- Email: [email protected]
Using Secondary 24-Hour Dietary Recall Data to Estimate Daily Dietary Factor Intake From the FLASHE Study Dietary Screener
RESEARCH METHODS
Using Secondary 24-Hour Dietary Recall Data to Estimate Daily Dietary Factor Intake From the FLASHE Study Dietary Screener Teresa M. Smith, PhD,1 Eric E. Calloway, PhD,1 Courtney A. Pinard, PhD,1 Erin Hennessy, PhD, MPH,2 April Y. Oh, PhD, MPH,3 Linda C. Nebeling, PhD, MPH, RD,3 Amy L. Yaroch, PhD1 Introduction: The National Cancer Institute’s 2014 Family Life, Activity, Sun, Health, and Eating Study utilized a 27-item Dietary Screener tailored to adolescent eating patterns that assessed the frequency of intake of several foods and beverages in parent–adolescent dyads. This study estimated intake of fruits and vegetables (FVs), dairy, added sugars, and whole grains for screener respondents using existing, nationally representative, 24-hour dietary recall data.
Methods: Dietary Screener items were converted from frequency responses to daily intake. Intake (dependent variable) was estimated using regression coefficients and portion sizes of foods and beverages (independent variables) generated from the 2003–2006 National Health and Nutrition Examination Survey 2-day 24-hour recall data set. Means (SDs) were used to examine daily dietary factor intake among parent and adolescents. Analysis was conducted in 2015–2016. The analytic sample consisted of 1,732 parents (aged Z18 years) and their adolescent aged 12–17 years (n¼1,632).
Results: Male parents consumed 3.6 cups of FVs, 1.8 cups of dairy, 22.6 teaspoons of added sugars, and 2.1 ounces of whole grains daily; female parents consumed 2.8 cups of FVs, 1.3 cups of dairy, 14.8 teaspoons of added sugars, and 1.4 ounces of whole grains daily. Male adolescents consumed 2.2 cups of FVs, 1.9 cups of dairy, 17.9 teaspoons of added sugars, and 1.0 ounces of whole grains daily; female adolescents consumed 2.2 cups FVs, 1.6 cups of dairy, 14.2 teaspoons of added sugars, and 0.8 ounces of whole grains daily.
Conclusions: Utilizing a dietary screener tailored to adolescent eating patterns in parent– adolescent dyads provided estimated dietary factor intake, underscoring existing 24-hour dietary recall data can be used to calibrate dietary habits. Am J Prev Med 2017;52(6):856–862. & 2017 American Journal of Preventive Medicine. Published by Elsevier Inc. All rights reserved.
INTRODUCTION
M
odifiable behaviors (e.g., poor diet and physical inactivity) that contribute to obesity and increase risk for chronic diseases, such as cancer, continue to afflict much of the population in the U.S. The National Cancer Institute (NCI),1 as well as other organizations such as the World Cancer Research Fund and American Institute for Cancer Research,2 American Cancer Society,3 and WHO,4 have emphasized diet, and especially specific food groups, as having potential to increase risk or prevent some cancers. Fruits and vegetables (FVs) and whole grains, for example, are often cited as potentially preventative of commonly 856 Am J Prev Med 2017;52(6):856–862
occurring cancers in the U.S. (e.g., breast, colorectal, and prostate), whereas foods high in added sugars and fats may increase risk for some cancers.1–4 Studies From the 1Gretchen Swanson Center for Nutrition, Omaha, Nebraska; 2 Tufts University Friedman School of Nutrition Science and Policy, Boston, Massachusetts; and 3Behavioral Research Program, National Cancer Institute, Bethesda, Maryland Address correspondence to: Erin M. Hennessy, PhD, MPH, Tufts University Friedman School of Nutrition Science and Policy, 150 Harrison Avenue, Boston MA 02111. E-mail: [email protected]. This article is part of a theme section titled The Family Life, Activity, Sun, Health, and Eating (FLASHE) Study: Insights Into Cancer-Prevention Behaviors Among Parent–Adolescent Dyads. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2017.01.015
& 2017 American Journal of Preventive Medicine. Published by Elsevier Inc. All rights reserved.
Smith et al / Am J Prev Med 2017;52(6):856–862
investigating individual food groups (e.g., FVs) and nutrients (e.g., fiber and sodium) have generated evidence supporting the relationship between dietary patterns with cancer incidence.5,6 Understanding correlates of dietary intake of these key food groups may help inform environmental-, familial-, and individual-based interventions focused on increasing healthy eating and cancer prevention. In particular, though it has been demonstrated that children (aged 6–12 years) and their parents have closely related diet quality and energy intake, perhaps because of shared food environments and parental modeling,7 this relationship and the mechanisms driving this relationship are unknown among adolescents and their parents. The Family Life, Activity, Sun, Health, and Eating (FLASHE) Study is a publicly available database developed by NCI that surveyed the psychosocial; generational (adolescents aged 12–17 years and their parent); and environmental correlates of cancer preventive behaviors.8 The survey was administered in 2014 with the goal of advancing the understanding of the relationships among the environment, psychosocial, and behavioral factors from a dyadic perspective. FLASHE included a 27-item Dietary Screener that asked about frequency of foods eaten and beverages consumed in the past 7 days (http://cancercontrol.cancer.gov/brp/hbrb/flashe.html). The FLASHE Dietary Screener was based on the Dietary Screener Questionnaire (DSQ) used in the 2009–2010 National Health and Nutrition Examination Survey (NHANES) but was tailored by NCI to inquire further on the foods and beverages commonly consumed among adolescents in the U.S., which were determined through literature review and by content experts, and reviewed by an external scientific advisory committee.8 The FLASHE Dietary Screener included common items (e.g., green salad, non-fried vegetables, cooked beans, fruit, cooked whole grains, whole grain bread, non-sugary cereal, fried chicken, sugary cereal, candy and chocolate, fried potatoes, other non-fried potatoes) as well as novel items (e.g., chips, processed meat, burgers, tacos, heat and serve, and others) to capture patterns of dietary behavior among adolescents and their families. Intake of screener items not described in this manuscript (e.g., novel items) will be described using frequencies of reported intake in a future manuscript. Assessing and characterizing dietary intake to identify areas for improvement in diet is an essential first step in promoting a more healthful diet (e.g., which may include greater FV intake) and ultimately reducing risk for certain cancers.9 NHANES typically collects diet information through 24-hour recalls; however, the 2009–2010 survey included the DSQ. In recognition of a need for characterizing a population’s median intakes or June 2017
857
examining inter-relationships between diet and other variables, the Risk Factor Assessment Branch (RFAB) of NCI developed scoring algorithms using 2003–2006 NHANES 24-hour recall data to convert screener responses into estimates of intake for several dietary factors that have remained of interest in dietary guidance in America.10–12 The dietary factors included cup equivalents of FVs and dairy, teaspoon equivalents of added sugars, ounce equivalents of whole grains, grams of fiber, and milligrams of calcium.10 The authors worked with RFAB to convert FLASHE Dietary Screener responses into intake of dietary factors believed to be of interest to researchers. RFAB tailored scoring algorithms developed for the DSQ based on the 2003–2006 NHANES 24-hour recall data.13 These derived variables are included in the publically available data set for researchers to explore dietary factor intake of the FLASHE sample. The purpose of this study was to estimate key dietary factor intake for this sample of adolescents and parents, as well as the relationship of adolescent intake to their parent’s intake.
METHODS Study Sample This study analyzed secondary data from the 2014 FLASHE study, a cross-sectional web survey administered from April to October 2014. The FLASHE study was approved by the U.S. Government’s Office of Management and Budget, National Cancer Institute’s Special Studies IRB, and Westat’s IRB. FLASHE methodology and sample design are described in detail elsewhere.14 In brief, the survey sample came from a commercial online consumer opinion panel; the response rate for parent surveys was 34.0% and for adolescent surveys was 31.6%.14 Sampled parents and their adolescent children completed self-reported items about their neighborhood and home environment, psychosocial factors, and behaviors related to cancer risk in two surveys, a diet-focused and physical activity–focused survey. The sample of participants that completed the diet-focused survey consisted of 1,745 adults and 1,657 adolescents. Respondents were excluded if their age or sex were not ascertained (n¼13) and for adolescents, if their age was outside of the range of 12–17 years (n¼25). Final analytic samples included 1,732 adults and 1,632 adolescents.
