regression Models ◽

Peanut Butter ◽

Digit Symbol Substitution Test ◽

Logistic Regression Models ◽

Funding Sources ◽

Gender Education ◽

Symbol Substitution ◽

Adjusted Logistic Regression ◽

Fluency Test

Abstract Objectives This study investigated the association of peanut and peanut butter (P/PB) consumption and cognitive function. Methods Using 2011–2014 NHANES data, we selected 60–80 year-olds who had two 24-hour diet recalls, cognitive function tests, and education information. P/PB and tree nut (TN) consumption was measured as well as the participant's performance on the CERAD Word Learning subtest (CERAD W-L), Animal Fluency test (AFT), and the Digit Symbol Substitution test (DSST). Scores from the three cognitive tests were dichotomized. Individuals were classified as either P/PB consumers or non-consumers and TN consumers or non-consumers. Unadjusted and adjusted logistic regression models examined associations between P/PB consumption, TN consumption, age, gender, education and cognitive function. Results Among the 2,857 individuals examined, participants who did not consume P/PB were more likely to do poorly on the CERAD W-L (adjusted OR = 1.55, 95% CI 1.25–1.92), AFT (adjusted OR = 1.30, 95% CI 1.06–1.60), and DSST (adjusted OR = 1.37, 95% CI 1.10–1.71) when compared to those who did consume P/PB. Conclusions These findings suggest an association between P/PB consumption and cognitive function. Funding Sources The Peanut Institute provided funding.

Download Full-text

A Traditional Costa Rican Adolescents' Diet Score Is a Valid Tool to Capture Diet Quality and Identify Sociodemographic Groups With Suboptimal Diet

Frontiers in Public Health ◽

10.3389/fpubh.2021.708956 ◽

2021 ◽

Vol 9 ◽

Author(s):

Rafael Monge-Rojas ◽

June O'Neill ◽

Michelle Lee-Bravatti ◽

Josiemer Mattei

Keyword(s):

Socioeconomic Status ◽

Costa Rica ◽

Diet Quality ◽

Rural Areas ◽

Urban Areas ◽

Fast Food ◽

Processed Meat ◽

Low Socioeconomic ◽

Traditional Diet ◽

Diet Score

Traditional diet indices may capture diet quality according to local food culture. Higher adherence to traditional diet scores may help prevent disease, yet evidence in adolescents is limited. This cross-sectional study aimed to develop and validate a Traditional Costa Rica Adolescents Diet Score (TCRAD) and determine its association with sociodemographic characteristics, under the hypothesis that girls, adolescents from rural areas, and with low socioeconomic status, have a more traditional healthy diet. A total of 804 urban and rural adolescents (13–18 years old) participated in the study. The TCRAD showed adequate internal validity as shown by significant associations with intake of 14 traditional foods and nutrients (legumes, vegetables, fruits, oils, dairy, and corn tortilla scored as healthy; and white rice, red/processed meat, solid fats, desserts/pastries, sugar-sweetened beverages, snacks, fast food, and bread and cookies scored as unhealthy). A high TCRAD score, indicative of a healthier and more traditional diet, was observed among adolescents in the low socioeconomic group vs. medium or high socioeconomic categories (42.9, 41.2, and 38.2%, respectively, p < 0.05), adolescents living in rural areas vs. urban (47.6 vs. 34.2%, p < 0.05), and among boys vs. girls (46.9 vs. 37.5%, p < 0.05). The TCRAD score is a valid tool to capture diet quality of adolescents in Costa Rica and could be used to measure association of diet with disease outcomes in this and similar populations. Public health nutrition programs in Costa Rica should focus on improving intake of foods and nutrients, and prioritize girls, adolescents in urban areas, and adolescents with high socioeconomic status.

Download Full-text

Predictors of Urinary Pyrethroid and Organophosphate Compound Concentrations among Healthy Pregnant Women in New York

International Journal of Environmental Research and Public Health ◽

10.3390/ijerph17176164 ◽

2020 ◽

Vol 17 (17) ◽

pp. 6164

Author(s):

Arin A. Balalian ◽

Xinhua Liu ◽

Eva Laura Siegel ◽

Julie Beth Herbstman ◽

Virginia Rauh ◽

...

Keyword(s):

New York ◽

Pregnant Women ◽

Regression Models ◽

Infant Development ◽

Limit Of Detection ◽

Processed Meat ◽

Linear Regression Models ◽

Logistic Regression Models ◽

Thyroid Disruption ◽

Modeling Strategy

Our study aimed to investigate dietary and non-dietary predictors of exposure to pyrethroids, organophosphates pesticides and 2,4-D herbicide in two cohorts of pregnant women in New York City: 153 women from the Thyroid Disruption and Infant Development (TDID) cohort and 121 from the Sibling/Hermanos Cohort(S/H). Baseline data on predictors were collected from the women at time of recruitment. We used three different modeling strategies to address missing data due to biomarker values below the limit of detection (<LOD): (1) logistic regression models with biomarkers categorized as (<median, ≥median); (2) linear regression models, imputing the <LOD values with (LOD/√2); (3) regression models, considering <LOD values as left-censored. Generally, all three models identified similar predictors of exposure. We found that ethnicity, higher income and education predicted higher concentrations of most of the biomarkers in both cohorts. Mothers who consumed processed meat in the TDID cohort, and broiled, barbequed food or burgers in the S/H cohort, tended to have lower concentrations of organophosphates and 2,4-D. The choice of modeling led to a few different predictors identified, and the selection of modeling strategy should be based on the study question.

Download Full-text