If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address for correspondence: Kendra K. Kattelmann, PhD, RD, Didactic Program in Dietetics, Health and Nutritional Sciences, South Dakota State University, Box 2203, Brookings, SD 57007; Phone: (605) 688-4045; Fax: (605) 688-5603.
† At the time this research was conducted, Dr Brown was at Purdue University, Dr Koenings was a doctoral candidate at the University of Wisconsin–Madison, and Dr Colby was at East Carolina University.
Sarah Colby
Footnotes
† At the time this research was conducted, Dr Brown was at Purdue University, Dr Koenings was a doctoral candidate at the University of Wisconsin–Madison, and Dr Colby was at East Carolina University.
Affiliations
Department of Nutrition, University of Tennessee, Knoxville, TN
† At the time this research was conducted, Dr Brown was at Purdue University, Dr Koenings was a doctoral candidate at the University of Wisconsin–Madison, and Dr Colby was at East Carolina University.
Mallory M. Koenings
Footnotes
† At the time this research was conducted, Dr Brown was at Purdue University, Dr Koenings was a doctoral candidate at the University of Wisconsin–Madison, and Dr Colby was at East Carolina University.
Affiliations
Department of Nutritional Sciences, Rutgers University, New Brunswick, NJ
† At the time this research was conducted, Dr Brown was at Purdue University, Dr Koenings was a doctoral candidate at the University of Wisconsin–Madison, and Dr Colby was at East Carolina University.
Onikia N. Brown
Footnotes
† At the time this research was conducted, Dr Brown was at Purdue University, Dr Koenings was a doctoral candidate at the University of Wisconsin–Madison, and Dr Colby was at East Carolina University.
Affiliations
Department of Nutrition, Dietetics, and Hospitality Management, Auburn University, Auburn, AL
† At the time this research was conducted, Dr Brown was at Purdue University, Dr Koenings was a doctoral candidate at the University of Wisconsin–Madison, and Dr Colby was at East Carolina University.
To assess the effectiveness of a tailored theory-based, Web-delivered intervention (Young Adults Eating and Active for Health) developed using community-based participatory research process.
Design
A 15-month (10-week intensive intervention with a 12-month follow-up) randomized, controlled trial delivered via Internet and e-mail.
Setting
Thirteen college campuses.
Participants
A total of 1,639 college students.
Intervention
Twenty-one mini-educational lessons and e-mail messages (called nudges) developed with the non-diet approach and focusing on eating behavior, physical activity, stress management, and healthy weight management. Nudges were short, frequent, entertaining, and stage-tailored to each behavior, and reinforced lesson content.
Main Outcome Measure
All participants were assessed at baseline, postintervention (3 months from baseline), and follow-up (15 months from baseline) for primary outcomes of weight, body mass index (BMI), fruit and vegetable intake (FVI), physical activity (PA), and perceived stress; and secondary outcomes of waist circumference, percent dietary fat, energy from sugar-sweetened beverages, servings of whole grains, self-instruction and regulation for mealtime behavior, hours of sleep, and stage of readiness for change for consuming 5 cups of FVI, completing 150 minutes of PA/wk, and managing stress on most days of the week. Demographics were collected at baseline.
Analysis
Chi-square analysis and mixed-models repeated measures analysis were performed to determine differences between experimental and control outcomes.
Results
There were no differences between experimental and control participants in BMI, weight, and waist circumference. There were small improvements in FVI (P = .001), vigorous PA in females (P = .05), fat intake (P = .002), self-instruction (P = .001), and regulation (P = .004) for mealtime behavior, and hours of sleep (P = .05) at postintervention, but improvements were not maintained at follow-up. At postintervention, a greater proportion of experimental participants were in the action/maintenance stages for FVI (P = .019) and PA (P = .002) than control.
Conclusions and Implications
Young Adults Eating and Active for Health is one of the first studies to use the community-based participatory research process of PRECEDE-PROCEED to develop a non-diet approach intervention. Although there were no differences between experimental and control participants in weight change or BMI, the intervention supported positive change in behaviors that may mediate excessive weight gain, such as increasing FVI and more healthful self-regulation mealtime behaviors immediately postintervention. Additional strategies to maintain the behavior changes need to be explored.
Emerging adults, aged 18–24 years, are at particularly high risk for weight gain because young adulthood is a transitional time when adverse changes in body weight often occur,
Approaches targeted to this life stage are needed to prevent unhealthy weight gain, thereby reducing the need for future weight loss and health problems.
Furthermore, these characteristics may be associated with an increase in perceived stress and lead to young adults using less healthful stress management strategies. Researchers have reported unhealthful eating, insufficient sleep and physical activity, and increased alcohol and tobacco use at these ages.
Nutrition and health educators are uniquely challenged by this transitional time because any intervention must be within the context of emerging adults' perceptions about health, quality of life, sense of independence, and personal life control. Thus, it is not surprising that few obesity prevention interventions target the sensibilities and interests of emerging adults.
The Internet has been used to deliver Web-based behavior change interventions for weight loss
Young Adults Eating and Active for Health (YEAH) was developed to determine the effects of a tailored theory-based, Web-delivered intervention on weight status in college students. The researchers used a community-based participatory research (CBPR) approach because it permits collaboration with the target audience to inform and understand the audience's needs and wants to formulate interventions desired.