Measures The FLASHE Dietary Screener is unique in that although it is based on the 26-item DSQ, it is tailored to capture foods and beverages more commonly consumed by adolescents in the U.S. The screener is administered similarly among parents. To estimate daily intake of dietary factors (i.e., food, beverage, or food groups) for the analytic sample, FLASHE Dietary Screener items were assigned to dietary factors (Appendix [available online] lists 18 of the 27 FLASHE Dietary Screener items used to estimate daily intake of dietary factors). Estimated daily intake of dietary factors from FLASHE Dietary Screener items were based upon
858
Smith et al / Am J Prev Med 2017;52(6):856–862
relationships detailed in the DSQ (e.g., fruit, fruit juice, salad, fried potatoes, other potatoes, dried beans, other vegetables, tomato sauce, salsa, and pizza were used to estimate fruits and vegetables cup equivalents per day in the DSQ),13 a review of the literature, expert consultation, and internal consensus. Estimated dietary factor intake was calculated for parents and adolescents using SAS programs obtained from RFAB at NCI. The SAS programs were originally developed to compare the 2009–2010 NHANES dietary screener responses (similar to the FLASHE Dietary Screener) to the What We Eat in America (WWEIA) 24-hour dietary recall data from the 2003–2006 NHANES, and are currently described in detail elsewhere.15 RFAB is currently developing direct calibrations of 2009–2010 NHANES 24-hour recall data to the DSQ, but they are not finalized; the most advanced science available, which included an indirect calibration with WWEIA 24-hour dietary recall data from the 2003–2006 NHANES, was utilized for this study.10 The SAS programs were tailored by RFAB at NCI to the FLASHE Dietary Screener items. A detailed description of this methodology, including how RFAB estimated portion size and regression coefficients from the 2003–2006 NHANES data set, can be found at epi.grants.cancer.gov/nhanes/dietscreen/scoring/. Additional steps, however, were warranted for aligning the FLASHE Dietary Screener items non-sugary cereal, sugary cereal, and non-soda sugar-sweetened beverages (SSBs) to the DSQ screener items before RFAB was able to tailor algorithms to this sample. Frequency of key foods, beverages, and food groups were assessed through the FLASHE 27-item Dietary Screener that asked respondents about frequency of foods eaten (green salad, non-fried vegetables, cooked beans, fruit, cooked whole grains, whole-grain bread, nonsugary cereal, fried chicken, sugary cereal, candy and chocolate, fried potatoes, other non-fried potatoes, chips, processed meat, burgers, tacos, heat and serve, cookies, cake, frozen desserts, pizza) and beverages consumed (water, fruit juice, milk, sweetened fruit drinks, regular soda, sports drinks, energy drinks) in the past 7 days. Respondents selected one of six response options that included did not consume in past 7 days; 1–3 times in past 7 days; 4–6 times in past 7 days; 1 time per day; 2 times per day; or 3 or more times per day. To estimate dietary factor intake, the first three responses (i.e., did not consume, 1–3 times, and 4–6 times in past 7 days) were converted into daily frequency by dividing by 7 days. In cases where the response included a range, the mid-value was selected, and divided by 7 days. There was one exception with regard to pizza, which was top-coded so that responses greater than or equal to two times per day were restricted to two. The top-code value of two was determined to be the maximum plausible intake value for pizza based on previously conducted work that used NHANES analysis of 2-day mean intakes from 24-hour recall data across all sex and age groups.16 Age; sex; household income (parents only); education level (parents only); race/ethnicity; health status; height; and weight were also self-reported in FLASHE.
Statistical Analysis Mean (SD) or percentage values were used to examine daily frequency of screener items, daily dietary factor intake, and sociodemographics among parent and adolescents. SAS, version 9.4, was used for statistical analysis. Daily dietary factor intake was calculated for the following: cup equivalents of fruits, vegetables, FVs combined (with and without fried potatoes); cup equivalents of dairy; teaspoon equivalents of added sugars (for all sugary foods and SSBs only); and ounce equivalents of whole grains based on
FLASHE Dietary Screener items. Additionally, the ratio of the estimated amount consumed by adolescents to the estimated amount consumed by their parent was reported. Analysis was conducted in 2015–2016.
RESULTS As shown in Table 1, parents in this sample were predominantly aged 35–44 years (44%); considered themselves to be non-Hispanic white (69%); reported an annual household income of o$100,000 per year (79%); and had earned a 4year college degree or higher (46%). Most of the parents reported that their health was very good or excellent (57%), and self-reported height and weight converted into BMI was considered to be in a normal or healthy range (BMI Z18.5 and r25.0) for 36% of parents. Roughly half (50%) of the adolescents in this sample were aged 12–14 years (Table 2). Most of the adolescents considered themselves to be non-Hispanic white (63%), followed by non-Hispanic black (17%); Hispanic (10%); and other or multiple races (9%). The majority of adolescents reported their health was very good or excellent (79%). The self-reported height and weight converted into BMI percentile was considered to be in a normal or healthy range (BMI percentile Z5.0 and r85.0) for 66% of adolescents. The means (SDs) of daily frequency intake of key screener items across adolescent respondents and their parents are shown in Table 3. Adolescents reported consuming milk and milk alternatives most frequently (0.99 [SD¼0.88] times per day) and cooked beans least frequently (0.22 [SD¼0.38] times per day). Their parents reported consuming fruit most frequently (0.97 [SD¼0.85] times per day) and sugary cereal least frequently (0.13 [SD¼0.30] times per day). Table 4 shows the means (SDs) of the daily dietary factor intake of key food groups and nutrients among adult and adolescent respondents. Based on the algorithm used to convert frequency of consumption to cup equivalents, parents in the FLASHE sample were consuming 3.6 cup equivalents (men) and 2.8 cup equivalents (women) of FVs; 1.8 cup equivalents (men) and 1.3 cup equivalents (women) of dairy; 22.6 teaspoon equivalents (men) and 14.8 teaspoon equivalents (women) of added sugars; and 2.1 ounce equivalents (men) and 1.4 ounce equivalents (women) of whole grains daily. Adolescents in the FLASHE sample reported consuming 2.2 cup equivalents (male adolescents) and 2.2 cup equivalents (female adolescents) of FVs; 1.9 cup equivalents (male adolescents) and 1.6 cup equivalents (female adolescents) of dairy; 17.9 teaspoon equivalents (male adolescents) and 14.2 teaspoon equivalents (female adolescents) of added sugars; and 1.0 ounce equivalents www.ajpmonline.org
Smith et al / Am J Prev Med 2017;52(6):856–862
859
Table 1. Distribution of Self-Reported Health Characteristics and Sociodemographics Among Parent Respondents (n¼1,732) Stratified by sex Variables n Age, years 18–34 35–44 45–59 Z60 Household income $0–$99,999 Z$100,000 Highest grade or level of education completed Less than a high school degree A high school degree or GED Some college but not a college degree A 4-year college degree or higher Race/Ethnicity Non-Hispanic white Non-Hispanic black Hispanic Other Self-reported health status Poor or fair Good Very good or excellent BMI Underweight (BMI o18.50) Normal weight (BMI Z18.50 and r24.99) Overweight (BMI Z25.00 and r29.99) Obese (BMI Z30.00)
Total, n (%)
Male, n (%)
Female, n (%)
454 (26.2)
1278 (73.8)
196 (11.3) 753 (43.5) 732 (42.3) 51 (2.9)
43 157 224 30
153 596 508 21
1359 (78.5) 353 (20.4)
324 (71.4) 127 (28.0)
1035 (81.0) 226 (17.7)
22 (1.3) 292 (16.9) 607 (35.0) 804 (46.4)
5 (1.1) 72 (15.9) 135 (29.7) 241 (53.1)
17 220 472 563
(1.3) (17.2) (36.9) (44.1)
1199 (69.2) 293 (16.9) 125 (7.2) 99 (5.7)
321 (70.7) 51 (11.2) 48 (10.6) 31 (6.8)
878 242 77 68
(68.7) (18.9) (6.0) (5.3)
204 (11.7) 535 (30.9) 983 (56.8)
47 (10.3) 148 (32.6) 255 (56.2)
157 (12.2) 387 (30.3) 728 (57.0)
23 (1.3) 627 (36.2) 526 (30.4) 535 (30.9)
4 (0.9) 132 (29.1) 185 (40.7) 130 (28.6)
19 495 341 405
(9.5) (34.6) (49.3) (6.6)
(12.0) (46.6) (39.7) (1.6)
(1.5) (38.7) (26.7) (31.7)
Note: Percentages do not add up to 100% because of missing responses. GED, General Educational Development test.
(male adolescents) and 0.8 ounce equivalents (female adolescents) of whole grains daily. In this sample, adolescents reported lower consumption of cup equivalents of vegetables, FVs, and FVs without fried potatoes when compared with their parent (68%, 80%, and 80%, respectively). They reported greater consumption, however, of cup equivalents of fruits (59% more); teaspoons of added sugars (30% more); cup equivalents of dairy (37% more); and ounce equivalents of whole grains (29% more). Of greatest note is that adolescents reported consuming about six times as many teaspoon equivalents of added sugars from SSBs compared with their parent.
