Dietary fat reduction strategies used by a group of adults aged 50 years and older – Statistical Data Included
Rayane Abusabha
ABSTRACT
Objective To investigate the fat-reduction strategies used by a group of older adults who successfully made and maintained positive dietary changes for 5 years or longer.
Design Participants completed 2 copies of a self-administered food frequency questionnaire: The first copy assessed diet before they began making changes and the second copy assessed diet after initiation of healthful dietary changes. Positive food changes were identified from the food frequency questionnaires. During in-person interviews, participants placed food changes onto a time line according to the nearest estimated date of initiation of the change.
Subjects Participants were 65 free-living older adults (aged [greater than]50 years) who had maintained substantial changes to decrease fat intake in their diet for at least 5 years.
Statistical analyses performed Quantitative and qualitative data were used to identify the fat-reduction strategies and to confirm and validate the fat-reduction strategy model. Confirmatory factor analysis was performed to confirm the new model. The Kuder-Richardson-20 reliability coefficient (kr) was used to determine internal consistency of the scales developed for the study.
Results The majority of participants decreased their fat intake gradually, at different time points in their lives, and over a long period of time (5 to 43 years). Mean percent energy intake from fat decreased from 44.3[+ or -]5.9% before dietary improvement to 25.9[+ or -]7.1% at the time of the study. The final model consisted of 5 fat-reduction strategies with 63 food changes. The strategies were: increase summer fruits (4 items; kr=0.66) , increase vegetables and grains (14 items; kr=0.79), decrease recreational foods (14 items; kr=0.76), decrease cooking fat (20 items; kr=0.86), and use fat-modified foods (11 items; kr=0.80).
Applications/conclusions Dietetics professionals should base their advice on the dietary strategies used by consumers rather than hypothetical premises such as food or nutrient groupings. Nutrition education interventions will have better chances for success if they are based on a set of customized programs that guide appropriate consumer segments through a series of small, comfortable, and sustainable dietary changes over a prolonged period of time. J Am Diet Assoc. 2001;101:1024-1050.
Epidemiological, clinical, and animal studies suggest that the adoption of diets low in fat and high in fiber are essential for decreasing the risk of chronic diseases such as heart disease and cancer [1-4]. Indeed, a major objective of the Healthy People 2010 initiative [5] is to reduce Americans’ average fat intake to 30% or less of total energy. For most Americans [6], this means making considerable dietary changes.
Many interventions have been conducted to help consumers decrease their fat intake [7-13]. When short-term success is considered, participants in most interventions have not been able to reach desired goals, like reaching fat intake of [less than]30% of energy [9,12]. In addition, with only individual exceptions [12], these interventions are generally unsuccessful in achieving long-term maintenance of the desired dietary behavior [9-14]. Devising successful and inexpensive dietary interventions is a challenge yet to be met, and it is evident that dietary change strategies that are effective for promoting the adoption and maintenance of desirable dietary behavior need to be discovered [9,15].
The wide ranges in dietary fat intake reports by different population groups [6] and the decline in fat intake during the past decades [1,6,16] indicate that there are persons who have successfully followed dietary recommendations and consume healthful diets. In-depth studies of the case histories of successful changers may provide dietetics professionals with dietary change strategies that could then be included and tested in nutrition education interventions designed to help other consumers make dietary changes. Therefore, the purpose of this project was to investigate the strategies used by a group of older adults ([greater than]50 years old) who made and maintained positive dietary changes for 5 years or longer.
METHODS
This research was a retrospective study that assessed dietary behavior change patterns using a combination of qualitative and quantitative methods. The research methods and procedures were previously tested for validity and reliability, and used in a different population [15,17]. All procedures were approved by a Committee for the Protection of Human Subjects at The Pennsylvania State University, University Park.
Recruitment and Subject Selection
Participants were recruited through radio and television announcements and newspaper advertisements. To be eligible, participants had to have lived in the United States for the past 5 years, be aged 50 years or older, and believe they had been making substantial dietary changes to decrease fat in their diet for at least 5 years. A telephone screening interview was administered to all interested persons to ascertain study eligibility. Participants were offered a monetary reward for completion of the study.
Data Collection
Eligible participants were mailed a packet of questionnaires that asked about demographic information, food preferences before and after diet improvement began, major life events that occurred in the past 15 years (assessed using the PERI Life Events Scale [18]), information on general motivation for change, and other relevant questions including some regarding exercise habits and alcohol consumption. Two identical copies of a food frequency questionnaire (FFQ) [19] were also included in the mailed packets. The FFQ was designed specifically to assess usual fat consumption. On 1 copy (post FFQ) participants assessed current food consumption patterns, that is, after initiation of healthful dietary changes. On the second copy (pre FFQ), participants assessed their food consumption patterns before the initiation of dietary improvement efforts. Participants were asked to set a date when they believed they began making dietary changes to decrease fat intake. This date was used as the reference point for recalling p ast diet.