PRECEDE-PROCEED is a CBPR model used with communities to untangle and understand the complex behavioral and environmental factors that influence health and quality of life.
In short, because students reported sleep, stress, and time management as the most salient issues in college, lessons devoted to these topics were added to a previous Web-based weight management intervention
Phases 4–6 of the PRECEDE-PROCEED model include intervention alignment, administrative and policy assessment, and implementation and process evaluation. Phases 7 and 8 are impact and outcome evaluations as tied to measurable objectives determined in the first steps of the process.
The purpose of this article was to report results of the implementation (PROCEED Phase 5), process evaluation (PROCEED Phase 6), and outcomes (PROCEED Phases 7 and 8) for Project YEAH. The authors hypothesized that young adult college students enrolled into Project YEAH, a tailored theory-based, Web-delivered intervention, would gain less weight, eat more fruits and vegetables, have greater physical activity, and report reduced perceived stress over the 15-month period than those in a non-intervention control group. Secondary outcomes also expected to be observed in the treatment group were lower waist circumference, better eating behaviors, greater hours of sleep, and greater readiness to change related behaviors.
Methods
The university partners in the US Department of Agriculture Multistate Research Project NC1028 (the Healthy Campus Research Consortium) developed and assessed Project YEAH. Each institutional review board from the 13 participating institutions (East Carolina University, Kansas State University, Michigan State University, Purdue University, Rutgers University, South Dakota State University, Syracuse University, Tuskegee University, University of Florida, University of New Hampshire, University of Rhode Island, University of Wisconsin–Madison, and West Virginia University) approved the study protocol. Informed consent was standardized and met all participating institutions' institutional review board requirements for the protection of human subjects.
Study Design
Project YEAH was a 15-month randomized, controlled trial delivered via Internet and e-mail. Participants were randomized into control or experimental group stratified by institution and gender via a computer-generated program. Both the experimental and control groups completed assessments at baseline, postintervention (approximately 3 months from baseline), and follow-up (approximately 15 months from baseline). Between baseline and postintervention assessments, the experimental group participated in a 10-week intervention. Control group participants had access to study materials after the follow-up assessment. The 10-week intervention was delivered the spring semester (January to May, 2011) with follow-up intervention assessments occurring in spring, 2012.
Participant Recruitment and Eligibility
Investigators recruited participants using face-to-face methods, ie, in-class and residential life housing meetings, and e-mails, letters, and flyers were posted on participating campuses. Recruitment materials directed participants to the YEAH Web site to determine their eligibility for participation.
Eligibility criteria included being 18–24 years of age and a full-time first-, second-, or third-year college student with regular access to an Internet-connected computer. The Web site excluded individuals from participation who reported being a declared major for nutrition, exercise science, and/or health promotion; being currently enrolled in a nutrition course; having a body mass index (BMI) < 18.5 kg/m2; and/or having a life-threatening illness or condition such as pregnancy or other diet- and/or activity-related medical restriction. A gender imbalance of ≥ 60% females at an institution during recruitment triggered a procedure to recruit additional males to rectify the imbalance. Researchers at the institution with the imbalance were notified to enhance efforts to recruit students from majors that traditionally have had large enrollments of males, such as engineering and mathematics.
Eligible participants who agreed to the online informed consent could then complete the baseline online questionnaire, after which they were prompted to schedule an appointment to have a baseline anthropometric assessment conducted on their campus. An automated e-mail confirmation reminded all participants of the date, time, and address for anthropometric assessment and instructed them to wear or bring light clothing and avoid food and/or beverages 3 hours before the appointment. If a participant failed to meet this protocol or was ill, the appointment was rescheduled. Paper copies of the same informed consent to which participants agreed online before enrollment were presented for signature at the start of the baseline anthropometric assessment appointment and subsequently were stored per institution requirements.
Intervention
The intervention aimed to foster lifestyle behaviors associated with healthy weight management. Only those in the experimental group had access to the intervention via a personalized Web site with password-protected access.
The intervention was developed using the CBPR process of PRECEDE-PROCEED and was delivered via the Web and e-mail.
Effects of a 16-month randomized controlled exercise trial on body weight and composition in young, overweight men and women: the Midwest Exercise Trial.
Anderson J. Millennials will make online sharing in networks a lifelong habit. Pew Internet & American Life Project. http://www.pewinternet.org/. Accessed July 9, 2014.
Briefly, the intervention consisted of 21 mini-educational lessons and e-mail messages (called nudges). The lessons were developed using Dick and Carey's
Model of Instructional Design and addressed eating behavior, physical activity, stress management, and healthy weight management through a non-diet approach. E-mail nudges were short, entertaining messages with videos personalized with the participant's name and stage-tailored to pre-contemplators, contemplators/preparers, or actor/maintainers
for fruit and vegetable consumption, physical activity, and stress management. The nudges reinforced behaviors promoted in the lessons encouraged participants to visit the Web portal to view the lessons and set goals.
During the intervention phase, participants received 3 of the stage-tailored nudges each week plus 1 encouraging them to view the new lessons. Participants were required to visit the Web site weekly to set goals for 1 or all 3 of targeted behavior(s). Within their personal Web portal, participants could view a graph of their goal(s), progress toward a goal, and recommendations for each target behavior. Participants who had not set a goal in the past 6 days were automatically sent e-mails prompting them to access their Web portal and set a goal. Viewing lessons was voluntary.
