Understanding self-reported staging of dietary behavior in low-income women

Abstract

This study examined self-reported staging for the goal of eating a low-fat diet and several specific dietary consumption behaviors to understand better readiness for dieting. Self-assessed motivation, food frequency measures, and psychosocial variables were obtained from 2057 low-income women enrolled in the Maryland Food for Life Program. Results indicated that staging of specific dietary consumption behaviors was significantly related to staging for the global goal of eating a low-fat diet. Women evaluate their motivation about eating low-fat diets based on perceived efforts and specific activities related to dietary consumption with important implications for dietary behavior change measurement and interventions.

Keywords

diet eating behavior socioeconomic status stages of change women’s health

Introduction

Most Americans consume a larger proportion of calories from fat than is recommended, and data suggest that fat intake remains high among women increasing the risk for obesity and its associated health conditions (Levi et al., 2010; Ogden et al., 2010, 2014; US Department of Agriculture and US Department of Health and Human and Services, 2010). Diets high in fat are associated with an increased body weight and elevated health risks. Eating a diet with less than 20–35 percent calories from fat, combined with balanced nutrition and energy expenditure, can help prevent weight gain and the health risks associated with excess fat, including stroke, cardiovascular disease, obesity, diabetes, and hypertension. Low-fat diets result in weight loss, overall reduction in waist circumference, decreased blood pressure, and lower levels of cholesterol (Shai et al., 2008).

Despite stable trends for fat intake among most ethnic groups, there has been an increase in total fat consumption among non-Hispanic Black males and females (Wright and Wang, 2010). This appears to be linked with socioeconomic status. While low-income black men are at lower risk for obesity than their higher income counterparts, low-income Black females are more likely to be obese (Ogden et al., 2010). Among Black females who fall below 130 percent poverty income ratio (PIR), 54.7 percent are obese. These rates are the highest when compared with non-Hispanic White and Mexican American females at the same poverty level (Dammann and Smith, 2009; Ogden et al., 2010). Women, Infants, and Children (WIC) is a federally funded food assistance program aimed at addressing nutrition-related risks among low-income women and children in an effort to improve birth outcomes and early childhood development as well as the health of the mother. This program provides nutrition education and food supplements to women who are pregnant, post-partum, or breastfeeding, and their infants and children as old as 5 years. Women are eligible for these benefits if their family income is at or below 185 percent of poverty level and are assessed as having a need for nutritional supplements. This program provides access to high-risk women and children and an opportunity for us to examine their motivation to eat a healthier lower fat diet.

A healthy diet requires daily dietary choices and modification of unhealthy food intake that involve changing attitudes and behaviors for those eating unhealthy diets. Although food and eating environments can be modified to offer healthier eating options (French et al., 2004; Story et al., 2008), individual choice and motivation are equally important in making the changes needed to eat healthy diets low in fat and high in fruits, vegetables, and fiber. This study attempts to understand and examine motivational considerations and dietary fat consumption of low-income women in the WIC program in an eastern urban city in the United States.

Assessing an individual’s dietary consumption and the motivation for making healthy dietary choices is challenging for many reasons. First, dietary choices that would make up a diet low in fat involve starting and stopping many varied eating behaviors that differ in their contribution to healthy eating. Second, a broad range of patterns of consumption (e.g. high-fat snacking, fried foods, fast foods, whole milk, and “supersizing”) contribute to dietary fat content (Greene and Rossi, 1998; Mozaffarian et al., 2011; Subar et al., 2001). Finally, the recommended goal for eating a diet low in fat includes eating less than 20–35 percent calories from fat depending on age and weight of the individual (Eckel et al., 2014; US Department of Agriculture and US Department of Health and Human and Services, 2010). Most individuals find this goal difficult to evaluate on a daily basis and hard to achieve (Sorensen et al., 1996; US Department of Agriculture and US Department of Health and Human and Services, 2010). All of these factors make assessing an individual’s motivation related to eating a healthy diet challenging (Bock et al., 1998; Curry et al., 1992; Glanz et al., 1994; Lechner et al., 1998).