The FFQs were used to determine what positive food changes were undertaken by participants to decrease fat intake in their diet, and to confirm whether the changes reported resulted in an actual decrease in fat intake. Only positive, sustained dietary changes resulting in decreased fat intake were investigated. In addition to behaviors to decrease fat intake, behaviors to increase fruit, vegetable, and grain intake were also examined as these behaviors have also been recognized as strategies used to decrease dietary fat intake [15,20]. A food change was considered to be positive if it moved in a favorable direction at least 2 frequency categories from pre to post FFQ. The frequency categories were: [less than]1 time per month/never, 2 to 3 times per month, 1 to 2 times per week, 3 to 4 times per week, 5 to 6 times per week, once a day, and [greater than]2 times per day. Therefore, a decrease in intake of hot dogs by 3 frequency categories (eg, from 3 to 4 times per week to [less than]1 time per month/never), and an increase in intake of broccoli by 2 frequency categories (eg, from 2 to 3 times per month to 3 to 4 times per week) were both considered positive changes. Movement of at least 2 frequency categories was used to decrease the likelihood of response error.
After a participant returned the questionnaire packet, positive food changes were identified from the pre and post FFQs. A long (eg, several feet) paper time line was then prepared by recording, on the left-hand side, all the positive food changes derived from the FFQs. Next, a horizontal line was drawn, indicating the number of years the participant reported making dietary changes. Equal amounts of space were left for each change year, and each month was labeled appropriately. Finally, major life events as reported on the PERI Life Events Scale were entered above the horizontal line to correspond with when they occurred. After construction of the time lines, all FFQs were mailed to the Fred Hutchinson Cancer Research Center, Seattle, Wash, for analysis.
In-person Interviews
After all survey forms were returned and the time lines were constructed, in-person interviews were scheduled. The audio-taped interviews provided qualitative data to describe changes and the context in which the changes occurred. During the interview, the participant and the interviewer worked together to fill in the time line with additional pertinent life events used as cues for deeper recall. These life events were often associated with specific dietary changes in terms of timing, if not causation. Positive food changes from the left-hand side of the timeline were added at the nearest estimated date of their initiation, As each dietary change was placed on the time line, the interviewer probed to determine additional information pertaining to the reason(s) why the participant made that particular dietary change, or group of changes, at that specific point in time. Dietary changes made at the same time and for the same reason were so marked and referred to as change clusters. Participants were encouraged t o review and correct their initial food frequency responses during the interview whenever they deemed necessary.
DATA ANALYSIS
The data analysis used in this research is unique in that qualitative data were used to identify fat-reduction strategies and to confirm and validate the fat-reduction strategy model. All audio tapes were transcribed and reviewed by the investigators. Descriptive statistics were run and Pearson correlation coefficients were calculated to detect relationships among dietary pattern variables (eg, change clusters, change in percent energy from fat). Using the pre and post FFQs, a data set of the sustained positive food changes was created. Factor analysis [21] was performed to confirm the structure previously established by Keenan et al [15]; modify the structure using findings from previous research [20,22] and from the qualitative interview data, if necessary; and confirm the new modified structure. The confirmatory factor analysis and the correlation matrixes calculated to estimate the fit of the model to the data were run in LISREL (version 7.16, 1989, Scientific Software, Mooresville, Ind). The Kuder-Richar dson-20 reliability coefficient was used to determine internal consistency for scales developed by confirmatory factor analysis. Nutrient analysis of the FFQs to determine dietary fat decrease was performed by means of the Nutrient Data System (release 2.4, 1992, Nutrition Coordinating Center, University of Minnesota, Minneapolis) at the Fred Hutchinson Cancer Research Center.
Assessment of Fit of the Model
In a confirmatory factor analysis model where specific hypotheses are being tested, several criteria need to be met to establish goodness-of-fit: the goodness-of-fit [X.sup.2] should be less than twice the degrees of freedom, the goodness-of-fit index should be [greater than]0.9 [23], the root-mean-square residual should be [less than]0.1, all factor loadings should be significant at the 0.05 level (presented by T values [greater than]2), factor intercorrelations should be low to moderate, and modification indexes should be low (for this study we aimed for modification indices [less than]10) [21]. A high modification index for an item exhibited on a factor suggests that the item is loading high on that factor. For example, if tofu was placed in the decrease cooking fat strategy, but exhibited a high modification index on the increase vegetable strategy, tofu might have fit better in the vegetable factor.
RESULTS
Of the 125 persons who responded to the advertisements, 78 met study criteria and were mailed the survey packets. Of all eligible participants, 13 (8 women and 5 men; mean age 61.8[+ or -]7.3 years) failed to return the survey packets and withdrew from the study. The remaining 65 persons completed the study. The sample consisted of more women than men (60% vs 40%). Participants were generally well educated and mostly white (Table 1).