During the follow-up phase (the period of time between postintervention assessment and follow-up assessment), the frequency of e-mail nudges was reduced to 4/mo (3 stage-based messages and a reminder message to visit the Web site). The Web site remained active for review but no new lessons were added.
Baseline, Postintervention, and Follow-up Assessments
At baseline, postintervention, and follow-up, both experimental and control participants completed an in-person anthropometric assessment and a series of online questionnaires evaluating eating behavior, physical activity, and stress perceptions and stages of change for readiness to consume 5 cups of fruit and vegetables, complete planned physical activity 5 times/wk for 30 minutes, and manage stress on most days of the week. Demographics (age, gender, year in school, academic major, on- or off-campus residence, university attending, and race/ethnicity) were collected at baseline. Specific measures are described below. All participants were given modest monetary compensation at each assessment.
Anthropometric assessments
Anthropometric assessments included weight, height, BMI, and waist circumference. All were primary study outcomes except waist circumference, which was a secondary outcome. Trained study personnel at each institution completed the anthropometric assessments after participants completed the online questionnaires at baseline, postintervention, and post–follow-up.
For weight, height, and BMI, each measurement was taken twice (and repeated as needed if the variance between measurements exceeded the standard) and the average was recorded and entered immediately into the online database. Weight was assessed to the nearest 0.1 kg using either a digital or balance beam scale calibrated using standard weights before measurements. (A list of university-specific scales and stadiometers is available on request.) Height was assessed to the nearest 0.1 cm using a wall-mounted stadiometer. Body mass index was calculated using the formula of weight (kg) / [height (m)]2.
Waist circumference was measured to the nearest 0.1 cm at the level of the iliac crest using a fiberglass, non-stretchable tension tape, ensuring the tape was in a horizontal plane and not twisted. Measurements were repeated if there was > 0.5 cm difference between the measurements until 2 were within 0.5 cm. Female participants with waist circumference ≥ 88 cm and male participants ≥ 102 cm were classified as being at risk for metabolic syndrome.
The researchers assessed intake of fruits and vegetables, sweetened beverages, and whole grains and percentage of dietary fat. In addition, self-instruction for mealtime behavior intention and self-regulation of healthy mealtime behaviors were assessed. Fruit and vegetable intake was a primary study outcome; other measures of eating behavior were secondary outcomes.
Fruit and vegetable intake
The authors used the National Cancer Institute Fruit and Vegetable Screener (short form) to assess fruits and vegetable daily intake as cups per day.
that queried both how often in the past and in what amounts college students consumed soft, fruit drinks, non-diet energy drinks, and sugar-sweetened specialty coffee drinks. Participants were asked how often in the past month (never or < 1/mo to ≥ 4/d) and what amount they drank of sugar-sweetened soft drinks (none, 12-oz can, restaurant glass or cup, 20-oz bottle, or 2-L bottle), fruit drinks (none, ≤ 11.5-oz can, 20-oz bottle, or 64-oz bottle), non-diet energy drinks (none, 2- to 6-oz shot, between 6 and 16 oz, or > 16 oz), and sugar-sweetened specialty coffee drinks (none, < 12 oz, or > 12 oz). Average kilocalories per day contributed by sugar-sweetened beverages were calculated by converting frequency and amount to ounces per day and multiplying by respective kilocalories per ounce. For this sample Cronbach alpha was .44.
Whole grain intake
The researchers assessed servings of whole grains consumed per day with the following question: “How many servings of whole grains do you eat on average per day? Examples of 1 serving = 1 slice of whole wheat bread; 5–6 whole grain crackers; 0.5 cup cooked brown rice; 0.5 cup oatmeal. (Note: All grains begin as whole grains; however, if after milling they keep all the parts of the original grain in their original proportions they are still considered a whole grain. Whole grains should be listed as the first ingredient on the label.)” Answers ranged from “< 1” to “≥ 6.” MyPlate serving sizes and description of whole grains were used as examples in the question (http://www.choosemyplate.gov/food-groups/grains-counts.html).
Self-instruction for healthful mealtime behavior intention and self-regulation of healthful mealtime behavior
Self-instruction for intention for healthful mealtime behavior (ie, planning, choosing, and assembling healthful meals) was measured with 6 items and self-regulation for engaging in healthful mealtime behavior was measured using 4 items adapted from Strong et al.
To determine self-instruction for healthful mealtime behavior intention, participants were asked how often in the past 3 months they had: “(1) reminded myself that planning quick and simple meals is important, (2) told myself that healthy meals do not require a lot of work, (3) reminded myself to eat in moderation, (4) told myself to allow room for an occasional treat food or dessert for just plain enjoyment, (5) reminded myself to think about my beverage choices, and (6) told myself that fruits and vegetables should be included in every meal.” For self-regulation for healthy meal behavior, participants indicated how often they had: (1) planned quick, easy, and healthy snacks; (2) selected beverages with health in mind; (3) purposely added vegetables to meals and snacks; and (4) been flexible and sensible in food choices. Likert scaled responses (1 = never and 5 = always) scores were summed. For this sample, Cronbach alpha was .73 for the self-instruction scale and .71 for the self-regulation scale.
Physical Activity Instrument
The authors used the International Physical Activity Questionnaire to assess amounts of physical activity performed at 3 intensity levels (ie, walking, moderate intensity, and vigorous intensity).