In an effort to meet this challenge, several of studies have used the stages of change (SOC) construct from the Transtheoretical Model (TTM) of Intentional Behavior Change (DiClemente and Prochaska, 1998; Prochaska et al., 1992) to measure the readiness or motivation of participants to change various eating behaviors or patterns of dietary consumption (Armitage et al., 2004; Reimer et al., 2004; Rosen, 2000); some have evaluated stages for “eating five or more fruits and vegetables” per day (Feldman et al., 2000; Greene et al., 2008), for “eating a diet low in fat” (<30% calories from fat; Greene and Rossi, 1998; Havas et al., 2003), and for “dietary fat modification” (consumption of low-fat or non-fat foods; Tassell and Flett, 2005). Stages offer a way of segmenting the process of change into a series of steps and tasks related to making a change: precontemplation (not seriously considering change), contemplation (considering and decision making), preparation (building commitment and planning), action (implementing plan), and maintenance (sustaining change over time). Individuals in different SOC use different coping activities described as processes of change in the TTM and reliably differ on important markers of the intentional process of change, namely, decisional considerations and self-efficacy (DiClemente, 2005).

Previous dietary change studies found stage evaluations to be useful in categorizing populations according to their stage status and reflecting their readiness to complete tasks related to making change (Curry et al., 1992; Glanz et al., 1993, 1994). Investigators who evaluated both dietary content using a food frequency questionnaire (FFQ) and stages have found good correspondence between dietary content and the measures of stage status and motivation (Heimendinger et al., 1995; Sorensen et al., 1996; Tassell and Flett, 2005). Studies examining stages and other behavioral and psychosocial variables have been generally supportive of significant relationships between eating behaviors and stage-based subgroups (Armitage and Arden, 2007; Armitage et al., 2004; Di Noia and Thompson, 2012; Finckenor and Byrd-Bredbenner, 2000; Glanz et al., 1994; Tassell and Flett, 2005).

Stage assessments have also been associated with significant changes in eating behaviors over time (Havas et al., 2003; Johnson et al., 2008; Wright et al., 2009). Havas et al. (2003) found stage status related to a significant reduction in percent of kilocalories derived from fat and significant increases in the consumption of both fruits and vegetables as well as the intake of grams of dietary fiber. In addition, significant relations have been found between participants’ stage status and other core constructs of the TTM including experiential and behavioral coping behaviors (Hall and Rossi, 2008; Riebe et al., 2005; Wright et al., 2009), implementation intentions from the Theory of Planned Behavior (Armitage, 2006), as well as a continuous measure of intention across a range of related health behaviors (Armitage, 2006; Emmons et al., 1994).

Generally, stage classification systems are designed to evaluate first whether an individual is already engaging in a target behavior for shorter or longer period of time. For those not already engaging in the behavior assessment is of attitudes and intentions focused on his or her current, self-reported readiness to change this target behavior (Bock et al., 1998; Wright et al., 2009). This study examined self-reported motivational stage of change both for the global dietary goal of “eating a diet low in fat” as well as the motivational stage of change for each of seven specific eating behaviors that would contribute to the global goal of eating a diet low in fat. Specific dietary consumption behaviors refer to those dietary behaviors that contribute to a diet low in fat include drinking low-fat (1%) or non-fat (skim) milk, avoiding fried foods, checking food labels for fat content, avoiding high-fat sweets, buying “low-fat” or “fat-free” foods, avoiding high-fat snacks, and avoiding high-fat meats.

Participants consisted of a large group of low-income women recruited from the Special Supplemental Nutrition Program for WIC programs. The Maryland WIC Food for Life (FFL) program was a randomized trial funded by the National Cancer Institute (NCI). The primary goals were for the intervention participants to show a net decrease of 2 percent in the percentage of calories from fat, a net increase of .5 servings in fruit and vegetable consumption, and a net increase of 2 g in fiber consumption compared to control participants.

Data derived from this randomized trial called FFL (Havas et al., 2003) were used to examine three primary questions: (1) how does self-reported motivational stage of change for the global goal of eating a diet low in fat relate to a measure of dietary fat consumption, (2) how do women’s self-reported motivational stage of change for the global goal of eating a diet low in fat relate to their self-reported stages on individual specific low-fat dietary behaviors (e.g. avoiding high-fat snacks, drinking skim milk, etc.), and (3) finally, in order to support the construct validity and utility of staging self-evaluations, we examined whether there were significant differences between subgroups of women based on stage of change for the global goal of eating a diet low in fat when compared to diet-related psychosocial variables (dietary decision-making considerations, self-efficacy for eating a healthy diet, and coping activities measured by the processes of change for eating a healthy diet).