FFQ analyses indicated that all participants had decreased their fat intake by [greater than]5% of total energy. Mean percent energy intake from fat decreased from 44.3[+ or -]5.9% before dietary improvement efforts to 25.9[+ or -]7.1% for current diet (Table 2). The length of time participants reported adopting dietary changes to decrease fat intake ranged from 5 to 43 years. Participants made changes in clusters averaging 10 foods per cluster. Only 1 participant reported adopting all her food changes (n=50) at 1 time point (ie, in 1 cluster). The other 64 made dietary changes at different time points (ie, in at least 3 or more clusters).
Correlation analyses indicated no relationships between gender and the total number of food changes made, the total number of change clusters, and the change in percent energy from fat. Similarly, significant differences in age were not observed in this sample. The total number of food changes adopted was significantly correlated with change in percent energy from fat (r=0.62; P[less than].01). Thus, the more food changes a participant made the higher the decrease in his or her total fat intake.
From the FFQs, 129 food changes to decrease fat intake were identified. However, qualitative analysis indicated that certain food changes should not be used. For example, the food items grouped under “shellfish” (shrimp, lobster, crab, and oysters) were viewed differently depending on the participant. Many participants believed these foods were part of a healthful diet and increased their intake, whereas others indicated these foods were high in cholesterol and eliminated them from their diet. Other food changes deleted due to similar inconsistencies included consumption of dark fish, white breads and bagels, cereals, and milk.
After removal of uninterpretable items, 120 potential food changes remained. Of those, 23 were removed from analysis because they were adopted by [less than]15% of the sample. Examples include increase corn, increase baked potatoes, and decrease lunch meat. Finally, increase spaghetti with tomato sauce was also removed from analysis because it correlated highly with decrease spaghetti with meat sauce.
The remaining 96 food changes were used in the factor analysis to identify the fat-reduction strategies used by this sample. A fat-reduction strategy consisted of food changes that were likely to have been adopted by a group of people, regardless of time. For example, if the food changes increase broccoli and increase cabbage belonged to the same fat-reduction strategy, then a person who increased her broccoli intake is very likely to have increased her cabbage intake at some point in time.
First, the fat-reduction strategies model previously established by Keenan et al [15] in a relatively younger group of adults (N = 145), aged 30 to 55 years, was forced into a confirmatory factor analysis in the current sample. The maximum likelihood estimate for the forced model would not converge to yield a unique solution. Thus, that model was not appropriate for use with our sample and a modified model needed to be established.
Knowledge from previous research [15,20,22] and information gained from the qualitative review of time lines and audio tapes were used to guide the construction of the new model. Eight strategies were tried in the confirmatory factor analysis model: increase fruit, increase vegetables, increase high-fiber foods (eg, beans and apricots), increase nutrient-specific foods (eg, tofu and bananas), decrease fried foods, decrease recreational foods, use fat-modified foods, and decrease cooking fat.
Inclusion of the 96 food changes in a single model was too large for the number of participants in this investigation. Therefore, the 8 identified strategies were first tested 1 at a time using reliabilities to examine the degree of internal consistency of the scales. Three strategies (increase high-fiber foods, increase nutrient-specific foods, and decrease fried foods) had low reliabilities ([less than]0.6) (ie, they did not measure the same concept) and were removed from further analysis. Reliability analysis was performed on each of the 5 remaining scales and the food items’ total correlations (the individual food change correlated with the total correlation of the other food changes in the strategy) were obtained (data not shown).
Again, because the number of cases was too small for the number of food changes, only 3 food changes from each of the 5 strategies were included in the confirmatory factor analysis. The 3 food changes chosen from each strategy were those that had the highest item correlation obtained from the reliability analysis. This model, which consisted of 5 strategies and 15 food changes, was designated as the general model. The general model provided an acceptable fit of the data and met all the goodness-of-fit criteria. The overall goodness-of-fit [[chi].sup.2] was 71.52 (df=80; P=.74), the goodness-of-fit index was 0.90, the root-mean-square residual was 0.08, and all the item loadings were significant at the 0.05 level (T values[greater than]2 (Table 3). In addition, except for a moderate intercorrelation (r=0.58) between use of fat-modified foods and decrease cooking fat strategies, all other fat-reduction strategies intercorrelations were [less than]0.4. All food changes had modification indexes [less than]8 and t he majority (89%) were [less than]5. The general model met all the fit criteria and appeared to be tenable.
Once the general model was established, we used information from the qualitative data to test all the food changes on their appropriate strategies. The food changes were added 1 at a time to the general model and tested for significance. A food change was dropped from the model for 1 of the following 3 reasons: if it was not significant and had low modification indexes on all the other strategies (eg, eggs); if it was significant, but exhibited a high modification index on 1 or several other strategies (eg, tofu); and if it was not significant, exhibited a high modification index on 1 or several other strategies, yet remained not significant after it was tried on each of the other strategies on which it exhibited high loadings (eg, baked/broiled chicken). Once all the significant food changes within 1 fat-reduction strategy were identified, we proceeded to test the remaining food changes on the next strategy. Using this systematic procedure, a total of 63 food changes were classified and 33 remained unclassif ied.