For each level, participants reported frequencies as days per week and average duration in minutes for each frequency over the past week. Researchers used the standard scoring protocol to convert these data into metabolic equivalents (MET-minutes) per week to generate total, walking, moderate activity, and vigorous activity scores.
to determine perceived stress level. This scale assesses the extent to which individuals considers their situations to be stressful, with coefficient alphas for reliability of .84–.86 in college students.
The items queried about how unpredictable, uncontrollable, and overloaded respondents found their lives, using a 5-point Likert scale (response format was 0 = never and 4 = very often). Scores were summed; higher scores indicated greater perceived stress. Perceived stress was a primary study outcome.
Hours of Sleep
One question from the Behavior Risk Factor Survey was used to evaluate sleep:
“Think about the time you actually spend sleeping or napping, not just the amount of sleep you think you should get. On average, how many hours of sleep do you get in a 24-hour period?” Answers were recorded as daily hours of sleep. Sleep behavior was a secondary study outcome.
Stage of Readiness to Change Assessment
At every assessment, participants completed algorithms for each of 3 target behaviors
(ie, precontemplation, contemplation/preparation, or action/maintenance). These target behaviors were: (1) consuming ≥ 5 cups of fruit and vegetables/d; (2) completing regular, planned physical activity 5 times/wk for 30 minutes; and (3) practicing daily stress management. For purposes of analysis, participants were categorized into pre-action (not ready or just thinking about the behavior change) or action (have made and/or maintaining the behavior).
Self-efficacy, perceived benefits, and weight satisfaction discriminate among stages of change for fruit and vegetable intakes for young men and women.
Stage of readiness to change was a secondary study outcome.
Process Evaluation
Per PROCEED Phase 6 (Process Evaluation), a process evaluation was conducted with experimental participants during the follow-up period via the Web portal. Participants were invited to complete an online survey to evaluate the overall program as excellent, good, satisfactory, or needing improvement. Time viewing the mini lessons was collected via the Web portal.
Data Analysis
To detect differences at alpha of .05 and power of 80%, an expected between-group difference of 0.7 kg for the primary outcome based on the pattern of weight gain of +0.6 kg/y found in the CARDIA study
an SD of 4.8 kg, and a 50% dropout rate, a minimum sample size of 1,476 was needed. Intervention assessment variables were examined for outliers and normality. Variables that were 3 SD beyond the average were considered outliers and evaluated for plausibility. Three responses (1 reporting > 6,000 kcal/d in sweetened beverage intake and 2 reporting > 15 cups of fruit and vegetables/d) were considered implausible and were not included in the analysis for these specific variables. Non-normal variables (ie, cups of fruit and vegetables, MET-minutes per week, and sweetened beverage energy intake) were log transformed before analyses.
One-way analysis of variance for continuous variables and chi-square analysis for categorical variables compared control and experimental dependent variables at baseline. To determine differences between experimental and control outcome variables, mixed models (PROC MIXED) repeated-measures analysis was performed using SAS statistical software (version 9.3, SAS, Gary, NC, 2011). Gender was included in the model to account for difference in retention rates between males and females. Significance was reported for group × time, group × time × gender, group, and time. Outcomes were determined on completers (defined as those who completed 15-month assessments). Unless otherwise noted, data are presented as means ± SD. Cohen d effect size is reported for outcomes that were statistically significant. Process evaluation data are presented as frequencies.
Results
Of the 6,277 potential participants who responded to recruitment announcements, 53% screened as eligible and approximately half consented (1,639; 824 experimental, 815 control) and were enrolled as participants (Figure).
At baseline the sample was aged 19.3 ± 1.1 years; 67% female; 72% white; and 38% first-year, 35% second-year, and 27% third-year students. The majority lived on campus (72%) and never used cigarettes (69%) or smokeless tobacco (93%). Comparing baseline stage of readiness to change goals set in this study (ie, ≥ 5 cups of fruits and vegetables daily, planned physical activity 5 times/wk for 30 minutes, and practicing daily stress management), only 12% met the fruit and vegetable goal but 82% achieved the physical activity goal (Table 1) and > 80% staged to post-action for managing stress most days of the week. Although most (68%) were within the normal BMI range, 60% of all students reported they were trying to lose weight. Approximately 20% of females and 8% of males were at risk for metabolic syndrome based on waist circumference (Table 1).
Table 1Comparison of Project Young Adults Eating and Active for Health Experimental and Control Study Participants at Baseline
P was derived from 1-way analysis of variance for continuous variables and chi-square analysis for categorical variables comparing control and experimental groups at baseline
P was derived from 1-way analysis of variance for continuous variables and chi-square analysis for categorical variable comparing individuals who completed the study through follow-up (completers) and those who did not complete all 3 study visits (non-completers)
Definition of on-campus residence includes students residing in residence halls, fraternities, or sororities and those living in other college- or university-owned units
Physical activity recommendations based on the US Department of Health and Human Services Physical Activity Guidelines Advisory Committee 2008 report suggesting 500–1,000 metabolic equivalent minutes of physical activity per week
P was derived from 1-way analysis of variance for continuous variables and chi-square analysis for categorical variable comparing individuals who completed the study through follow-up (completers) and those who did not complete all 3 study visits (non-completers)
At risk for metabolic syndrome based on waist circumference > 88 cm in female participants and >102 cm in male participants.