DISCUSSION The intent of this study was to estimate key daily dietary factor intake for a sample of parents and adolescents using FLASHE Dietary Screener responses and to illustrate subsequent values among the participating sample. June 2017
Although this sampling design of FLASHE was different from that of NHANES, it is noteworthy that when indirectly compared to 24-hour recall data from WWEIA, 2009–2010 NHANES weighted Mean Daily Food Patterns Cup Equivalents tables,17 parents in the FLASHE sample reported consuming amounts relatively close to adults aged Z20 years in the WWEIA sample. For illustrative purposes, FLASHE male and female parents reported consuming 1.8 cups and 1.4 cups of vegetables a day, respectively, compared with male and female WWEIA participants who consumed 1.7 cups and 1.5 cups, respectively. FLASHE male and female adolescents reported consuming 1.0 cups of vegetables a day, compared with male and female participants aged 12–19 years in the WWEIA sample who consumed 1.2 cups and 1.0 cups, respectively. Though the age ranges of the samples differ, when comparing to the national sample, these data suggest that variables derived in the current study are reasonable estimates of dietary intake for parents and adolescents who participated in FLASHE.
860
Smith et al / Am J Prev Med 2017;52(6):856–862
Table 2. Distribution of Self-Reported Health Characteristics and Sociodemographics Among Adolescent Respondents (N¼1,632) Stratified by Sex Total, n (%)
Male, n (%)
Female, n (%)
809 (49.6)
823 (50.4)
820 (50.2) 812 (49.8)
394 (48.7) 415 (51.3)
426 (51.8) 397 (48.2)
1,030 (63.1) 272 (16.7) 160 (9.8) 152 (9.3)
501 (61.9) 137 (16.9) 83 (10.3) 80 (9.9)
529 (64.3) 135 (16.4) 77 (9.4) 72 (8.7)
91 (5.6) 241 (14.8) 1,295 (79.4)
36 (4.4) 102 (12.6) 669 (82.7)
55 (6.7) 139 (16.9) 626 (76.1)
66 (4.0) 1,085 (66.5) 245 (15.0) 201 (12.3)
42 514 130 109
24 (2.9) 571 (69.4) 115 (14.0) 92 (11.2)
Variables n Age, years 12–14 15–17 Race/ethnicity Non-Hispanic white Non-Hispanic black Hispanic Other Self-reported health status Poor or fair Good Very good or excellent BMI-for-age percentile classification Underweight (BMI percentile o5.00) Normal weight (BMI percentile Z5.00 and r84.99) Overweight (BMI percentile Z85.00 and r94.99) Obese (BMI percentile Z95.00)
(5.2) (63.5) (16.1) (13.5)
Note: Percentages do not add up to 100% because of missing responses.
When considering the dyad-level ratio of amounts of dietary factors consumed by adolescents and their parent, adolescents generally reported consuming lesser amounts of Table 3. Mean ⫾ SD of Daily Frequency Intake of Key Screener Items Across Adolescent Respondents and Their Parent
a
FLASHE screener items
Adolescent
Parent
100% fruit juice Candy/chocolate Cooked beans Cooked whole grains Cookies/cakes Fried potatoes Frozen desserts Fruit Green salad Milk and milk alternatives Non-soda SSBa Non-sugary cereal Other non-fried potatoes Other non-fried vegetables Pizza Soda Sugary cereal Whole grain bread
0.51⫾0.63 0.49⫾0.53 0.22⫾0.38 0.23⫾0.39 0.40⫾0.46 0.33⫾0.36 0.29⫾0.36 0.86⫾0.80 0.38⫾0.50 0.99⫾0.88 0.83⫾1.04 0.24⫾0.38 0.25⫾0.36 0.60⫾0.59 0.28⫾0.28 0.48⫾0.65 0.27⫾0.40 0.45⫾0.54
0.40⫾0.56 0.41⫾0.48 0.25⫾0.36 0.22⫾0.34 0.30⫾0.39 0.26⫾0.32 0.21⫾0.30 0.97⫾0.85 0.50⫾0.51 0.37⫾0.39 0.50⫾0.93 0.22⫾0.34 0.25⫾0.32 0.75⫾0.67 0.22⫾0.24 0.46⫾0.74 0.13⫾0.30 0.43⫾0.49
Composite of sweetened fruit drinks, energy drinks, and sports drinks. FLASHE, Family Life, Activity, Sun, Health, and Eating Study; SSB, sugarsweetened beverages.
vegetables, as well as FVs combined (with and without fried potatoes). However, adolescents reported consuming greater amounts of fruits; added sugars (including those in SSBs); dairy; and whole grains. A recent study found that female adolescents (aged 12.0–18.8 years) tend to mimic their parents’ eating behaviors with regard to the same foods.18 Though most of the ratio values in the current study would support this finding, adolescents reported consuming more than six times as many teaspoons of added sugars from SSBs, which can equate to nearly half a can of regular soda per day. Adolescent SSB intake has been found by a different study to be potentially influenced by their parent(s), but also other people, such as peers or other adult authority figures.19 Future research should seek to understand correlates of dietary behaviors between members of parent–adolescent dyads, especially among the dietary factors identified in the current study, in order to inform family-based interventions.
Limitations This study has several limitations to report. First, because data collected using the FLASHE Survey were self-report, they cannot be independently verified for accuracy. Second, the sample was drawn using a consumer opinion panel and not a random sample, though similar data collected technique have been used in other large health behavior surveys (e.g., NCI’s Food Attitudes and Behaviors Survey and the Centers for Disease Control and Prevention’s Styles surveys). Third, these estimates were www.ajpmonline.org
June 2017
1.30 (1.22, 1.36) 6.04 (5.32, 6.77) 1.37 (1.32, 1.41) 1.29 (1.16, 1.41) 14.81⫾8.60 7.86⫾10.43 1.33⫾0.46 1.43⫾3.33 22.55⫾12.75 14.35⫾15.01 1.75⫾0.68 2.14⫾4.83 16.86⫾10.46 9.58⫾12.18 1.44⫾0.56 1.62⫾3.81 eq, equivalent; SSB, sugar-sweetened beverage; tsp, teaspoon.
17.83⫾7.88 7.74⫾7.50 1.85⫾0.75 1.00⫾1.79 16.00⫾7.66 6.68⫾7.00 1.73⫾0.73 0.89⫾1.57
14.20⫾6.94 5.69⫾6.32 1.61⫾0.70 0.79⫾1.37
(1.42, 1.75) (0.67, 0.70) (0.78, 0.82) (0.78, 0.82) 1.59 0.68 0.80 0.80 1.24⫾0.88 1.41⫾0.52 2.79⫾1.33 2.69⫾1.29 1.64⫾1.09 1.79⫾0.61 3.61⫾1.45 3.44⫾1.39 1.34⫾0.95 1.51⫾0.57 3.00⫾1.41 2.88⫾1.36 1.08⫾0.73 0.96⫾0.36 2.18⫾0.99 2.07⫾0.96
Fruits (cup eq) Vegetables (cup eq) Fruits and vegetables (cup eq) Fruits and vegetables without fried potatoes (cup eq) Added sugars (tsp) Added sugars from SSB (tsp) Dairy (cup eq) Whole grains (ounce eq)
1.06⫾0.72 0.96⫾0.38 2.16⫾1.05 2.05⫾0.99
1.04⫾0.71 0.96⫾0.41 2.15⫾1.10 2.03⫾1.03
Mean (95% CI) Female Male Male Dietary factor
Total
Female
Total
Parent Adolescent
Table 4. Mean ⫾ SD of Daily Dietary Factor Intake of Food Groups and Nutrients Across Adolescent Respondents and Their Parent
Ratio of adolescent to parent
Smith et al / Am J Prev Med 2017;52(6):856–862
861
based on scoring algorithms developed to estimate dietary factor intake using 2003–2006 NHANES 24-hour recall data and eating patterns may differ in portion size and overall contribution to the estimated nutrients, though portion sizes were estimated based on age and sex and applied accordingly to participants of the current study. RFAB is currently developing direct calibrations of 2009–2010 NHANES 24-hour recall data to the DSQ, but as they are not finalized, the most advanced science available was utilized.10 Lastly, it is unknown how accurate responses are for adolescents participating in frequency-based questionnaires. Despite these limitations, this study has many strengths to report, namely, the use of 24-hour recall data to estimate dietary intake from Dietary Screener responses. Dietary screeners, including the DSQ and FLASHE Dietary Screener, ask about intake of food, beverages, or food groups of particular interest, and then the frequency of consumption of screener items or patterns of eating practices can be used as indicators of particular aspects of diet.20 It is important to note that some dietary screeners probe respondents about the portion sizes of food/beverages they consume (i.e., FV screeners in NCI’s Food Attitudes and Behaviors Survey),21,22 which can allow for analysis and reporting on a standard amount consumed (e.g., cups, teaspoons).23 However, there can be methodologic issues, such as inconsistent results, high error rates, and compliance issues, when utilizing participant-reported portion sizes for estimating dietary intake.23 The FLASHE Dietary Screener was a feasible instrument, as it provided a short, relatively simple, and a moderately valid way to rank individuals with regard to consumption of certain foods and nutrients.21,24 In this study, it was important to be able to estimate a standardized amount of food consumed. Future studies using FLASHE data could have a similar need, and researchers may seek to compare the intake of a study sample to a national average or dietary recommendation.