For the remaining 33 food changes, an exploratory factor analysis was performed to detect any fat-reduction strategies that may have been missed using the qualitative data. No new strategies were identified. Therefore, the unclassified food changes were not incorporated into the confirmatory factor analysis model.
The final model consisted of 5 fat-reduction strategies with 63 food changes (Table 3). These strategies were: increase summer fruits, increase vegetables and grains, decrease recreational foods, decrease cooking fat, and use fat-modified foods. Scales consisting of all the food changes that loaded significantly on their corresponding strategy were constructed. Scales demonstrated good internal consistency with reliability coefficients ranging from 0.66 (4 food items) to 0.86 (20 food items).
DISCUSSION
Several limitations of this study should be recognized. The small sample size limits the generalizability of the data and may have masked additional findings. With a larger sample, there would have been a greater possibility of categorizing most of the 33 unclassified food changes. For example, qualitative data strongly suggested that many participants consumed certain foods mainly because they were rich in specific nutrients, as if these foods were pills. Tofu was consumed because of its high phytoestrogen content and bananas because of their high potassium content. Quite possibly, each of these food changes may make up a fat-reduction strategy on its own. On the other hand, they could all belong to 1 strategy we labeled the nutrient-specific strategy. The small sample size made it difficult to test either of these hypotheses.
Another limitation is that the research was based on self-reported, retrospective recall of diet. Using self-reported dietary accounts introduces potential errors in the data. Some of these errors include participants’ failure to recall all foods consumed and participants’ desire to please the interviewer, which may lead them to exaggerate the use or non-use of certain food items. These errors of self-report are not only limited to FFQs, but are true for all dietary assessment methods, including 24-hour recalls. Assessment of retrospective food intake via semi-quantitative FFQs provides a reasonable estimation of food change trends [24-27]. Although retrospective recall may be appropriate as a preliminary step in developing hypotheses about dietary patterns of older adults, future research should follow persons who are in the process of making dietary changes, Dietary data could then be collected using well-controlled, multiple 24-hour dietary recalls [28,29] to confirm the identified strategies.
The majority of participants in this study were white, highly educated, and came from middle- to upper-socioeconomic groups, which limits extrapolation of the findings to the general population. The characteristics of this sample are consistent with previous research reporting that most dietary changers in the United States are white and more economically and more educationally advantaged [17,30-34]. These findings also agree with the Diffusion of Innovations theory developed by Rogers [35], who contends that innovators, or early adopters, are those who “buy-in” first when new ideas or technologies are introduced into
society. According to Rogers, early adopters make up a small percentage of the population ([less than]15%) and are usually highly educated and more economically advantaged. This may be a pattern being observed with the adoption of low-fat diets–the groups studied thus far are mainly the early adopters.
Keeping in mind the limitations of this research, several conclusions can be drawn. First, the majority of older adults in our sample decreased their fat intake gradually, at different time points in their lives, and over a long period of time. Second, while making dietary changes, these older adults followed a different pattern than that established by Keenan et al [15]. Third, older adults seem to make dietary changes using specific fat-reduction strategies.
The data confirm that dietary change is complex, dynamic, and does not happen overnight. Approximately two-thirds of participants indicated that they have unsuccessfully attempted to improve their diets on at least 1 other occasion. Qualitative data suggest that these previous attempts were not failures, but small successes. Every time a participant tried to change his or her diet, for example as a result of joining a weight loss program, a number of dietary changes were adopted, several of which were maintained, contributing to the overall lower-fat diet. In the majority of cases, dietary changes were made gradually and took a number of years and many clusters (an average of 16 clusters/person) before the goal was achieved. In fact, most participants reported that they were still in the process of making positive dietary changes.
The data also suggest that older adults’ dietary behavior is different from their younger counterparts, especially in men. Whereas studies of successful changers in younger adults have reported difficulties in recruiting younger men [15,30,34], this was not the case when recruiting older men. We found that older men were as eager to participate and as interested in nutrition and changing their dietary behavior as were older women. Future research should take into consideration the effect of age when studying men’s dietary behavior.
Age differences were also found in the way the fat-reduction strategies were classified. Keenan et al [15] identified 9 dietary rat-reduction strategies (in a group aged 30 to 55 years). Only 5 strategies were identified in our study. The 5 strategies were made up of certain food changes that were similar to those described by Keenan and colleagues [15], whereas other food changes emerged as different by comparison. These age differences need to be confirmed using a larger sample that is more distinctly different in age, and a more prospective approach to studying dietary patterns. If distinct age or life stage differences are established, nutrition educators would need to take age into account when developing future interventions.