(%)
8.0
7.5
7.8
.60
7.7
7.9
.98
a P was derived from 1-way analysis of variance for continuous variables and chi-square analysis for categorical variables comparing control and experimental groups at baseline
b P was derived from 1-way analysis of variance for continuous variables and chi-square analysis for categorical variable comparing individuals who completed the study through follow-up (completers) and those who did not complete all 3 study visits (non-completers)
c Definition of on-campus residence includes students residing in residence halls, fraternities, or sororities and those living in other college- or university-owned units
d Physical activity recommendations based on the US Department of Health and Human Services Physical Activity Guidelines Advisory Committee 2008 report suggesting 500–1,000 metabolic equivalent minutes of physical activity per week
e At risk for metabolic syndrome based on waist circumference > 88 cm in female participants and >102 cm in male participants.
There were no differences in demographic characteristics and anthropometric measurements between experimental (n = 824) and control (n = 815) participants at baseline. Most participants (76%) completed the 10-week postintervention anthropometric assessment data and 59% completed the 15-month follow-up anthropometric assessment (Figure). When baseline characteristics of 973 participants who completed the 15-month follow-up were compared with the 666 non-completers, the only differences were that more completers (70.4% vs 60.7%) were female and had never used cigarettes (71.7% vs 65.6%).
Intervention Results
Table 2, Table 3, Table 4 present intervention results. Baseline, postintervention, and follow-up findings were compared for statistically significant differences.
Table 2Comparison of Project Young Adults Eating and Active for Health Participant Primary Study Outcomes (Anthropometric, Eating Behavior, Physical Activity, and Perceived Stress) at Baseline, Postintervention and Follow-up
Significance was set at P ≤ .05 and was determined using PROC MIXED repeated-measures analysis with the fixed effects of time, gender, and group using SAS statistical software. Different numbers within groups indicate significant differences with respect to time when group × time interaction was significant
Average of measurement taken at least twice (and repeated as needed when the variance between measurements exceeded the standard). Body mass index was calculated using the standard formula (ie, weight [kg] / height [m]2)
Average of measurement taken at least twice (and repeated as needed when the variance between measurements exceeded the standard). Body mass index was calculated using the standard formula (ie, weight [kg] / height [m]2)
Fruit and vegetable intake was assessed using the National Cancer Institute Fruit and Vegetable Screener.39 Variables were transformed before analysis. Sample means are reported
Physical activity was measured using the International Physical Activity Questionnaire.44 Variable were transformed before analysis. Sample means are reported
Perceived stress was measured with Cohen Perceived Stress Scale (14 items).45 Higher scores indicate greater perceived stress
Experimental
22.4 ± 7.2
22.8 ± 7.9
22.9 ± 7.6
.80
.65
.37
.005
Control
22.4 ± 7.1
23.2 ± 7.7
23.5 ± 8.1
MET indicates metabolic equivalents.
a Significance was set at P ≤ .05 and was determined using PROC MIXED repeated-measures analysis with the fixed effects of time, gender, and group using SAS statistical software. Different numbers within groups indicate significant differences with respect to time when group × time interaction was significant
b Average of measurement taken at least twice (and repeated as needed when the variance between measurements exceeded the standard). Body mass index was calculated using the standard formula (ie, weight [kg] / height [m]2)
c Fruit and vegetable intake was assessed using the National Cancer Institute Fruit and Vegetable Screener.
Table 3Comparison of Project Young Adults Eating and Active for Health Participant Secondary Study Outcomes (Anthropometric, Eating Behavior, Physical Activity, and Perceived Stress) at Baseline, Postintervention and Follow-up
Significance was set at P ≤ .05 and was determined using PROC MIXED repeated measures-analysis SAS software including the fixed effects of time, gender, and group. Different numbers within groups indicate significant differences with respect to time when group × time interaction was significant
Sugar-sweetened beverage intake was assessed with 8 questions adapted from West et al.42 Participants were asked how often in the past month (never or < 1/mo to ≥ 4/d) and what amount for sugar-sweetened soft drinks (none, 12-oz can, restaurant glass or cup, 20-oz bottle, or 2-L bottle), fruit drinks (none, ≤ 11.5-oz can, 20-oz bottle, or 64-oz bottle), non-diet energy drinks (none, 2- to 6-oz shot, between 6 and 16 oz, or > 16 oz), and sugar-sweetened specialty coffee drinks (none, < 12 oz, or > 12 oz). Average number of kilocalories per day was calculated by converting frequency and amount to ounces per day and multiplying by respective kilocalories per ounce. Variables were log transformed for analysis. Sample means are reported
Whole grain intake was assessed using the question, “How many servings of whole grains do you eat on average per day?” Answers ranged from “< 1” to “≥ 6”
Experimental
2.1 ± 1.4
2.2 ± 1.4
2.3 ± 1.4
.13
.47
.73
.08
Control
2.2 ± 1.5
2.2 ± 1.5
2.2 ± 1.4
Self-instruction for intention for healthful mealtime behavior
Six questions for self-instruction for intention for healthful meal behavior and 4 questions for self-regulation for reported healthful meal behavior were adapted from Strong et al.43 Reponses were on a scale of 1 to 5, in which 1 = “never” and 5 = “always”. Scale scores were summed, with higher scores indicating greater intention and behavior
Significant differences between experimental and control group at respective time periods.
