Critical Review of NC-219:
NC219 Accomplishments and Progress Summary
The North Central USDA/CREES Regional Research Project, which began October 1, 1996, is a collaboration of nutrition education scientists from thirteen states including Alabama, Arizona, Iowa, Kansas, Maine, Michigan, Nebraska, New York, Rhode Island, Ohio, Oregon, South Dakota, and Wisconsin. The project, entitled "Using Stages of Change model to promote consumption of grains, vegetables and fruits by young adults" was guided by two objectives. The first objective was to identify young adults' readiness to change food habits associated with consumption of grain products, vegetables and fruits. The second objective was to identify messages and recommend strategies that will promote young adults' progress toward increased consumption of grain products, vegetables and fruits. In completing objective one, i.e. identifying the readiness to change, instruments to assess stage of change, decisional balance, and self-efficacy were developed, tested for validity and reliability, and evaluated as to their appropriateness to serve as a cognitive marker for assignment to and evaluation of an educational intervention. In addition, a food frequency questionnaire was developed and validated. A survey using these instruments was completed by 1545 young adults in ten different states. This multi-step process was completed in years one through four (1996 - 2000). Identifying messages and recommending strategies (objective two) constitutes the focus of the fifth research year (2000-2001).
Instrument Development
Instruments were developed to assess stage of change, decisional balance and self efficacy as well as a food frequency. The stage of change algorithm has been described previously (see page 5).
Fruit and Vegetable Intake. The food frequency questionnaire was originally based on 1994 and 1995 CSFII fruit and vegetable intake data for 19 to 24 year olds. The survey was revised based on pre-testing and pilot studies with young adults and low-income clients. The resulting instrument was tested for criterion validity by comparing findings with 24-hour recall data and food frequency questionnaire responses. Two semi-quantitative food frequency questionnaires (FFQs), one for fruit and one for vegetables, were developed, then revised after the pilot test. The final instrument consists of 12 fruit items and 14 vegetable items. Frequency of consumption was measured on an eight-level rating scale from one time per month or less to twice per day or more. Three options of serving sizes (i.e. small, medium, and large) were provided. Most foods listed a medium serving as comparable to a Food Guide Pyramid serving (49) except for a few that specified an amount more comparable to usually consumed portion sizes.
Self-Efficacy and Decisional Balance. Twenty self-efficacy
items and 63 decisional balance (pros and cons) items were first
generated as a result of literature review and in-depth individual
interviews with convenience samples of young adults. Items were
deleted or revised through pilot-tests according to results from
principal component analysis, inter-item correlation coefficients,
and Cronbach's a coefficients. A final five-item per food group
(10 items total), five-point Likert scale, ranging from very confident
to not confident at all, was developed to assess self-efficacy
to perform tasks regarding the consumption of fruits and vegetables.
Cronbach's a coefficients of the self-efficacy scale were 0.86
for fruit and 0.85 for vegetables. The final decisional balance
scale consisted of 10 con and 8 pro items per food group (36 items
total). Respondents rated how important each item was in their
decision making for consuming fruits and vegetables. The five-point
importance scale ranged from very important to not at all important.
The overall decisional balance scale had a Cronbach's a coefficient
of 0.80 for fruit and 0.79 for vegetables; the alpha coefficients
of the pro and con items were 0.73 and 0.72 for fruit and 0.73
and 0.70 for vegetables, respectively.
Once initial instrument development was completed, instruments
were validated in a pilot test utilizing a telephone interview
followed by a mail survey. The telephone interview was completed
by 209 young adults. Subjects were interviewed about one specific
food group: fruits (N=47), vegetables (N=44) or grains (N=118).
Subjects were asked about perceived number of servings of the
food group, then completed a food group specific 24 hour recall
followed by adjustment based on usual food habits. Adjusted servings
were used for stage classification (if ³ criterion how long
for Action/Maintenance; if < criterion motivational readiness
to meet the criterion was assessed (e.g., you are currently eating
two servings a day of vegetables, do you intend to increase to
three servings) for Precontemplation, Contemplation, and Preparation.
Proportion by stage is listed in the table below.
Table 2: Proportion by stage by food group based on the interview (N=209)
| Food Group | Precontemplation | Contemplation | Preparation | Action | Maintenance |
| Fruit | 13% | 2% | 17% | 17% | 51% |
| Vegetables | 14% | 9% | 14% | 2% | 61% |
| Grains | 36% | 4% | 8.5% | 3% | 47.5% |
Table 3. Proportion by stage1 and servings by food group (N=141)
|
|
|
|
|
||||||
| % | Mean2 | Median3 | % | Mean | Median | % | Mean | Median | |
| Precontemplation | 10% | 1.8±1.9a | 1.31 | 31% | 2.61±1.8 a | 1.94 | 28% | 5.27±2.5a | 4.76 |
| Contemplation | 17% | 2.5±2.6ab | 2.31 | 24% | 2.84±1.7 ab | 2.55 | 14% | 4.79±2.4a | 4.79 |
| Preparation | 7% | 2.6±1.5ab | 1.88 | 9 % | 2.60±1.9 ab | 1.81 | 7% | 5.80±1.5ab | 5.23 |
| Action | 12% | 6.1±3.1bc | 3.92 | 4% | 5.43±0.9bc | 5.22 | 5% | 4.98±1.8ab | 5.36 |
| Maintenance | 53% | 6.8±3.7c | 4.62 | 32% | 4.80±2.2c | 4.87 | 46% | 7.78±4.2b | 6.87 |
1. Proportion by stage based on the survey (servings self-assessed
for stage)
2. Mean±SD intake as determined by FFQ
3. Median intake as determined by FFQ
abc Within each food group, difference superscripts are significantly
different at p<.05
Proportions by stage follow similar patterns to those found in the interview, lending support to the validity of the stage classification system. The one exception was vegetables which is consistent with the underestimation in Maintenance found in the interview. In addition, the number of servings from the survey FFQ supported the validity of the self-assessment system for stage with the exception of grains. However, the large number of servings of fruit for those in Action and Maintenance suggested that the FFQ needed to be modified (this was done for the 4th year survey).