CONCLUSIONS This study advances the literature in that it utilized a dietary screener tailored to adolescent eating patterns to estimate dietary factor intake of adolescents and their parents. Because of this design, data estimated from this screener, as well as the screener itself, is useful in dyadic and familial dietary research. Further, it underscores the utility of advanced dietary assessment and analysis in assessing population-level dietary habits. This screener and the accompanying publically available data set, including the derived variables described in this study, are available for researchers to explore dietary factor intake of parent–adolescent dyads; characteristics and
862
Smith et al / Am J Prev Med 2017;52(6):856–862
dietary intake of subsamples; and relationship of dietary intake to other health outcomes and lifestyles (e.g., BMI and self-reported health status).
ACKNOWLEDGMENTS This article is part of a theme issue supported by the National Institutes of Health. The findings and conclusions in this article are those of the author(s) and do not necessarily represent the official position of the National Institutes of Health. Leslie A. Lytle, Ph.D. and Louise C. Mâsse, Ph.D. served as Guest Editors of the theme issue. Thank you to Frances E. Thompson, MPH, PhD, epidemiologist in the Risk Factor Assessment Branch at the National Cancer Institute for developing the screener and its algorithms and Lisa L. Kahle of Information Management Services for developing the SAS programs used in this analysis. Thank you to Casey Blaser for advising on the statistical analysis. The Family Life, Activity, Sun, Health, and Eating (FLASHE) study was funded by the National Cancer Institute under contract number HHSN261201200039I issued to Westat. The Gretchen Swanson Center for Nutrition (Authors Smith, Calloway, Pinard, and Yaroch) worked on FLASHE via subcontract to Westat under this contract. At the time this study was conducted, Dr. Hennessy was a Senior Behavioral Scientist (Contractor), Clinical Research Directorate/ Clinical Monitoring Research Program, Leidos Biomedical Research, Inc., Frederick National Laboratory for Cancer Research. No financial disclosures were reported by the authors of this paper.
SUPPLEMENTAL MATERIAL Supplemental materials associated with this article can be found in the online version at http://dx.doi.org/10.1016/j. amepre.2017.01.015.
REFERENCES 1. Klein WMP, Bloch M, Hesse BW, et al. Behavioral research in cancer prevention and control: a look to the future. Am J Prev Med. 2014; 46(3):303–311. http://dx.doi.org/10.1016/j.amepre.2013.10.004. 2. World Cancer Research Fund, American Institute for Cancer Research (AICR). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington DC: AICR; 2007. 3. Kushi LH, Doyle C, McCullough M, et al. American Cancer Society guidelines on nutrition and physical activity for cancer prevention: reducing the risk of cancer with healthy food choices and physical activity. CA Cancer J Clin. 2012;62(1):30–67. http://dx.doi.org/10.3322/caac.20140. 4. Joint WHO-FAO Expert Consultation on Diet, Nutrition, and the Prevention of Chronic Diseases. Diet, Nutrition, and the Prevention of Chronic Diseases: Report of a WHO-FAO Expert Consultation. Geneva: WHO; 2003. 5. Key TJ, Schatzkin A, Willett WC, Allen NE, Spencer EA, Travis RC. Diet, nutrition and the prevention of cancer. Public Health Nutr. 2004;7(1A):187–200. http://dx.doi.org/10.1079/phn2003588. 6. Emmett PM, Jones LR, Northstone K. Dietary patterns in the Avon Longitudinal Study of Parents and Children. Nutr Rev. 2015;73(suppl 3): 207–230. http://dx.doi.org/10.1093/nutrit/nuv055. 7. Robson SM, Couch SC, Peugh JL, et al. Parent diet quality and energy intake are related to child diet quality and energy intake. J Acad Nutr Diet. 2016;116 (6):984–990. http://dx.doi.org/10.1016/j.jand.2016.02.011.
8. Nebeling LC, Hennessy E, Oh AY, et al. The FLASHE study: survey development, dyadic perspectives, and participants characteristics. Am J Prev Med. 2017;52(6):839–848. 9. Patrick K, Norman GJ, Calfas KJ, et al. Diet, physical activity, and sedentary behaviors as risk factors for overweight in adolescence. Arch Pediatr Adolesc Med. 2004;158(4):385–390. http://dx.doi.org/10.1001/ archpedi.158.4.385. 10. Division of Cancer Control and Population Sciences, National Cancer Institute. Dietary Screener Questionnaire in the NHANES 2009-10: background. Epidemiology and Genomics Research Program. http:// epi.grants.cancer.gov/nhanes/dietscreen/. Published October 15, 2015. Accessed December 1, 2015. 11. Dietary Guidelines Advisory Committee. Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans, 2010, to the Secretary of Agriculture and the Secretary of Health and Human Services. Agric Res Serv. 2010. 12. U.S. DHHS, U.S. Department of Agriculture. 2015–2020 Dietary Guidelines for Americans. http://health.gov/dietaryguidelines/2015/ guidelines/. Published 2015. Accessed November 16, 2016. 13. National Cancer Institute, Applied Research Cancer Control and Population Sciences. Dietary Screener Questionnaire in the NHANES 2009-10. http://appliedresearch.cancer.gov/nhanes/dietscreen/. Published April 11, 2014. 14. Westat. Family Life, Activity, Sun, Health, and Eating (FLASHE) Study Methodology Report. http://cancercontrol.cancer.gov/brp/hbrb/docs/FLA SHE_Methods_Report.pdf. Published 2015. Accessed October 6, 2016. 15. Division of Cancer Control and Population Sciences, National Cancer Institute. Dietary Screener Questionnaire (DSQ) in the NHANES 200910: data processing and scoring procedures. Epidemiology and Genomics Research Program. http://epi.grants.cancer.gov/nhanes/dietscreen/scor ing/. Published October 15, 2015. Accessed December 29, 2015. 16. Epidemiology and Genomics Research Program. Identifying extreme exposure values (earlier methods). https://epi.grants.cancer.gov/ nhanes/dietscreen/scoring/earlier/identify.html. Published October 15, 2015. Accessed December 20, 2016. 17. U.S. Department of Agriculture, Agricultural Research Service. Food patterns equivalents intakes from food: mean amounts consumed per individual, by gender and age. What We Eat in America. NHANES 2009-2010. www.ars.usda.gov/ba/bhnrc/fsrg. Published 2013. Accessed December 29, 2015. 18. Sharps M, Higgs S, Blissett J, et al. Examining evidence for behavioural mimicry of parental eating by adolescent females. An observational study. Appetite. 2015;89:56–61. http://dx.doi.org/10.1016/j.appet.2015.01.015. 19. Riebl SK, MacDougal C, Hill C, et al. Beverage choices of adolescents and their parents using the theory of planned behavior: a mixed methods analysis. J Acad Nutr Diet. 2016;116(2):226–239. http://dx. doi.org/10.1016/j.jand.2015.10.019. 20. National Cancer Institute. Screeners at a Glance. Dietary Assessment Primer. https://dietassessmentprimer.cancer.gov/profiles/screeners/. Accessed March 13, 2015. 21. Yaroch AL, Tooze J, Thompson FE, et al. Evaluation of three short dietary instruments to assess fruit and vegetable intake: the National Cancer Institute’s food attitudes and behaviors survey. J Acad Nutr Diet. 2012;112(10):1570–1577. http://dx.doi.org/10.1016/j.jand.2012.06.002. 22. Subar AF, Thompson FE, Kipnis V, et al. Comparative validation of the Block, Willett, and National Cancer Institute food frequency questionnaires: the Eating at America’s Table Study. Am J Epidemiol. 2001;154(12):1089–1099. http://dx.doi.org/10.1093/aje/154.12.1089. 23. Young LR, Nestle MS. Portion sizes in dietary assessment: issues and policy implications. Nutr Rev. 1995;53(6):149–158. http://dx.doi.org/ 10.1111/j.1753-4887.1995.tb01542.x. 24. Kim DJ, Holowaty EJ. Brief, validated survey instruments for the measurement of fruit and vegetable intakes in adults: a review. Prev Med. 2003;36(4):440–447. http://dx.doi.org/10.1016/S0091-7435(02) 00040-3.
www.ajpmonline.org
Using Secondary 24-Hour Dietary Recall Data to Estimate Daily Dietary Factor Intake From the FLASHE Study Dietary Screener Teresa M. Smith, PhD,1 Eric E. Calloway, PhD,1 Courtney A. Pinard, PhD,1 Erin Hennessy, PhD, MPH,2 April Y. Oh, PhD, MPH,3 Linda C. Nebeling, PhD, MPH, RD,3 Amy L. Yaroch, PhD1 Introduction: The National Cancer Institute’s 2014 Family Life, Activity, Sun, Health, and Eating Study utilized a 27-item Dietary Screener tailored to adolescent eating patterns that assessed the frequency of intake of several foods and beverages in parent–adolescent dyads. This study estimated intake of fruits and vegetables (FVs), dairy, added sugars, and whole grains for screener respondents using existing, nationally representative, 24-hour dietary recall data.