7.2 ± 1.1
.05
.30
.19
.07
Control
7.8 ± 3.71
6.9 ± 1.12
7.1 ± 1.13
a Significance was set at P ≤ .05 and was determined using PROC MIXED repeated measures-analysis SAS software including the fixed effects of time, gender, and group. Different numbers within groups indicate significant differences with respect to time when group × time interaction was significant
b Fat intake was estimated using the National Cancer Institute Fat Screener;
Participants were asked how often in the past month (never or < 1/mo to ≥ 4/d) and what amount for sugar-sweetened soft drinks (none, 12-oz can, restaurant glass or cup, 20-oz bottle, or 2-L bottle), fruit drinks (none, ≤ 11.5-oz can, 20-oz bottle, or 64-oz bottle), non-diet energy drinks (none, 2- to 6-oz shot, between 6 and 16 oz, or > 16 oz), and sugar-sweetened specialty coffee drinks (none, < 12 oz, or > 12 oz). Average number of kilocalories per day was calculated by converting frequency and amount to ounces per day and multiplying by respective kilocalories per ounce. Variables were log transformed for analysis. Sample means are reported
d Whole grain intake was assessed using the question, “How many servings of whole grains do you eat on average per day?” Answers ranged from “< 1” to “≥ 6”
e Six questions for self-instruction for intention for healthful meal behavior and 4 questions for self-regulation for reported healthful meal behavior were adapted from Strong et al.
Reponses were on a scale of 1 to 5, in which 1 = “never” and 5 = “always”. Scale scores were summed, with higher scores indicating greater intention and behavior
f Sleep patterns were assessed based on the question, “On average, how many hours of sleep do you get in a 24-hour period?”
∗ Significant differences between experimental and control group at respective time periods.
Differences between total number reported and number completed (experimental n = 497; control n = 476) equals missing values for respective behavior.
P
n
%
n
%
n
n
%
n
%
n
n
%
n
%
n
Fruit and vegetable intake
Experimental
217
43.8
278
56.0
495
.19
147
31.8
316
68.2
463
.02
168
35.3
309
64.9
476
.92
Control
185
39.1
288
60.5
473
160
37.6
266
62.5
426
163
35.4
297
64.6
460
Physical activity
Experimental
323
65.3
172
34.7
495
.52
211
45.6
252
54.4
463
.002
228
47.8
249
52.0
477
.09
Control
302
63.9
171
36.1
473
235
55.2
191
44.8
426
252
54.8
208
45.2
460
Stress management
Experimental
81
16.4
413
83.6
494
.06
71
15.3
392
84.7
463
.10
77
16.1
400
83.9
477
.91
Control
53
11.2
420
88.8
473
68
16.0
358
84.0
426
76
16.5
384
83.5
460
Note: Movement is expressed as the percentage within each group at each assessment period. Participants (who completed) movements from pre-action (not ready or just thinking about the behavior change) to post-action (have made and/or maintaining the behavior) in staging within respective behavior were evaluated using chi-square analysis.
a Differences between total number reported and number completed (experimental n = 497; control n = 476) equals missing values for respective behavior.
Analysis of intervention results indicated there were no differences between experimental and control participants in the primary anthropometric outcomes of BMI and weight (Table 2) or secondary outcome of waist circumference (Table 3) at postintervention or follow-up. First-year students gained significantly more weight than non–first year students over the 15 months of the intervention (0.73 ± 0.18 vs 0.29 ± 0.11 kg; P = .04); however, there was no difference in weight gain between experimental and control first-year participants.
Eating Behavior
For the primary dietary outcome of cups of fruits and vegetables consumed daily, there was a significant group × time interaction and a significant within-group effect for time. The only between-group difference occurred at postintervention. Experimental participants reported small but statistically significant increases in cups of total fruits and vegetable from baseline (2.6 ± 2.1 cups) to postintervention (2.8 ± 2.1 cups), whereas the control participants decreased cups of fruits and vegetables from baseline (2.7 ± 1.9 cups) to postintervention (2.5 ± 2.1 cups), resulting in a significant group difference with a small effect size (Cohen d = 0.05) between experimental and control group at postintervention. These between-group differences were not maintained at the 15-month follow-up (Table 2).
There was a significant group × time difference in the secondary dietary outcome of percentage of energy from fat but there were no differences in sugar-sweetened beverage consumption and servings of whole grains. Experimental participants decreased fat intake from baseline (31.3% ± 5.2% of kilocalories) to postintervention (30.4% ± 4.4%) and maintained the change through follow-up (30.5% ± 4.1%), whereas control participants did not change. The effect size at both postintervention and follow-up was small (Cohen d = 0.14). There was a significant group × time and a significant within-group effect for time for self-instruction for intention for healthful mealtime behavior and self-regulation for reported healthful mealtime behavior. Experimental participants reported higher scores in both self-instruction (3.6 ± 0.7 vs 3.4 ± 0.8; Cohen d = 0.28) and self-regulation (3.6 ± 0.7 vs 3.5 ± 0.7; Cohen d = 0.18) than control participants at postintervention (Table 3).