Dietetics professionals are trained to give advice based on different meal patterns, specific food groups, specific nutrients, r adherence to a prescribed dietary regimen [36-38]. Our study suggests that older adults may use different food groupings than what is traditionally advised. The food changes were not grouped based on a set meal pattern or on a specific regimen, rather they were grouped into strategies that may fit best within the domains proposed by Kristal et al [20]. Kristal and colleagues [20] validated a diet behavior questionnaire for selecting diets low in fat in a group of 99 highly educated, white women aged 40 to 59 years. The questionnaire was based on anthropological theory of dietary change [39] and was developed to assess 4 broad behavioral domains (modification, substitution, replacement, and exclusion). Five scales (or strategies) were identified that corresponded to the hypothesized domains: modification of high-fat foods (eg, take skin off chicken), substitution of high-fat foods wi th specially manufactured lower-fat foods, replacement of high-fat foods with low-fat alternatives (eg, increase fruits and vegetables), avoid fat as seasoning, and avoid meat. Avoid fat as seasoning and avoid meat both belong to the same domain of exclusion of high-fat items. In our study, older adults used all of the strategies described by Kristal et al [20], but other new strategies emerged as well.
One new strategy, decrease recreational foods, also emerged in previous work [15]. It includes food changes that aim at decreasing foods often consumed during social occasions such as parties or picnics (Table 3). The decrease recreational foods strategy may represent a third dimension of the exclusion domain proposed by Kristal et al [20].
Decrease cooking fat was another strategy identified in this group of older adults. Prewitt and colleagues [22] used the diet behavior questionnaire developed by Kristal et al [20] in 235 African-American men and women (mean age 48 years) to measure dimensions of dietary fat behavior. The investigators found that the avoidance domain split into 2 strategies: the avoidance of fat as seasoning and the avoidance of high-fat food preparation. Our findings support those result and confirm the importance of “decrease cooking fat” as a dietary fat behavior. Keenan and colleagues [15] also identified a decrease cooking fat strategy in their sample.
Another notable difference is that, in our sample, fruits and vegetables were 2 distinct food categories. Whereas dietetics professionals often group fruits and vegetables together, such as in the 5-A-Day for Better Health message [37], our results, as well as previous research [15,40], indicate that fruits and vegetables represent different constructs. Fruits are more likely to be consumed in raw form, individually as a snack; in contrast, vegetables are usually consumed as part of a meal (often cooked), in a mixed dish, or in a sandwich. In short, fruits and vegetables are conceived of and used differently by consumers and should be separated when providing dietary advice.
APPLICATIONS
* Helping consumers successfully adopt and maintain new dietary patterns is a challenge faced regularly by dietetics professionals. It may be more appropriate for dietetics professionals to move away from providing dietary advice based on hypothetical premises, such as those based on food or nutrient groupings, and use the dietary strategies practiced by consumers. For example, nutrition educators should separate fruits and vegetables when providing dietary advice or planning educational interventions.
* Dietetics professionals should be aware that a short intervention, such as a 1-hour dietary counseling session or reading a brochure, will not achieve substantial reductions in a person’s dietary fat intake. Most persons need many more visits than currently allotted and repeated exposure to positive nutrition messages over a prolonged period of time to achieve desired dietary goals.
* Research is needed to confirm our dietary fat-reduction strategies prospectively and to determine the age differences in dietary behavior, especially in men.
R. AbuSabha is a nutritionist with the US Department of Agriculture, Food and Nutrition Service Alexandria, Va. At the time the study was completed, she was a research associate with The Pennsylvania State University, University Park. K-H Hsieh is a research associate in the Department of Human Development and Family Studies and C. Achterberg is dean of the Schreyer Honors College at The Pennsylvania State University, University Park.
References
(1.) US Dept of Health and Human Services, Centers for Disease Control and Prevention. Chronic Diseases and Their Risk Factors: The Nation’s Leading Causes of Death. Atlanta, Ga: National Chronic Disease Center; 1999.
(2.) National Research Council. Committee on Diet and Health: Diet and Health: Implications for reducing Chronic Disease Risk. Washington, DC: National Academy Press; 1989.
(3.) World Cancer Research Fund and American Institute for Cancer Research. Food, Nutrition and the Prevention of Cancer: A Global Perspective. Washington, DC: American Institute for Cancer Research; 1997.
(4.) Wynder EL, Weisburger JH, Ng SK. Nutrition: the need to define “optimal” intake as a basis for public policy decisions. Am J Public Health. 1992;82:346- 350.
(5.) Healthy People 2010: National Health Promotion and Disease Prevention Objectives. 2nd ed. Washington, DC: US Government Printing Office: 2000.
(6.) Enns CW. Goldman JD, Cook A. Trends in food and nutrient intakes by adults: NFCS 1977-78, CSFII 1989-91, and CSFII 1994-95. Fam Econ Nutr Rev. 1997;10:2-15.
(7.) Kristal AR, White E, Shattuck AL, Anderson GL, Fowler A, Urban N. Long-term maintenance of low-fat diets: durability of fat-related dietary habits in the Women’s Health Trial. J Am Diet Assoc. 1992;92:553-559.
(8.) White E, Shattuck AL, Kristal AR, Urban N, Prentice RL, Henderson MM, Insull W Jr, Moskowitz M, Goldman S, Woods M. Maintenance of low-fat diet: follow-up of the Women’s Health Trial. Cancer Epidemiol Biomarkers Prev. 1992;1:315-323.