Methods

Study design

FFL was a site randomized trial examining the efficacy of a multicomponent dietary intervention that lasted over 6 months among women served by WIC nutrition programs at 10 sites in Maryland (Havas et al., 2003). Study eligibility requirements included enrollment in the WIC program, age 18 years or older, English-speaking, planning to remain at the WIC services location for the next 8 months, and not having previously participated in a prior study called the Maryland WIC 5-A-Day Promotion Program (Havas et al., 1994, 1998). All participants signed informed consent forms approved by the University of Maryland’s and the State Health Departments’ Institutional Review Boards (IRBs).

The FFL trial employed a crossover design in which participants were recruited from five sites randomized initially to intervention and five randomized to control status for Phase 1 of the trial (June through December of 1999). Then site intervention status was reversed during Phase 2 (August 2000 through January 2001) when a new group of participants was recruited from each of the 10 sites (five intervention and five control).

Using a site as its own control and a separate group of participants for each phase controlled for variance attributable to site in the interventions study. However, these analyses examined only the baseline data for these women across intervention and control conditions. Participants in the intervention received a multidimensional intervention consisting of health fairs, mailed materials, demonstrations, and trainings focused on making and eating foods low in fat targeting the seven specific targeted dietary behavior patterns (avoiding high-fat snacks, sweets, meats, fried foods; checking food labels, buying low and fat-free foods, and drinking low-fat milk) assessed at baseline. The intervention also targeted increasing the intake of foods high in fiber, fruits, and vegetables. Control participants received the normal WIC program with a minimal amount of time dedicated to nutrition education. Outcomes supported efficacy of the intervention for promoting dietary changes assessed at 6-month follow-up post-intervention for decreasing fat, and increasing fiber and fruits and vegetables (Havas et al., 2003). As noted above, this study examined only baseline data from these women prior to any FFL intervention.

Participants

A total of 2066 women enrolled in the Maryland WIC FFL Program. Overall, this sample is ethnically diverse (58% reporting being an ethnic minority), largely single (58%), and low-income with over half reporting earning less than US$15,000 in the previous year with almost one-quarter receiving food stamps (supplemental Table 1). Participants completed a self-administered survey upon enrollment and 2 months following the 6-month intervention. The evaluation sample for this study was 2057 due to critical missing data for 9 women.

Table 1.

Dietary outcomes and self-reported stages of change for eating a diet low in fat.

Dietary outcomes
Self-reported stages of change	Dietary (kcal)^a		% calories from fat^b		F&V servings^c		Fiber (g)^c
Self-reported stages of change	Mean	SD	Mean	SD	Mean	SD	Mean	SD
PC (n = 798)	2114.0	899.3	36.8	6.5	3.3	2.4	13.0	6.9
C (n = 536)	1970.3	809.6	36.0	6.7	3.2	2.2	12.6	7.0
P (n = 304)	1971.2	807.2	34.8	6.6	3.6	2.3	13.3	6.4
A (n = 254)	1814.8	828.6	31.3	7.0	3.9	2.3	14.6	7.2
M (n = 165)	1892.9	764.4	30.6	7.3	4.6	2.7	16.8	8.7
F-test	7.61		55.00		14.04		13.92
p value	<.001		<.001		<.001		<.001
R ²	.01		.10		.03		.03

F&V: fruit and vegetable; SD: standard deviation; PC: precontemplation; C: contemplation; P: preparation; A: action; M: maintenance.

Degrees of freedom = 2052 for all analyses.

A, M < PC.

M = A < P = C < PC.

PC = C < P = A < M.

Measures

The FFL 20-page scannable survey included demographics, smoking status, diet-related knowledge, attitudes, self-efficacy, social support, stage of change, and typical foods and quantities consumed in the past month. Measures were piloted and reviewed by peer counselors to insure readability and comprehension.