Methods: Dietary Screener items were converted from frequency responses to daily intake. Intake (dependent variable) was estimated using regression coefficients and portion sizes of foods and beverages (independent variables) generated from the 2003–2006 National Health and Nutrition Examination Survey 2-day 24-hour recall data set. Means (SDs) were used to examine daily dietary factor intake among parent and adolescents. Analysis was conducted in 2015–2016. The analytic sample consisted of 1,732 parents (aged Z18 years) and their adolescent aged 12–17 years (n¼1,632).
Results: Male parents consumed 3.6 cups of FVs, 1.8 cups of dairy, 22.6 teaspoons of added sugars, and 2.1 ounces of whole grains daily; female parents consumed 2.8 cups of FVs, 1.3 cups of dairy, 14.8 teaspoons of added sugars, and 1.4 ounces of whole grains daily. Male adolescents consumed 2.2 cups of FVs, 1.9 cups of dairy, 17.9 teaspoons of added sugars, and 1.0 ounces of whole grains daily; female adolescents consumed 2.2 cups FVs, 1.6 cups of dairy, 14.2 teaspoons of added sugars, and 0.8 ounces of whole grains daily.
Conclusions: Utilizing a dietary screener tailored to adolescent eating patterns in parent– adolescent dyads provided estimated dietary factor intake, underscoring existing 24-hour dietary recall data can be used to calibrate dietary habits. Am J Prev Med 2017;52(6):856–862. & 2017 American Journal of Preventive Medicine. Published by Elsevier Inc. All rights reserved.
INTRODUCTION
M
odifiable behaviors (e.g., poor diet and physical inactivity) that contribute to obesity and increase risk for chronic diseases, such as cancer, continue to afflict much of the population in the U.S. The National Cancer Institute (NCI),1 as well as other organizations such as the World Cancer Research Fund and American Institute for Cancer Research,2 American Cancer Society,3 and WHO,4 have emphasized diet, and especially specific food groups, as having potential to increase risk or prevent some cancers. Fruits and vegetables (FVs) and whole grains, for example, are often cited as potentially preventative of commonly 856 Am J Prev Med 2017;52(6):856–862
occurring cancers in the U.S. (e.g., breast, colorectal, and prostate), whereas foods high in added sugars and fats may increase risk for some cancers.1–4 Studies From the 1Gretchen Swanson Center for Nutrition, Omaha, Nebraska; 2 Tufts University Friedman School of Nutrition Science and Policy, Boston, Massachusetts; and 3Behavioral Research Program, National Cancer Institute, Bethesda, Maryland Address correspondence to: Erin M. Hennessy, PhD, MPH, Tufts University Friedman School of Nutrition Science and Policy, 150 Harrison Avenue, Boston MA 02111. E-mail: [email protected]. This article is part of a theme section titled The Family Life, Activity, Sun, Health, and Eating (FLASHE) Study: Insights Into Cancer-Prevention Behaviors Among Parent–Adolescent Dyads. 0749-3797/$36.00 http://dx.doi.org/10.1016/j.amepre.2017.01.015
& 2017 American Journal of Preventive Medicine. Published by Elsevier Inc. All rights reserved.
Smith et al / Am J Prev Med 2017;52(6):856–862
investigating individual food groups (e.g., FVs) and nutrients (e.g., fiber and sodium) have generated evidence supporting the relationship between dietary patterns with cancer incidence.5,6 Understanding correlates of dietary intake of these key food groups may help inform environmental-, familial-, and individual-based interventions focused on increasing healthy eating and cancer prevention. In particular, though it has been demonstrated that children (aged 6–12 years) and their parents have closely related diet quality and energy intake, perhaps because of shared food environments and parental modeling,7 this relationship and the mechanisms driving this relationship are unknown among adolescents and their parents. The Family Life, Activity, Sun, Health, and Eating (FLASHE) Study is a publicly available database developed by NCI that surveyed the psychosocial; generational (adolescents aged 12–17 years and their parent); and environmental correlates of cancer preventive behaviors.8 The survey was administered in 2014 with the goal of advancing the understanding of the relationships among the environment, psychosocial, and behavioral factors from a dyadic perspective. FLASHE included a 27-item Dietary Screener that asked about frequency of foods eaten and beverages consumed in the past 7 days (http://cancercontrol.cancer.gov/brp/hbrb/flashe.html). The FLASHE Dietary Screener was based on the Dietary Screener Questionnaire (DSQ) used in the 2009–2010 National Health and Nutrition Examination Survey (NHANES) but was tailored by NCI to inquire further on the foods and beverages commonly consumed among adolescents in the U.S., which were determined through literature review and by content experts, and reviewed by an external scientific advisory committee.8 The FLASHE Dietary Screener included common items (e.g., green salad, non-fried vegetables, cooked beans, fruit, cooked whole grains, whole grain bread, non-sugary cereal, fried chicken, sugary cereal, candy and chocolate, fried potatoes, other non-fried potatoes) as well as novel items (e.g., chips, processed meat, burgers, tacos, heat and serve, and others) to capture patterns of dietary behavior among adolescents and their families. Intake of screener items not described in this manuscript (e.g., novel items) will be described using frequencies of reported intake in a future manuscript. Assessing and characterizing dietary intake to identify areas for improvement in diet is an essential first step in promoting a more healthful diet (e.g., which may include greater FV intake) and ultimately reducing risk for certain cancers.9 NHANES typically collects diet information through 24-hour recalls; however, the 2009–2010 survey included the DSQ. In recognition of a need for characterizing a population’s median intakes or June 2017
857
examining inter-relationships between diet and other variables, the Risk Factor Assessment Branch (RFAB) of NCI developed scoring algorithms using 2003–2006 NHANES 24-hour recall data to convert screener responses into estimates of intake for several dietary factors that have remained of interest in dietary guidance in America.10–12 The dietary factors included cup equivalents of FVs and dairy, teaspoon equivalents of added sugars, ounce equivalents of whole grains, grams of fiber, and milligrams of calcium.10 The authors worked with RFAB to convert FLASHE Dietary Screener responses into intake of dietary factors believed to be of interest to researchers. RFAB tailored scoring algorithms developed for the DSQ based on the 2003–2006 NHANES 24-hour recall data.13 These derived variables are included in the publically available data set for researchers to explore dietary factor intake of the FLASHE sample. The purpose of this study was to estimate key dietary factor intake for this sample of adolescents and parents, as well as the relationship of adolescent intake to their parent’s intake.
METHODS Study Sample This study analyzed secondary data from the 2014 FLASHE study, a cross-sectional web survey administered from April to October 2014. The FLASHE study was approved by the U.S. Government’s Office of Management and Budget, National Cancer Institute’s Special Studies IRB, and Westat’s IRB. FLASHE methodology and sample design are described in detail elsewhere.14 In brief, the survey sample came from a commercial online consumer opinion panel; the response rate for parent surveys was 34.0% and for adolescent surveys was 31.6%.14 Sampled parents and their adolescent children completed self-reported items about their neighborhood and home environment, psychosocial factors, and behaviors related to cancer risk in two surveys, a diet-focused and physical activity–focused survey. The sample of participants that completed the diet-focused survey consisted of 1,745 adults and 1,657 adolescents. Respondents were excluded if their age or sex were not ascertained (n¼13) and for adolescents, if their age was outside of the range of 12–17 years (n¼25). Final analytic samples included 1,732 adults and 1,632 adolescents.