(9.) Glanz K. Reducing breast cancer risk through changes in diet and alcohol intake: from clinic to community. Ann Behav Med. 1994;16:334-346.
(10.) Boyd NF, Cousins M, Beaton M. Quantitative changes in dietary fat intake and serum cholesterol in women: results from a randomized, controlled trial. Am J Clin Nutr. 1990;52:470-476.
(11.) Sorensen G, Morris DH, Hunt MK, Herbert JH, Harris DR, Stoddard A, Ockene J. The effect of a worksite nutrition intervention on employees’ dietary habits: the Treatwell program. Am J Public Health. 1992;82:877-880.
(12.) Sorensen G, Thompson K, Glanz K, Feng Z, Kinne S, DiClimente C, Emmons K, Heimendinger J, Probart C, Lichtenstein E. Worksite-based cancer prevention: primary results from the Working Well trial. Am J Public Health. 1996;86:939-947.
(13.) Finckenor M, Byrd-Bredbenner C. Nutrition intervention group program based on preaction-stage-oriented change processes of the Transtheoretical Model promotes long-term reduction in dietary fat intake. J Am Diet Assoc. 2000;100:335-342.
(14.) Fortmann SP, Taylor CB, Flora JA, Winkelby VA. Effect of community health education on plasma cholesterol levels and diet: the Stanford Five-City project. Am J Epidemiol. 1993;37:269-282.
(15.) Keenan DP, Achterberg C, Kris-Etherton PM, AbuSabha R, von Eye A. Use of qualitative and quantitative methods to define behavioral fat-reduction strategies and their relationship to dietary fat reduction in the Patterns of Change Study. J Am Diet Assoc. 1996;96:1245-1253.
(16.) Nationwide Food Consumption Survey: Food and Nutrient Intakes by Individuals in the United States, 1 Day, 1987-88. Washington, DC: US Dept of Agriculture; 1993.
(17.) Keenan DP, AbuSabha R, Sigman-Grant M, Achterberg C, Greenspan N. Factors that influence dietary fat reduction behaviors. J Nutr Educ. 1999;31:134-144.
(18.) Dohrenwend BS, Askenasy AR, Krasnoff L, Dohrenwend BP. Exemplification of a method for scaling life events: the PERI Life Events Scale. J Health Soc Behav. 1978;19:205-229.
(19.) Kristal AR, Shattuck AL, Williams A. Food Frequency Questionnaire for Diet intervention Research. Proceedings of the 17th National Nutrient Databank Conference, Baltimore, Maryland, June 7-10, 1992. Washington, DC: International Life Sciences Institute; 1994:110-25.
(20.) Kristal AR, Shattuck AL, Henry HJ. Patterns of dietary behavior associated with selecting diets low in fat: reliability and validity of a behavioral approach to dietary assessment. J Am Diet Assoc. 1990;90:214-220.
(21.) Long JS. Confirmatory Factor Analysis. Beverly Hills, Calif: Sage; 1983. Quantitative Applications in the Social Sciences series.
(22.) Prewitt TE, Durazo-Arivizu R, McGee DL, Luke A, Cooper RS. One size fits all: implications for assessing dietary behavior. J Am Diet Assoc. 1997;97(suppl 7):S70-S72.
(23.) Pedhazur EJ, Schmelkin LP. Measurement, Design, and Analysis: An Integrated Approach. Hillsdale, NJ: Erlbaum; 1991:653-658.
(24.) Van Staveren WA, West CE, Hoffmans MDAF, Bos P, Kardinaal AF, van Poppel GA, Schipper HJ, Hautvast JG, Hayes RB. Comparison of contemporary and retrospective estimates of food consumption made by a diet history method, Am J Epidemiol. 1986;123:884-893.
(25.) Byers TE, Rosenthal RI, Marshall JR, Rzepka TF, Cummings KM, Graham S. Diet history from the distant past: a methodological study. Nutr Cancer. 1983:5:69-77.
(26.) Xing X, Burr JA, Brasure JR, Neugut Al, Marshall JR. Reproducibility of nutrient intake in a food frequency questionnaire used in a general population. Nutr Cancer. 1996;25:259-268.
(27.) Thompson FE, Metzner HL, Lamphiear DE, Hawthorne VM. Characteristics of individuals and long-term reproducibility of dietary reports: the Tecumseh Diet Methodology study. J Clin Epidemiol. 1990:43:1169-1178.
(28.) Johnson RK, Driscoll P, Goran MI. Comparison of multiple-pass 24-hour recall estimates of energy intake with total energy expenditure determined by the doubly labeled water method in young children. J Am Diet Assoc. 1996:96:1140-1144.
(29.) Wynder EL, Cohen LA, Winters BL. The challenges of assessing fat intake in cancer research investigations. J Am Diet Assoc. 1 997;97(Suppl 7):S5-S8.