Physical Activity
There were no differences between experimental and control participants in total MET-minutes per week. However, when evaluating changes by intensity of activity, there was a significant time effect for moderate MET-minutes per week. Both groups increased in moderate MET-minutes per week from baseline to postintervention. There was a significant group × time × gender effect in vigorous MET-minutes per week (Table 2). Females in the experimental group had a significantly higher albeit small effect size (Cohen d = 0.2) in vigorous MET-minutes per week at postintervention than female control participants. These differences were not maintained through follow-up. There were differences at postintervention in the males; experimental males reported significantly fewer vigorous Met-minutes at postintervention than control males. Similar to females, the differences were not maintained through follow-up (Table 2).
Stress
There were no differences in perceived stress (Table 2). Perceived stress scores ranged from 22.4 to 23.9 on a scale of 0 to 56, with higher scores indicating greater stress.
Hours of Sleep
There was a significant group × time interaction for sleep; experimental participants reported a higher number of hours of sleep than controls at postintervention (7.3 ± 1.1 vs 6.9 ± 1.1; Cohen d = 0.4) (Table 3). Experimental participants maintained hours of sleep and reported no changes in hours of sleep from baseline to postintervention and follow-up. Control participants reported sleeping fewer hours at both postintervention and follow-up than at baseline.
Stage of Readiness to Change
There was no difference between groups in stage of readiness to change for any behavior at baseline. At postintervention, a significantly greater proportion of experimental participants were in the post-action stages than control participants for fruit and vegetable intake (P = .02) and physical activity (P = .002) but differences were not maintained at follow-up (Table 4). There were no differences between groups for stress management at any time point.
Process Evaluation
Of the 825 experimental participants, 471 (57%) completed the process evaluation postintervention. Most (87%) evaluated the program as good or excellent. When restricting this sample to those who viewed at least 1 lesson in each of the 3 main topic categories (n = 402), 75% were “at least moderately motivated” by the Eating Lessons, along with 74% by the Physical Activity Lessons and 57% by the Stress Management Lessons. Although lesson completion was a voluntary activity, 70% of experimental participants (n = 579) accessed the lessons during the intervention period and 252 participants (31%) spent ≥ 30 minutes viewing the lessons. As planned, there was no lesson access for control participants.
Discussion
Project YEAH is one of the first studies to use the CBPR process of PRECEDE-PROCEED to develop a non-diet approach intervention reinforcing behaviors that prevent excessive weight gain in the young adult, college population. Work and outcomes from a non-college vocational education population are described elsewhere.
Extensive work was conducted via the PRECEDE Phases to determine quality of life issues and respective predisposing, reinforcing, and enabling factors that needed to be addressed in the intervention to foster healthful behaviors supporting weight maintenance.
Furthermore, the intervention content was individually tailored by the Transtheoretical Model and educational materials developed using Dick and Carey's24 Model for Instructional Design to enhance the attention, relevance, confidence, and satisfaction of the intervention. Although there were no differences between experimental and control participants in Project YEAH in weight change or BMI, the intervention supported positive change in behaviors that may mediate excessive weight gain, such as increasing fruit and vegetable intake and more healthful self-regulation mealtime behaviors.
Experimental group participants reported a small but significant increase in fruit and vegetable intake during the intervention period, whereas control participants decreased their intake. Other programs, including previous work from the authors' group,
reported similar outcomes in that they did not influence weight but affected health behaviors such as fruit and vegetable intake, which may influence long-term weight maintenance.
The majority of participants in Project YEAH (68%) were at a normal weight, similar to the recent American College Health Association National College Health Assessment
population in which 62% reported normal weight. Although participants in this study tended to be of normal weight, National Health and Nutrition Examination Survey 2009–2010
reported that 67% of males and 56% of females in the 20- to 39-year age range had a BMI ≥ 25 kg/m2. Weight gain is still a concern because the first-year students in YEAH gained significantly more weight than the second-year students.
Web-based programs have successfully increased physical activity. Richardson et al
Web-enabled feedback control over energy balance promotes an increase in physical activity and a reduction of body weight and disease risk in overweight sedentary adults.
reported that Web-enabled feedback not only increased physical activity but also decreased body weight in sedentary adults. In contrast, an increase in physical activity behavior was not reported in Project YEAH, which may be partly because of a ceiling affect, as the YEAH participants were relatively active at baseline compared with national standards. Approximately 80% of YEAH participants self-reported meeting the recommended physical activity guidelines of 150 minutes of moderate physical activity per day
and Kraushaar and Kramer targeted weight reduction in sedentary adults, whereas Project YEAH aimed to prevent excessive weight gain among college students, most of whom walk around campus to go to classes and other activities.
In the formative phases of the work to develop the Project YEAH, college students had identified stress from lack of finances and time to manage daily activities and schoolwork as a factor that affected quality of life,
and this stress was perceived as a barrier to healthful behavior. In contrast, when the college students in this intervention were queried about perceived stressed, they reported relative satisfaction with life and had relatively low perceived stress scores (22.4–23.5) similar to those reported by other student samples.
Furthermore, > 80% of YEAH participants reported being in post-action stages for managing stress on most days of the week at baseline, which could also account for the lack of change in perceived stress scores.