Measures The FLASHE Dietary Screener is unique in that although it is based on the 26-item DSQ, it is tailored to capture foods and beverages more commonly consumed by adolescents in the U.S. The screener is administered similarly among parents. To estimate daily intake of dietary factors (i.e., food, beverage, or food groups) for the analytic sample, FLASHE Dietary Screener items were assigned to dietary factors (Appendix [available online] lists 18 of the 27 FLASHE Dietary Screener items used to estimate daily intake of dietary factors). Estimated daily intake of dietary factors from FLASHE Dietary Screener items were based upon
858
Smith et al / Am J Prev Med 2017;52(6):856–862
relationships detailed in the DSQ (e.g., fruit, fruit juice, salad, fried potatoes, other potatoes, dried beans, other vegetables, tomato sauce, salsa, and pizza were used to estimate fruits and vegetables cup equivalents per day in the DSQ),13 a review of the literature, expert consultation, and internal consensus. Estimated dietary factor intake was calculated for parents and adolescents using SAS programs obtained from RFAB at NCI. The SAS programs were originally developed to compare the 2009–2010 NHANES dietary screener responses (similar to the FLASHE Dietary Screener) to the What We Eat in America (WWEIA) 24-hour dietary recall data from the 2003–2006 NHANES, and are currently described in detail elsewhere.15 RFAB is currently developing direct calibrations of 2009–2010 NHANES 24-hour recall data to the DSQ, but they are not finalized; the most advanced science available, which included an indirect calibration with WWEIA 24-hour dietary recall data from the 2003–2006 NHANES, was utilized for this study.10 The SAS programs were tailored by RFAB at NCI to the FLASHE Dietary Screener items. A detailed description of this methodology, including how RFAB estimated portion size and regression coefficients from the 2003–2006 NHANES data set, can be found at epi.grants.cancer.gov/nhanes/dietscreen/scoring/. Additional steps, however, were warranted for aligning the FLASHE Dietary Screener items non-sugary cereal, sugary cereal, and non-soda sugar-sweetened beverages (SSBs) to the DSQ screener items before RFAB was able to tailor algorithms to this sample. Frequency of key foods, beverages, and food groups were assessed through the FLASHE 27-item Dietary Screener that asked respondents about frequency of foods eaten (green salad, non-fried vegetables, cooked beans, fruit, cooked whole grains, whole-grain bread, nonsugary cereal, fried chicken, sugary cereal, candy and chocolate, fried potatoes, other non-fried potatoes, chips, processed meat, burgers, tacos, heat and serve, cookies, cake, frozen desserts, pizza) and beverages consumed (water, fruit juice, milk, sweetened fruit drinks, regular soda, sports drinks, energy drinks) in the past 7 days. Respondents selected one of six response options that included did not consume in past 7 days; 1–3 times in past 7 days; 4–6 times in past 7 days; 1 time per day; 2 times per day; or 3 or more times per day. To estimate dietary factor intake, the first three responses (i.e., did not consume, 1–3 times, and 4–6 times in past 7 days) were converted into daily frequency by dividing by 7 days. In cases where the response included a range, the mid-value was selected, and divided by 7 days. There was one exception with regard to pizza, which was top-coded so that responses greater than or equal to two times per day were restricted to two. The top-code value of two was determined to be the maximum plausible intake value for pizza based on previously conducted work that used NHANES analysis of 2-day mean intakes from 24-hour recall data across all sex and age groups.16 Age; sex; household income (parents only); education level (parents only); race/ethnicity; health status; height; and weight were also self-reported in FLASHE.
Statistical Analysis Mean (SD) or percentage values were used to examine daily frequency of screener items, daily dietary factor intake, and sociodemographics among parent and adolescents. SAS, version 9.4, was used for statistical analysis. Daily dietary factor intake was calculated for the following: cup equivalents of fruits, vegetables, FVs combined (with and without fried potatoes); cup equivalents of dairy; teaspoon equivalents of added sugars (for all sugary foods and SSBs only); and ounce equivalents of whole grains based on
FLASHE Dietary Screener items. Additionally, the ratio of the estimated amount consumed by adolescents to the estimated amount consumed by their parent was reported. Analysis was conducted in 2015–2016.
RESULTS As shown in Table 1, parents in this sample were predominantly aged 35–44 years (44%); considered themselves to be non-Hispanic white (69%); reported an annual household income of o$100,000 per year (79%); and had earned a 4year college degree or higher (46%). Most of the parents reported that their health was very good or excellent (57%), and self-reported height and weight converted into BMI was considered to be in a normal or healthy range (BMI Z18.5 and r25.0) for 36% of parents. Roughly half (50%) of the adolescents in this sample were aged 12–14 years (Table 2). Most of the adolescents considered themselves to be non-Hispanic white (63%), followed by non-Hispanic black (17%); Hispanic (10%); and other or multiple races (9%). The majority of adolescents reported their health was very good or excellent (79%). The self-reported height and weight converted into BMI percentile was considered to be in a normal or healthy range (BMI percentile Z5.0 and r85.0) for 66% of adolescents. The means (SDs) of daily frequency intake of key screener items across adolescent respondents and their parents are shown in Table 3. Adolescents reported consuming milk and milk alternatives most frequently (0.99 [SD¼0.88] times per day) and cooked beans least frequently (0.22 [SD¼0.38] times per day). Their parents reported consuming fruit most frequently (0.97 [SD¼0.85] times per day) and sugary cereal least frequently (0.13 [SD¼0.30] times per day). Table 4 shows the means (SDs) of the daily dietary factor intake of key food groups and nutrients among adult and adolescent respondents. Based on the algorithm used to convert frequency of consumption to cup equivalents, parents in the FLASHE sample were consuming 3.6 cup equivalents (men) and 2.8 cup equivalents (women) of FVs; 1.8 cup equivalents (men) and 1.3 cup equivalents (women) of dairy; 22.6 teaspoon equivalents (men) and 14.8 teaspoon equivalents (women) of added sugars; and 2.1 ounce equivalents (men) and 1.4 ounce equivalents (women) of whole grains daily. Adolescents in the FLASHE sample reported consuming 2.2 cup equivalents (male adolescents) and 2.2 cup equivalents (female adolescents) of FVs; 1.9 cup equivalents (male adolescents) and 1.6 cup equivalents (female adolescents) of dairy; 17.9 teaspoon equivalents (male adolescents) and 14.2 teaspoon equivalents (female adolescents) of added sugars; and 1.0 ounce equivalents www.ajpmonline.org
Smith et al / Am J Prev Med 2017;52(6):856–862
859
Table 1. Distribution of Self-Reported Health Characteristics and Sociodemographics Among Parent Respondents (n¼1,732) Stratified by sex Variables n Age, years 18–34 35–44 45–59 Z60 Household income $0–$99,999 Z$100,000 Highest grade or level of education completed Less than a high school degree A high school degree or GED Some college but not a college degree A 4-year college degree or higher Race/Ethnicity Non-Hispanic white Non-Hispanic black Hispanic Other Self-reported health status Poor or fair Good Very good or excellent BMI Underweight (BMI o18.50) Normal weight (BMI Z18.50 and r24.99) Overweight (BMI Z25.00 and r29.99) Obese (BMI Z30.00)
Total, n (%)
Male, n (%)
Female, n (%)
454 (26.2)
1278 (73.8)
196 (11.3) 753 (43.5) 732 (42.3) 51 (2.9)
43 157 224 30
153 596 508 21
1359 (78.5) 353 (20.4)
324 (71.4) 127 (28.0)
1035 (81.0) 226 (17.7)
22 (1.3) 292 (16.9) 607 (35.0) 804 (46.4)
5 (1.1) 72 (15.9) 135 (29.7) 241 (53.1)
17 220 472 563
(1.3) (17.2) (36.9) (44.1)
1199 (69.2) 293 (16.9) 125 (7.2) 99 (5.7)
321 (70.7) 51 (11.2) 48 (10.6) 31 (6.8)
878 242 77 68
(68.7) (18.9) (6.0) (5.3)
204 (11.7) 535 (30.9) 983 (56.8)
47 (10.3) 148 (32.6) 255 (56.2)
157 (12.2) 387 (30.3) 728 (57.0)
23 (1.3) 627 (36.2) 526 (30.4) 535 (30.9)
4 (0.9) 132 (29.1) 185 (40.7) 130 (28.6)
19 495 341 405
(9.5) (34.6) (49.3) (6.6)
(12.0) (46.6) (39.7) (1.6)
(1.5) (38.7) (26.7) (31.7)
Note: Percentages do not add up to 100% because of missing responses. GED, General Educational Development test.
(male adolescents) and 0.8 ounce equivalents (female adolescents) of whole grains daily. In this sample, adolescents reported lower consumption of cup equivalents of vegetables, FVs, and FVs without fried potatoes when compared with their parent (68%, 80%, and 80%, respectively). They reported greater consumption, however, of cup equivalents of fruits (59% more); teaspoons of added sugars (30% more); cup equivalents of dairy (37% more); and ounce equivalents of whole grains (29% more). Of greatest note is that adolescents reported consuming about six times as many teaspoon equivalents of added sugars from SSBs compared with their parent.