(30.) Peterson 5, Sigman-Grant M, Eisenstat B, Kris-Etherton P. Impact of adopting lower-fat food choices on energy and nutrient intakes of American adults. J Am Diet Assoc. 1999:99:177-183.
(31.) Sigman-Grant M. Can you have your low-fat cake and eat it too? The role of fat-modified products. J Am Diet Assoc. 1997; 97(Suppl 7):S76-S81.
(32.) Patterson RE, Kristal AR, Coatas RJ, Tylvasky FA, Ritenbaugh C, van Horn L, Caggiula AW, Snetselaar L. Low-fat diet practices of older women: prevalence and implications for dietary assessment. J Am Diet Assoc. 1996:97:670-679.
(33.) Subar AS, Heimendinger J, Krebs-Smith SM, Patterson BH, Kessler R Pivonka E. 5-A-Day for Better Health. A Baseline Study for Americans’ Fruit and Vegetable Consumption. Washington, DC: National Cancer Institute; 1992.
(34.) Auld GW, Bruhn CM, McNulty J, Bock MA, Gabel K, Lauritzen G, Medeiros D, Newman R, Nitzke 5, Oritz M, Read M, Schutz H, Sheehan ET. Reported adoption of dietary fat and fiber recommendations among consumers. J Am Diet Assoc. 2000:100:52-58.
(35.) Rogers EM. Diffusion of Innovations. 4th ed. New York, NY: The Free Press; 1995.
(36.) Glanz K. Compliance with dietary regimens: its magnitude, measurement, and determinants. Prey Med. 1980:9:787-792.
(37.) Heimendinger J, Chapelsky D. The National 5-A-Day for Better Health Program. Adv Exper Med Biol. 1996:401:199-206.
(38.) Williams SR. Nutrition and Diet Therapy. 8th ed. St Louis, Mo: Mosby: 1997.
(39.) Jerome NW. On determining food patterns of urban dwellers in contemporary United States society. In: Arnott ML, ed. Gastronomy, the Anthropology of Food Habits. Paris, France: Mouton Publishers; 1976.
(40.) Trudeau E, Kristal AR, Li S, Patterson RE. Demographic and psychosocial predictors of fruit and vegetable intakes differ: implications for dietary interventions. J Am Diet Assoc, 1998;98:1412-1417.
This research was funded by National Institutes of Health, National Cancer Institute grant No. CA-70971-01.
Table 1
Characteristics of older adults (N=65) who
successfully decreased dietary fat intake
Characteristic n %
Age (years)
50-59 31 48
60-69 20 31
70-79 10 15
80-89 4 6
Gender
Female 39 60
Male 26 40
Adults in household
Living alone 20 31
Living with a partner/other adult 37 57
Living with children 8 12
Ethnicity
White 63 97
African-American 2 3
Highest education level attained
High school or less 8 12
Graduated college/some college 26 40
Completed graduate school 31 48
Occupation
Retired 28 43
Professional occupation 16 25
Service occupation 12 18
Housewife 9 14
Number of previous attempts to improve diet
Never 17 26
Once or twice 25 39
Many times 23 35
Primary reason for making dietary changes
Improving personal health 35 54
Combination of improving personal health
and reducing own weight 21 32
Improving health of partner 6 9
Reducing own weight 3 5
Mean[+ or -]SD Range
Age by gender
Female 62.8[+ or -]9.1 50-88
Male 62.5[+ or -]8.3 52-84
SD=standard deviation.
Table 2
Characteristics of dietary patterns of older
adults (N=65) who were successful at
decreasing their fat intake
Variable Mean[+ or -]SD Range
Percent energy from fat
Before dietary improvement 44.3[+ or -]5.9 28.2-55.8
Current diet, after improvement 25.9[+ or -]7.1 7.0-42.3
Change 18.4[+ or -]8.0 3-43
Number of years decreasing fat intake 17.1[+ or -]8.6 5-43
Number of food changes made per person 40.9[+ or -]11.8 18-66
Number of change clusters per person 16.5[+ or -]9.9 1-50
Number of food changes made per cluster 10.4[+ or -]8.6 1-50
SD=standard deviation.