Food frequency

A 90-item modified version of the Block ’95 FFQ (Subar et al., 2001) assessed frequency of a wide array of foods typically consumed usually recorded in one of nine categories, with quantities ranging from never to weekly to 2 or more times per day and with three possible portion sizes listed. Additional precision in nutrient estimates was achieved with questions about low-fat or no-fat versions of 10 types of food including milk, lunch meats, and cookies. Nutrient estimates were calculated by multiplying the frequency of consumption by portion size by nutrient content of the food and summing across all foods, using the Block Dietary Analysis System, 1998, version 5.4, Berkeley, CA (see Havas et al., 2003; for more details on this measure).

Self-reported SOC

Self-reported SOC for the global goal of “eating a diet low in fat” were measured with a 5-item self-classification matrix. Participants were asked to indicate stage status for eating a diet low in fat (defined as ⩽30% calories from fat) by selecting one of the following categories: not thinking about it (precontemplation), thinking about starting in next 6 months (contemplation), definitely planning to start in the next 30 days (preparation), already doing it for less than 6 months (action), and already doing it for 6 or more months (maintenance), consistent with descriptions of stages in the TTM (Prochaska et al., 1992). In addition to this global low-fat diet item, using the same five staging categories, participants then indicated their stage status for seven specific dietary behavior patterns that contribute to a low-fat diet and were targeted by the FFL intervention (i.e. (1) “drinking low-fat (1%) or non-fat (skim) milk,” (2) “avoiding eating fried foods,” (3) “checking food labels for fat content,” (4) “avoiding eating high-fat sweets like cookies, cakes, pastries, and donuts,” (5) “buying foods with the words ‘low-fat’ or ‘fat-free’ on the label,” (6) “avoiding eating high-fat snacks like chips and nuts,” and (7) “avoiding eating high-fat meats like hamburgers, hot dogs, sausage, and lunch meats”). Stage of change measures has been used with a variety of populations, including low-income women and men (Wright et al., 2009) and validity of staging measures has been supported in multiple studies (Armitage and Arden, 2007; Sorensen et al., 1996).

Processes of change

Processes of change represent covert and overt coping activities that are related to stage of change status and progression through the stages. Experiential coping processes typically are more frequent in early stages and behavioral processes more often in later stages (Perz et al., 1996). An abbreviated 8-item assessment evaluated these coping processes of change for eating a healthier diet (4 items each measuring experiential and behavioral processes). Respondents rated how often they engage in activities, feelings, or thoughts related to eating a healthier diet from 1 (never) to 5 (very often). Items included, “I checked food labels at the grocery store” (behavioral) and “I have noticed that more people are eating healthy (experiential).” Internal consistency reliability (alpha) coefficient for the total scale was .88 and for the behavioral and experiential subscales .82 and .79, respectively.

Self-efficacy for eating a diet low in fat

Self-efficacy for eating a diet low in fat (Bandura, 1977) was measured using self-reported confidence to engage in eating a diet low in fat across seven situations rated on a scale of 1 (not at all) to 5 (agree a lot), for example, “I am sure I can eat low-fat foods when I eat out.” Internal consistency coefficient alpha was .87.

Decisional balance for eating a diet low in fat

Decisional balance for eating a diet low in fat (Janis and Mann, 1979) was evaluated using a decisional “balance sheet” of potential gains and losses typically related to an individual’s stage of change (Prochaska et al., 1994). The scale assessed the importance of pros (5 items) and cons (5 items) of eating a diet low in fat on a scale of 1 (disagree a lot) to 5 (agree a lot), including, “I feel better about myself when I eat low-fat foods” and “Checking food labels takes too much time.” Internal consistency coefficients were .77 for the pros and .51 for the cons.

Analysis plan

Women were grouped based on self-reported stage and these groups were compared using both univariate and multivariate and statistical analyses. In order to compare stage status for the goal of eating a diet low in fat to those reported for each of the specific targeted behaviors, we aggregated the stage status for each of seven specific targeted behaviors as described below. Pearson’s chi-square statistics were used to examine relations between subcomponent behaviors and participants’ stage for eating a diet low in fat. Differences were considered statistically significant when p values of .05 or lower were obtained. Finally, multivariate analyses (General Linear Model) were used to compare the low in fat diet stage groups on the key constructs from the TTM (i.e. processes of change, self-efficacy, and decisional balance for eating a diet low in fat). Pairwise comparisons derived from the overall general linear model evaluated differences on the mean scores of dependent variables.