DISCUSSION The intent of this study was to estimate key daily dietary factor intake for a sample of parents and adolescents using FLASHE Dietary Screener responses and to illustrate subsequent values among the participating sample. June 2017
Although this sampling design of FLASHE was different from that of NHANES, it is noteworthy that when indirectly compared to 24-hour recall data from WWEIA, 2009–2010 NHANES weighted Mean Daily Food Patterns Cup Equivalents tables,17 parents in the FLASHE sample reported consuming amounts relatively close to adults aged Z20 years in the WWEIA sample. For illustrative purposes, FLASHE male and female parents reported consuming 1.8 cups and 1.4 cups of vegetables a day, respectively, compared with male and female WWEIA participants who consumed 1.7 cups and 1.5 cups, respectively. FLASHE male and female adolescents reported consuming 1.0 cups of vegetables a day, compared with male and female participants aged 12–19 years in the WWEIA sample who consumed 1.2 cups and 1.0 cups, respectively. Though the age ranges of the samples differ, when comparing to the national sample, these data suggest that variables derived in the current study are reasonable estimates of dietary intake for parents and adolescents who participated in FLASHE.
860
Smith et al / Am J Prev Med 2017;52(6):856–862
Table 2. Distribution of Self-Reported Health Characteristics and Sociodemographics Among Adolescent Respondents (N¼1,632) Stratified by Sex Total, n (%)
Male, n (%)
Female, n (%)
809 (49.6)
823 (50.4)
820 (50.2) 812 (49.8)
394 (48.7) 415 (51.3)
426 (51.8) 397 (48.2)
1,030 (63.1) 272 (16.7) 160 (9.8) 152 (9.3)
501 (61.9) 137 (16.9) 83 (10.3) 80 (9.9)
529 (64.3) 135 (16.4) 77 (9.4) 72 (8.7)
91 (5.6) 241 (14.8) 1,295 (79.4)
36 (4.4) 102 (12.6) 669 (82.7)
55 (6.7) 139 (16.9) 626 (76.1)
66 (4.0) 1,085 (66.5) 245 (15.0) 201 (12.3)
42 514 130 109
24 (2.9) 571 (69.4) 115 (14.0) 92 (11.2)
Variables n Age, years 12–14 15–17 Race/ethnicity Non-Hispanic white Non-Hispanic black Hispanic Other Self-reported health status Poor or fair Good Very good or excellent BMI-for-age percentile classification Underweight (BMI percentile o5.00) Normal weight (BMI percentile Z5.00 and r84.99) Overweight (BMI percentile Z85.00 and r94.99) Obese (BMI percentile Z95.00)
(5.2) (63.5) (16.1) (13.5)
Note: Percentages do not add up to 100% because of missing responses.
When considering the dyad-level ratio of amounts of dietary factors consumed by adolescents and their parent, adolescents generally reported consuming lesser amounts of Table 3. Mean ⫾ SD of Daily Frequency Intake of Key Screener Items Across Adolescent Respondents and Their Parent
a
FLASHE screener items
Adolescent
Parent
100% fruit juice Candy/chocolate Cooked beans Cooked whole grains Cookies/cakes Fried potatoes Frozen desserts Fruit Green salad Milk and milk alternatives Non-soda SSBa Non-sugary cereal Other non-fried potatoes Other non-fried vegetables Pizza Soda Sugary cereal Whole grain bread
0.51⫾0.63 0.49⫾0.53 0.22⫾0.38 0.23⫾0.39 0.40⫾0.46 0.33⫾0.36 0.29⫾0.36 0.86⫾0.80 0.38⫾0.50 0.99⫾0.88 0.83⫾1.04 0.24⫾0.38 0.25⫾0.36 0.60⫾0.59 0.28⫾0.28 0.48⫾0.65 0.27⫾0.40 0.45⫾0.54
0.40⫾0.56 0.41⫾0.48 0.25⫾0.36 0.22⫾0.34 0.30⫾0.39 0.26⫾0.32 0.21⫾0.30 0.97⫾0.85 0.50⫾0.51 0.37⫾0.39 0.50⫾0.93 0.22⫾0.34 0.25⫾0.32 0.75⫾0.67 0.22⫾0.24 0.46⫾0.74 0.13⫾0.30 0.43⫾0.49
Composite of sweetened fruit drinks, energy drinks, and sports drinks. FLASHE, Family Life, Activity, Sun, Health, and Eating Study; SSB, sugarsweetened beverages.
vegetables, as well as FVs combined (with and without fried potatoes). However, adolescents reported consuming greater amounts of fruits; added sugars (including those in SSBs); dairy; and whole grains. A recent study found that female adolescents (aged 12.0–18.8 years) tend to mimic their parents’ eating behaviors with regard to the same foods.18 Though most of the ratio values in the current study would support this finding, adolescents reported consuming more than six times as many teaspoons of added sugars from SSBs, which can equate to nearly half a can of regular soda per day. Adolescent SSB intake has been found by a different study to be potentially influenced by their parent(s), but also other people, such as peers or other adult authority figures.19 Future research should seek to understand correlates of dietary behaviors between members of parent–adolescent dyads, especially among the dietary factors identified in the current study, in order to inform family-based interventions.
Limitations This study has several limitations to report. First, because data collected using the FLASHE Survey were self-report, they cannot be independently verified for accuracy. Second, the sample was drawn using a consumer opinion panel and not a random sample, though similar data collected technique have been used in other large health behavior surveys (e.g., NCI’s Food Attitudes and Behaviors Survey and the Centers for Disease Control and Prevention’s Styles surveys). Third, these estimates were www.ajpmonline.org
June 2017
1.30 (1.22, 1.36) 6.04 (5.32, 6.77) 1.37 (1.32, 1.41) 1.29 (1.16, 1.41) 14.81⫾8.60 7.86⫾10.43 1.33⫾0.46 1.43⫾3.33 22.55⫾12.75 14.35⫾15.01 1.75⫾0.68 2.14⫾4.83 16.86⫾10.46 9.58⫾12.18 1.44⫾0.56 1.62⫾3.81 eq, equivalent; SSB, sugar-sweetened beverage; tsp, teaspoon.
17.83⫾7.88 7.74⫾7.50 1.85⫾0.75 1.00⫾1.79 16.00⫾7.66 6.68⫾7.00 1.73⫾0.73 0.89⫾1.57
14.20⫾6.94 5.69⫾6.32 1.61⫾0.70 0.79⫾1.37
(1.42, 1.75) (0.67, 0.70) (0.78, 0.82) (0.78, 0.82) 1.59 0.68 0.80 0.80 1.24⫾0.88 1.41⫾0.52 2.79⫾1.33 2.69⫾1.29 1.64⫾1.09 1.79⫾0.61 3.61⫾1.45 3.44⫾1.39 1.34⫾0.95 1.51⫾0.57 3.00⫾1.41 2.88⫾1.36 1.08⫾0.73 0.96⫾0.36 2.18⫾0.99 2.07⫾0.96
Fruits (cup eq) Vegetables (cup eq) Fruits and vegetables (cup eq) Fruits and vegetables without fried potatoes (cup eq) Added sugars (tsp) Added sugars from SSB (tsp) Dairy (cup eq) Whole grains (ounce eq)
1.06⫾0.72 0.96⫾0.38 2.16⫾1.05 2.05⫾0.99
1.04⫾0.71 0.96⫾0.41 2.15⫾1.10 2.03⫾1.03
Mean (95% CI) Female Male Male Dietary factor
Total
Female
Total
Parent Adolescent
Table 4. Mean ⫾ SD of Daily Dietary Factor Intake of Food Groups and Nutrients Across Adolescent Respondents and Their Parent
Ratio of adolescent to parent
Smith et al / Am J Prev Med 2017;52(6):856–862
861
based on scoring algorithms developed to estimate dietary factor intake using 2003–2006 NHANES 24-hour recall data and eating patterns may differ in portion size and overall contribution to the estimated nutrients, though portion sizes were estimated based on age and sex and applied accordingly to participants of the current study. RFAB is currently developing direct calibrations of 2009–2010 NHANES 24-hour recall data to the DSQ, but as they are not finalized, the most advanced science available was utilized.10 Lastly, it is unknown how accurate responses are for adolescents participating in frequency-based questionnaires. Despite these limitations, this study has many strengths to report, namely, the use of 24-hour recall data to estimate dietary intake from Dietary Screener responses. Dietary screeners, including the DSQ and FLASHE Dietary Screener, ask about intake of food, beverages, or food groups of particular interest, and then the frequency of consumption of screener items or patterns of eating practices can be used as indicators of particular aspects of diet.20 It is important to note that some dietary screeners probe respondents about the portion sizes of food/beverages they consume (i.e., FV screeners in NCI’s Food Attitudes and Behaviors Survey),21,22 which can allow for analysis and reporting on a standard amount consumed (e.g., cups, teaspoons).23 However, there can be methodologic issues, such as inconsistent results, high error rates, and compliance issues, when utilizing participant-reported portion sizes for estimating dietary intake.23 The FLASHE Dietary Screener was a feasible instrument, as it provided a short, relatively simple, and a moderately valid way to rank individuals with regard to consumption of certain foods and nutrients.21,24 In this study, it was important to be able to estimate a standardized amount of food consumed. Future studies using FLASHE data could have a similar need, and researchers may seek to compare the intake of a study sample to a national average or dietary recommendation.