Table 3
Confirmatory factor analysis of fat-reduction
strategies by adults aged [greater than or
equal to]50 years (N=65): general model,
single-factor tests, and internal consistency
of scales
Factors T Values [a]
General
model
Decrease/change recreational foods
(n=14, reliability coefficient=0.76) [b]
Decrease cookies 4.37
Decrease chocolate candy 3.32
Decrease coleslaw 3.25
Increase low-fat cookies
Increase low-fat frozen desserts
Decrease pancakes, waffles
Decrease ice cream
Decrease pudding, custard
Decrease chips and crackers
Increase low-fat pizza
Decrease pies
Decrease chili with meat and beans
Decrease doughnuts, cakes
Increase low-fat cakes, pastries
Use fat-modified foods
(n=11, reliability coefficient=0.80)
Use low-fat dressing 7.34
Use low-fat mayonnaise 5.36
Use low-fat spreads [c] after cooking 4.78
vegetables
Use low-fat spreads[c] on bread
Remove skin off chicken
Decrease spreads [c] on vegetables, grains
Trim fat off beef
Change cooking fat
Use low-fat spreads [c] when cooking
vegetables
Use low-fat sour cream
Drink low-fat milk
Increase vegetables and grains
(n=14, reliability coefficient=0.79)
Increase sweet potatoes 5.24
Increase summer squash 4.16
Increase cooked greens 4.04
Increase broccoli
Decrease salad dressing
Increase winter squash
Increase beans
Increase vegetable soups
Increase rice, grains, noodles
Increase mixed salads
Increase onions, leeks
Increase bean soups
Increase cauliflower, cabbage
Increase red peppers, red chilies
Increase summer fruits
(n=4, reliability coefficient=0.66)
Increase cantaloupes 4.29
Increase watermelon 2.98
Increase peaches, nectarines, plums 2.77
Increase other melons, honeydew
Decrease cooking fat
(n=20, reliability coefficient=0.86)
Decrease noodles with cream sauce 5.24
Decrease spreads on bread [c] 4.60
Decrease beef, pork or lamb 4.42
Decrease ground meat
Decrease cream soups
Change to light popcorn [d]
Decrease noodles with meat sauce
Decrease meat casseroles
Change to plain tuna [e]
Decrease meat sandwiches
Decrease mayonnaise-based salads
Factors
Single-factor
test
Decrease/change recreational foods
(n=14, reliability coefficient=0.76) [b]
Decrease cookies 3.47
Decrease chocolate candy 4.15
Decrease coleslaw 5.55
Increase low-fat cookies 3.47
Increase low-fat frozen desserts 3.47
Decrease pancakes, waffles 3.27
Decrease ice cream 3.10
Decrease pudding, custard 2.88
Decrease chips and crackers 2.79
Increase low-fat pizza 2.79
Decrease pies 2.66
Decrease chili with meat and beans 2.54
Decrease doughnuts, cakes 2.20
Increase low-fat cakes, pastries 2.11
Use fat-modified foods
(n=11, reliability coefficient=0.80)
Use low-fat dressing 7.85
Use low-fat mayonnaise 5.21
Use low-fat spreads [c] after cooking 5.74
vegetables
Use low-fat spreads[c] on bread 4.66
Remove skin off chicken 4.33
Decrease spreads [c] on vegetables, grains 3.89
Trim fat off beef 3.53
Change cooking fat 3.25
Use low-fat spreads [c] when cooking 2.85
vegetables
Use low-fat sour cream 2.69
Drink low-fat milk 2.09
Increase vegetables and grains
(n=14, reliability coefficient=0.79)
Increase sweet potatoes 5.66
Increase summer squash 4.50
Increase cooked greens 5.50
Increase broccoli 4.20
Decrease salad dressing 4.18
Increase winter squash 4.13
Increase beans 2.92
Increase vegetable soups 2.89
Increase rice, grains, noodles 2.82
Increase mixed salads 2.60
Increase onions, leeks 2.57
Increase bean soups 2.55
Increase cauliflower, cabbage 2.43
Increase red peppers, red chilies 2.24
Increase summer fruits
(n=4, reliability coefficient=0.66)
Increase cantaloupes 6.05
Increase watermelon 3.22
Increase peaches, nectarines, plums 2.77
Increase other melons, honeydew 3.89
Decrease cooking fat
(n=20, reliability coefficient=0.86)
Decrease noodles with cream sauce 4.79
Decrease spreads on bread [c] 4.93
Decrease beef, pork or lamb 5.19
Decrease ground meat 4.97
Decrease cream soups 4.16
Change to light popcorn [d] 4.00
Decrease noodles with meat sauce 3.95
Decrease meat casseroles 3.90
Change to plain tuna [e] 3.87
Decrease meat sandwiches 3.61
Decrease mayonnaise-based salads 3.54
General
model
Decrease butter on popcorn 3.52
Decrease French fries 3.49
Decrease cheese, including in cooking 3.46
Change from dark to light chicken meat 3.44
Decrease cottage, ricotta cheese 3.25
Decrease pizza 3.05
Decrease meat gravies 3.02
Decrease biscuits, muffins 2.77
Decrease fried chicken 2.53
(a)General model [x.sup.2]=71.52 (df=80; P=
.74); T values between 2.0 and 2.65 are
significant at .05 level; T values [greater
than]2.65 are significant at .01 level.
(b)Food listed is representative of several
similar foods grouped together on the food
frequency questionnaire.
(c)Spread=butter, margarine, oil, or other
fat such as sour cream.
(d)Includes microwave “lite” or popcorn
without oil.
(e)Refers to plain tuna vs tuna salad with
mayonnaise or in a casserole.
COPYRIGHT 2001 American Dietetic Association
COPYRIGHT 2001 Gale Group