Results

SOC for eating a diet low in fat and dietary consumption

Results indicated that subgroups based on self-reported motivational stage for the global goal of eating a diet low in fat were as follows for the 2057 women participants: 798 (38.8%) in precontemplation, 536 (26.1%) in contemplation, 304 (14.8%) in preparation, 254 (12.3%) in action, and 165 (8.0%) in maintenance. Most women placed themselves in pre-action stages with only 20 percent or so in action or maintenance. These five stage subgroups differed significantly on dietary consumption. For example, women in precontemplation and contemplation consumed significantly more calories, had a greater percent of calories from fat, and ate fewer fruits and vegetables and less fiber than those in action and maintenance (Table 1). The pattern of differences across dietary consumption behaviors was consistent with self-reported stage for the global goal of eating a low-fat diet, but as highlighted in Table 2, subgroups of women in some adjacent stages for some dietary outcomes often had values that were not significantly different (e.g. PC = C or A = M). To examine potential confounds to weight data, separate analyses were conducted co-varying pregnancy status, weight, and breastfeeding status and showed no differences in outcomes.

Table 2.

Cross-tabulation of self-reported stages of change (SOC) for eating a diet low in fat by component behaviors.

Component behaviors	Self-reported stage for eating a diet low in fat
	PC	C	P	A	M
Completely consistent (all low or all high) across seven behaviors (n = 369)
PC on all seven component behaviors	91.8	6.6	.9	.6	0
M on all seven component behaviors	0	5.9	3.9	15.7	74.5
Mostly consistent across seven behaviors (n = 869)
PC or C on all seven component behaviors	45.5	52.5	1.9	0	0
C or P on all seven component behaviors	3.3	75.6	20.0	1.1	0
P or A on all seven component behaviors	0	1.4	72.6	26.0	0
A or M on all seven component behaviors	0	7.5	6.3	65.0	21.3
Inconsistent across seven behaviors (n = 1188)
Inconsistent pattern across seven component behaviors	32.5	25.4	18.4	14.5	9.3

PC: precontemplation; M: maintenance; C: contemplation; P: preparation; A: action.

Bold values represent consistent response patterns for overall stage for eating a diet low in fat and component behaviors.

Examining specific elements of dietary consumption across subgroups based on stage status for eating a diet low in fat indicated that approximately 82 percent of the women who reported that they were in pre-action stages had dietary consumption data indicating that they were consuming more than the dietary goal of 30 percent calories from fat. However, only 41.8 percent of the women who classified themselves as being in action or maintenance stages for more global goal of eating a diet low in fat actually were eating diets that had less than 30 percent calories from fat. In fact, over one-third of the women who reported that they were in action for this goal and about 20 percent of those in maintenance reported diets that indicated that they were consuming more than 35 percent calories from fat according to the Block Food Frequency data. Although there are some discrepancies between reported stages and dietary consumption, almost three quarters of the sample (73.7%) reported eating diets that were consistent with pre-action or action stages for eating a diet low in fat.

Comparing staging subgroups for eating a diet low in fat to self-reported stages for the seven specific dietary behaviors

Several sequential analytic steps were taken to evaluate women’s motivational stage self-classifications for each of the seven specific dietary behaviors before comparing stage subgroups for eating a diet low in fat. First, we created stage scores for each woman by assigning a numerical score (1 to 5) to classify self-reported stage for each of the seven specific diet behaviors with 1 indicating precontemplation and 5 indicating maintenance for each behavior (e.g. eating high-fat snacks). Then we summed stage scores for each woman across all seven component behaviors with a range from 7 (all precontemplation) to 35 (all maintenance). Then we averaged them for each participant by dividing that total by 7 so each woman had an overall stage score across all seven behaviors. The mean of the averaged stage scores for all seven behaviors across the entire group was 2.4 (standard deviation (SD) = 1.1).