CONCLUSIONS This study advances the literature in that it utilized a dietary screener tailored to adolescent eating patterns to estimate dietary factor intake of adolescents and their parents. Because of this design, data estimated from this screener, as well as the screener itself, is useful in dyadic and familial dietary research. Further, it underscores the utility of advanced dietary assessment and analysis in assessing population-level dietary habits. This screener and the accompanying publically available data set, including the derived variables described in this study, are available for researchers to explore dietary factor intake of parent–adolescent dyads; characteristics and
862
Smith et al / Am J Prev Med 2017;52(6):856–862
dietary intake of subsamples; and relationship of dietary intake to other health outcomes and lifestyles (e.g., BMI and self-reported health status).
ACKNOWLEDGMENTS This article is part of a theme issue supported by the National Institutes of Health. The findings and conclusions in this article are those of the author(s) and do not necessarily represent the official position of the National Institutes of Health. Leslie A. Lytle, Ph.D. and Louise C. Mâsse, Ph.D. served as Guest Editors of the theme issue. Thank you to Frances E. Thompson, MPH, PhD, epidemiologist in the Risk Factor Assessment Branch at the National Cancer Institute for developing the screener and its algorithms and Lisa L. Kahle of Information Management Services for developing the SAS programs used in this analysis. Thank you to Casey Blaser for advising on the statistical analysis. The Family Life, Activity, Sun, Health, and Eating (FLASHE) study was funded by the National Cancer Institute under contract number HHSN261201200039I issued to Westat. The Gretchen Swanson Center for Nutrition (Authors Smith, Calloway, Pinard, and Yaroch) worked on FLASHE via subcontract to Westat under this contract. At the time this study was conducted, Dr. Hennessy was a Senior Behavioral Scientist (Contractor), Clinical Research Directorate/ Clinical Monitoring Research Program, Leidos Biomedical Research, Inc., Frederick National Laboratory for Cancer Research. No financial disclosures were reported by the authors of this paper.
SUPPLEMENTAL MATERIAL Supplemental materials associated with this article can be found in the online version at http://dx.doi.org/10.1016/j. amepre.2017.01.015.
REFERENCES 1. Klein WMP, Bloch M, Hesse BW, et al. Behavioral research in cancer prevention and control: a look to the future. Am J Prev Med. 2014; 46(3):303–311. http://dx.doi.org/10.1016/j.amepre.2013.10.004. 2. World Cancer Research Fund, American Institute for Cancer Research (AICR). Food, Nutrition, Physical Activity, and the Prevention of Cancer: A Global Perspective. Washington DC: AICR; 2007. 3. Kushi LH, Doyle C, McCullough M, et al. American Cancer Society guidelines on nutrition and physical activity for cancer prevention: reducing the risk of cancer with healthy food choices and physical activity. CA Cancer J Clin. 2012;62(1):30–67. http://dx.doi.org/10.3322/caac.20140. 4. Joint WHO-FAO Expert Consultation on Diet, Nutrition, and the Prevention of Chronic Diseases. Diet, Nutrition, and the Prevention of Chronic Diseases: Report of a WHO-FAO Expert Consultation. Geneva: WHO; 2003. 5. Key TJ, Schatzkin A, Willett WC, Allen NE, Spencer EA, Travis RC. Diet, nutrition and the prevention of cancer. Public Health Nutr. 2004;7(1A):187–200. http://dx.doi.org/10.1079/phn2003588. 6. Emmett PM, Jones LR, Northstone K. Dietary patterns in the Avon Longitudinal Study of Parents and Children. Nutr Rev. 2015;73(suppl 3): 207–230. http://dx.doi.org/10.1093/nutrit/nuv055. 7. Robson SM, Couch SC, Peugh JL, et al. Parent diet quality and energy intake are related to child diet quality and energy intake. J Acad Nutr Diet. 2016;116 (6):984–990. http://dx.doi.org/10.1016/j.jand.2016.02.011.
8. Nebeling LC, Hennessy E, Oh AY, et al. The FLASHE study: survey development, dyadic perspectives, and participants characteristics. Am J Prev Med. 2017;52(6):839–848. 9. Patrick K, Norman GJ, Calfas KJ, et al. Diet, physical activity, and sedentary behaviors as risk factors for overweight in adolescence. Arch Pediatr Adolesc Med. 2004;158(4):385–390. http://dx.doi.org/10.1001/ archpedi.158.4.385. 10. Division of Cancer Control and Population Sciences, National Cancer Institute. Dietary Screener Questionnaire in the NHANES 2009-10: background. Epidemiology and Genomics Research Program. http:// epi.grants.cancer.gov/nhanes/dietscreen/. Published October 15, 2015. Accessed December 1, 2015. 11. Dietary Guidelines Advisory Committee. Report of the Dietary Guidelines Advisory Committee on the Dietary Guidelines for Americans, 2010, to the Secretary of Agriculture and the Secretary of Health and Human Services. Agric Res Serv. 2010. 12. U.S. DHHS, U.S. Department of Agriculture. 2015–2020 Dietary Guidelines for Americans. http://health.gov/dietaryguidelines/2015/ guidelines/. Published 2015. Accessed November 16, 2016. 13. National Cancer Institute, Applied Research Cancer Control and Population Sciences. Dietary Screener Questionnaire in the NHANES 2009-10. http://appliedresearch.cancer.gov/nhanes/dietscreen/. Published April 11, 2014. 14. Westat. Family Life, Activity, Sun, Health, and Eating (FLASHE) Study Methodology Report. http://cancercontrol.cancer.gov/brp/hbrb/docs/FLA SHE_Methods_Report.pdf. Published 2015. Accessed October 6, 2016. 15. Division of Cancer Control and Population Sciences, National Cancer Institute. Dietary Screener Questionnaire (DSQ) in the NHANES 200910: data processing and scoring procedures. Epidemiology and Genomics Research Program. http://epi.grants.cancer.gov/nhanes/dietscreen/scor ing/. Published October 15, 2015. Accessed December 29, 2015. 16. Epidemiology and Genomics Research Program. Identifying extreme exposure values (earlier methods). https://epi.grants.cancer.gov/ nhanes/dietscreen/scoring/earlier/identify.html. Published October 15, 2015. Accessed December 20, 2016. 17. U.S. Department of Agriculture, Agricultural Research Service. Food patterns equivalents intakes from food: mean amounts consumed per individual, by gender and age. What We Eat in America. NHANES 2009-2010. www.ars.usda.gov/ba/bhnrc/fsrg. Published 2013. Accessed December 29, 2015. 18. Sharps M, Higgs S, Blissett J, et al. Examining evidence for behavioural mimicry of parental eating by adolescent females. An observational study. Appetite. 2015;89:56–61. http://dx.doi.org/10.1016/j.appet.2015.01.015. 19. Riebl SK, MacDougal C, Hill C, et al. Beverage choices of adolescents and their parents using the theory of planned behavior: a mixed methods analysis. J Acad Nutr Diet. 2016;116(2):226–239. http://dx. doi.org/10.1016/j.jand.2015.10.019. 20. National Cancer Institute. Screeners at a Glance. Dietary Assessment Primer. https://dietassessmentprimer.cancer.gov/profiles/screeners/. Accessed March 13, 2015. 21. Yaroch AL, Tooze J, Thompson FE, et al. Evaluation of three short dietary instruments to assess fruit and vegetable intake: the National Cancer Institute’s food attitudes and behaviors survey. J Acad Nutr Diet. 2012;112(10):1570–1577. http://dx.doi.org/10.1016/j.jand.2012.06.002. 22. Subar AF, Thompson FE, Kipnis V, et al. Comparative validation of the Block, Willett, and National Cancer Institute food frequency questionnaires: the Eating at America’s Table Study. Am J Epidemiol. 2001;154(12):1089–1099. http://dx.doi.org/10.1093/aje/154.12.1089. 23. Young LR, Nestle MS. Portion sizes in dietary assessment: issues and policy implications. Nutr Rev. 1995;53(6):149–158. http://dx.doi.org/ 10.1111/j.1753-4887.1995.tb01542.x. 24. Kim DJ, Holowaty EJ. Brief, validated survey instruments for the measurement of fruit and vegetable intakes in adults: a review. Prev Med. 2003;36(4):440–447. http://dx.doi.org/10.1016/S0091-7435(02) 00040-3.
www.ajpmonline.org