Because lack of sleep and or the ability to manage schedules can interfere with obtaining adequate sleep, amount of sleep was used as a secondary indicator of stress management. Project YEAH included lessons on the importance of getting enough sleep and strategies on how to manage time to get enough sleep. Experimental participants were able to maintain sleep hours whereas the control group reported a decrease in the hours of sleep during from baseline to postintervention. This difference in behavior may be important when considered in the context of the academic calendar. Baseline sleep measures occurred at the beginning of the semester as students returned from winter break, whereas the postintervention and follow-up measures both occurred near the semesters' end. Increased end-of-semester academic demands may result in college students getting less sleep. Both groups reported a low average amount of sleep, however: closer to 7 than the 9 recommended hours per day and less than amounts reported by other researchers.
The YEAH experimental group participants reported greater self-instruction for intention to plan meals and snacks and self-regulation in execution of meals and snacks postintervention than control participants. Project YEAH eating lessons and corresponding nudges may have accounted for these differences in intention to and/or execution of healthful mealtime behaviors. Eating lessons focused on activities to support healthful mealtime behaviors, such as planning for regular meals and snacks, shopping for food, budgeting, assembling meals and snacks, and choosing foods in campus food venues. These changes in mealtime behavior are important for weight maintenance. The ability for young adults to manage eating regular meals with recommended amounts of fruits and vegetables has been associated with protection from obesity.
reported a positive association between preparing their own meals and consumption of fruits and vegetables among college students. In additionally, Strong et al
reported that college students who planned and tracked meals and snacks had lower energy intake, greater fruit and vegetable consumption, and less energy from added sugars.
Results indicate that the intervention was moderately successful; a greater proportion of experimental than control participants were in post-action stages at postintervention for consuming ≥ 5 cups/d of fruit and vegetable as well as completing 150 min/wk of physical activity, but there was no significant difference between groups for perceived stress. These differences were not maintained at follow-up. Although theory-based, Web-delivered programs offer promise to support behavior changes for obesity prevention, and Transtheoretical Model tailored interventions have been effective,
Like all studies, this one has limitations, most of which are unavoidable in human studies. Limitations include the self-selected sample, attrition rates, and self-reported measures of eating behaviors, physical activity, and perceived stress. Although participants in this study were self-selected, the baseline characteristics are similar to the spring, 2013 American College Health Association–National College Health Assessment II (ACHA-NCHA II)
sample in gender and consumption of fruit and vegetables. The spring, 2013 ACHA sample was 65% female vs 67% in YEAH. Six percent of participants in spring, 2013 ACHA reported consuming ≥ 5 cups of fruit and vegetable vs 11% in YEAH. The living arrangements differed in that 72% of YEAH participants reported living on campus vs 37.3% of ACHA-NCHA II respondents and approximately 35% of general student population from participating universities. The higher percentage of students living on campus in Project YEAH is more than likely the result of the recruitment protocol targeting first- and second-year students. The retention rate of 59% of participants is slightly less than that of 67% in previous work with college students,
but there was still sufficient size per the power calculation. The healthful behavior enhancements were small and limited, but the geographically dispersed college students enhanced the generalizability of the intervention.
Implications for Research and Practice
Project YEAH used an Internet delivery mode to reach young adult college students. It was built on best practices, using the ingredients for success reported by others. For instance, successful interventions are theory-based and have components of tailoring the intervention to the participant, goal setting, and some reminders or messaging.
Interventions to promote physical activity and dietary lifestyle changes for cardiovascular risk factor reduction in adults: a scientific statement from the American Heart Association.
Project YEAH developers extensively used theory to ensure that the lesson activities caught and held the attention of the target audience and motivated participants to learn.
Model for Instructional Design because use of these models improves the quality of the educational lessons and acceptability of the product to the learner.
Indeed, participants indicated that the YEAH lesson activities were interesting and relevant, and nearly three-quarters of experimental participants accessed and viewed the lessons even though this was not a requirement of the study. Furthermore, interventions that use participatory research strategies are more effective and sustainable than a research only–driven intervention.
to guide the development of Project YEAH and included extensive target audience input throughout the development and delivery of the intervention. The PRECEDE-PROCEED process
Project YEAH used the ingredients of best practices with fidelity; however, study findings make it clear that there is much more to learn about the recipe for using these ingredients to create interventions that successfully and substantially facilitate behavior changes that promote healthy weights. Project YEAH has the potential to provide basic health and wellness information for the college student, particularly the incoming freshman.
Acknowledgment
This project was supported by National Research Initiative Grant (2009-55215-05460) from the US Department of Agriculture National Institute for Food and Agriculture. The authors acknowledge Emily Hansen, MS, and Ren Curiong, PhD, for assistance with data management and statistical analysis, and the multiple graduate students from each institution for data collection.
References
Ogden C.L.
Carroll M.D.
Kit B.K.
Flegal K.M.
Prevalence of obesity and trends in body mass index among US children and adolescents, 1999-2010.
Effects of a 16-month randomized controlled exercise trial on body weight and composition in young, overweight men and women: the Midwest Exercise Trial.
Anderson J. Millennials will make online sharing in networks a lifelong habit. Pew Internet & American Life Project. http://www.pewinternet.org/. Accessed July 9, 2014.
Self-efficacy, perceived benefits, and weight satisfaction discriminate among stages of change for fruit and vegetable intakes for young men and women.
Web-enabled feedback control over energy balance promotes an increase in physical activity and a reduction of body weight and disease risk in overweight sedentary adults.
Interventions to promote physical activity and dietary lifestyle changes for cardiovascular risk factor reduction in adults: a scientific statement from the American Heart Association.
00637-X&id=gr1.jpg){kind=link}