We then compared this mean composite stage score for the seven specific behaviors across the five groups of women based on their stage for eating a low-fat diet. Results indicated that there was a significant association between stage status for the global goal of eating a diet low in fat and women’s stage status on the seven component behaviors, F = 574.42, df = 2052, p < .001. All of the between-group comparisons were significant (i.e. PC < C < P < A < M). Women reporting that they were in precontemplation for eating a low-fat diet had an averaged stage score for the seven specific diet behaviors of 1.65 (SD = .77). Averaged stage score on the seven behaviors for the contemplation subgroup was 2.20 (SD = .74), preparation 2.87 (SD = .73), action 3.53 (SD = .86), and maintenance 4.14 (SD = .86). The more the women endorsed more advanced stage status for the global goal of eating a diet low in fat, the more they reported being in more advanced stages for multiple specific low-fat diet behaviors. For example, women who endorsed being in action for eating a diet low in fat had a mean score of 3.53 across the seven specific behaviors, which indicated that overall they averaged between preparation and action for these specific behaviors. Women who checked that they were in contemplation for eating a diet low in fat had average scores that indicated they were between contemplation and preparation on the seven component behaviors.

Second, we examined intra-individual consistency in these women’s self-reported stage status for the specific behaviors and their global low-fat diet stage status using cross-tabulations. We examined subgroups of women whose self-reported stage status for the seven behaviors was (1) completely consistent in stage for all the seven behaviors (i.e. all precontemplation or all maintenance across the seven behaviors (n = 369)), (2) mostly consistent in stage status (i.e. they reported being in adjacent stages (precontemplation or contemplation; preparation or action) for all seven component behaviors (n = 869)), or (3) rather inconsistent or scattered in reported stage status across the component behaviors (N = 1188; see Table 2). Almost half the sample reported that they were in very different stages of readiness for these seven specific low-fat diet behaviors. However, almost all women (91.8%) who were consistent and reported being in precontemplation for all seven component behaviors reported that they were in precontemplation for eating a low-fat diet. Similarly, the majority of the women who were completely consistent and reported being in maintenance for all component behaviors (74.5%) reported being in maintenance for eating a diet low in fat. Women who were largely consistent in reporting themselves in adjacent stages for all seven subcomponent behaviors reported a global low-fat diet stage status that was in one of these two stages with almost all the women (98.0%) in either precontemplation or contemplation for the seven component behaviors putting themselves in either precontemplation or contemplation for eating a diet low in fat and the vast majority (86.3%) who endorsed action or maintenance on all specific dietary behaviors rating themselves as in action or maintenance for eating a low-fat diet. There were significant associations between stages for component dietary behaviors and the stage status for the more global eating a diet low in fat.

Change-related variables and stage for eating a low-fat diet

A final evaluation examined the relations between women’s stage status for the global goal of eating a diet low in fat and a set of psychosocial variables that have been found to be related to stage status in previous studies. Significant differences emerged among the global low-in-fat diet stage subgroups on dietary decision-making considerations (pros and cons), self-efficacy for eating a low-fat diet, and women’s use of coping activities measured by the processes of change. Women who reported they were in earlier stages for eating a diet low in fat were significantly different from women in later SOC on all these variables (see supplemental Table 2). In fact, there were significant differences between each of the stage groups with two exceptions (differences were not significant between women in precontemplation and those in contemplation on experiential and behavioral processes and between women in contemplation and preparation on the cons of eating a diet low in fat). Self-reported motivational stage status for eating a diet low in fat was very consistent with the motivation, decision-making and coping activities of these women supporting the validity of these self-reported stage classifications in terms of women being consistent in their perceptions of where they are in the SOC and what change-related activities and evaluations they report.

Discussion

This descriptive study offers an interesting and comprehensive view of low-income, predominantly African American women’s self-assessed stage status both for a global eating goal of eating a low-fat diet and for a series of specific dietary behaviors that would support eating a low-fat diet. There were four key findings. First, women’s self-classification of their readiness to change or engage in various dietary behaviors as defined by the SOC is very consistent with their dietary attitudes and behaviors related to consumption. Their stage status for a more global goal of eating a diet low in fat reflects both their reported overall dietary consumption in terms of calories from fat, is associated with levels of dietary consumption of fruits, vegetables, and fiber, and reflects their attitudes, expectancies, and efforts with regard to engaging in specific dietary behaviors that contribute to a low-fat diet. These women are able to self-assess their stage of change for dietary behaviors in a remarkably coherent manner.