Survey
Research participants were from 10 states (Al, KS, ME, MI, NE, NY, RI, OR, SD, and WI). All states collaborated to develop and follow the research protocol. Each state purchased a random name list of 3500 young adults between 18 to 24 y, except for Nebraska (19-24 y). Up to four attempts were made to reach each selected person by phone by trained interviewers from the same age group. Of the 3570 young adults contacted, 640 (14%) refused to participate. A total of 2930 agreed to participate, completed a brief screening interview, and received questionnaires by mail. Reminder postcards and a second survey were mailed 10 and 20 days, respectively, after the initial mailing to maximize response rate. An incentive drawing of $25.00 per 25 respondents was completed in each state. In all, 1545 usable surveys were returned with an overall response rate of 53%. Response rates for phone recruitment ranged from 70% in WI to 96% in SD, while mailed survey return rates ranged from 35% in MI to 69% in OR. Compared with respondents to the mail survey, non-respondents were more likely to be male and in the Precontemplation stage. Stage distribution has been described previously (page 6), as was the strong effect of self-efficacy on meeting the guidelines (page 7). Decisional balance was also found to be a strong predictor of behavior; Precontemplators and Contemplators had higher cons than pros with the crossover between pros and cons occurring in Preparation. Pros were higher than cons in Action and Maintenance. The following figure describes the servings by stage for the food groups. The problem with fruits (excessive number of servings in Action and Maintenance) has been resolved. It is clear that self-assessed servings used for the stage classification is validated by the FFQ (< Guidelines for Precontemplation, Contemplation and Preparation, ³ Guidelines for Action and Maintenance).
Overall, the research confirmed the applicability of the Stages of Change model for fruit and vegetable consumption by young adults and demonstrated the importance of the decisional balance and self-efficacy constructs. Valid instruments measuring stage of change, decisional balance, self-efficacy and dietary behavior have been developed. However, because the decisional balance instrument is 36 items, a short form of the instrument is being developed by RI as its' 5th year project in order to reduce subject burden.
Identification of Messages and Recommendation of Intervention Strategies
Knowledge of young adults' fruit and vegetable intake profiles
supported the need to identify messages and recommend strategies
to promote increased consumption of these foods. The availability
of a valid and reliable tool to assess stage of change operationalized
the objective. The fifth year of this regional project is using
focus group research to test the acceptability of stage-tailored
newsletters using messages developed by this project. This will
enable the current project to develop recommendations about intervention
strategies. The proposed renewal will test intervention strategies.
Newsletters, adapted by RI from previous materials, are brief,
2 - 4 page newsletters with stage matched messages designed to
increase motivational readiness to increase intake (Precontemplation),
to increase intake to meet the criterion (Preparation), and to
increase amount and variety of intake (Maintenance). Material
is presented in column style with color graphics and uses 14-point
type. Two issues for each stage have been be modified by the 5th
Year Committee and are being evaluated. These six newsletters
were be modified as follows: 1) tailored using messages identified
by the current project 2) tailored for states (logo on masthead
and state specific resources included), 3) a clear theme per issue
has been inserted, 4) a story has been included, and 5) language
was be modified to be appropriate for young adults. A specific
set of questions has been developed including questions assessing
preferred method of delivery of nutrition information. Focus group
leaders have been trained and are assisted by a recorder to tape
record the groups. Twelve focus groups are being be conducted
with college students (four groups per stage). In addition 3 focus
groups (one per stage) are being held with non-college young adults.
Focus groups are being transcribed and the leader and recorder
are informally summarizing data about the preferred delivery method.
Transcripts will be analyzed.
Conclusion
The initial objective to identify stage of motivational readiness to change in young adults for fruits and vegetables was accomplished. However the project was unable to develop a practical instrument for grains. The project decided during the annual Meeting in 1999 to discontinue investigation of grains and to focus on fruits and vegetables. Valid instruments measuring stage of change, self-efficacy, decisional balance, and behavior were developed. A survey was completed describing stage of change, self-efficacy, decisional balance, and behavior in a representative sample of 1545 young adults. The second objective to identify messages and recommend strategies is currently being completed. Prototype materials have been developed and are being evaluated in focus groups during the fifth year of the project. The renewal of this regional project develop and test stage tailored interventions for young adults.