Second, some women endorsing action and maintenance overestimated their success and underestimated the fat content of their diets. It seems that they may be endorsing action stages because they are trying to eat a diet “lower” in fat rather than having achieved eating a diet “low” in fat reflecting effort and not goal achievement. Interestingly, there was also a small subgroup of women who, although they endorsed being in pre-action stages, actually had diets that met the targeted percent calories from fat. These women may be classifying themselves in pre-action because they are not intentionally trying to engage in specific behaviors for eating a diet low in fat and/or are not engaging in any activities to eat such a diet. Researchers and clinicians trying to evaluate motivation for dietary changes and predict behavior from motivation and stage assessments need to consider that stage status may reflect experiences related to dietary change and not necessarily achievement of eating a diet that meets a pre-established definition of a low-fat diet (⩽30% calories from fat in this study).

Third, when evaluating a complex behavioral goal like dietary fat content, it seems useful to evaluate component behaviors and not simply an overall goal like eating a diet low in fat. Women differed dramatically on their motivation for each of the seven behaviors that would contribute to eating a low-fat diet and were identified as targets of this intervention. About 60 percent of the women placed themselves in precontemplation for drinking skim or 1 percent milk, whereas only about 36 percent saw themselves in precontemplation for avoiding high-fat snacks. These differences in self-evaluations on specific diet behaviors are significantly related to overall self-classification about eating a diet low in fat. However, assessing readiness and stage status for specific behaviors, even the seven broad categories used in this study, offers a more differentiated way to explore motivation for changing specific patterns of behaviors that are the targets of dietary intervention. Information about motivation for specific behaviors that are related to a larger goal would have important implications for tailoring interventions.

Assessing and targeting specific patterns of behavior can make interventions more culturally sensitive as well. In this urban population of low-income women, there are both environmental barriers (few large markets, low-fat options being hard to access and costly) and cultural norms (ubiquity of fried foods in preparing meals) that can be assessed along with stage status for specific behavior patterns in order to create multicomponent interventions for this at risk population of women (Ogden et al., 2010).

Fourth, self-reported stage seems to reflect a complicated self-evaluation of coping activities, intentions, and expended effort even in this group of less educated women who are living in poverty. There are significant differences between pros of eating a low-fat diet, experiential and behavioral processes of change, and self-efficacy to eat a low-fat diet across the subgroups of women based on their stage for the global goal of eating a diet low in fat. Self-reported attitudes, evaluation, and coping activities indicate that self-reported global stage is related in an organized fashion to these personal mechanisms of change (DiClemente, 2003). However, when there seems to be a discrepancy between where they see themselves on the various change process variables (decision-making self-efficacy and process of change) and whether they are achieving the dietary goal of <30 percent calories from fat, they seem to rely on their perceived efforts and attitudes for calculating their stage for the global goal of eating a low-fat diet. For example, although women reporting themselves in the action stage for eating a low-fat diet reported a mean of 31.3 percent calories from fat, somewhat above the targeted goal, their decisional considerations, self-efficacy for eating a low-fat diet, and experiential and behavioral processes of change scores were higher than women in precontemplation and more comparable to women reporting themselves to be in maintenance. It seems that when a goal consists of multiple behaviors and is not clearly defined or measureable, individuals tend to assess their effort and activities related to that goal in order to evaluate their stage status and where they are in the process of change.

Stages represent an attempt to segment the entire process of intentional behavior change based on the tasks needed to be accomplished on the path to successfully sustained change (DiClemente, 2005, 2007). Self-staging as we have done in this study allows individuals to place themselves at different places along the path based on the chosen definitional algorithm. In this case, we simply defined the stage (thinking about doing it in the next 6 months or planning to do it in the next 30 days) and found important between-stage differences in behaviors and attitudes. This study and the findings from the parent study indicate that self-reported stage assessment is meaningfully related to dietary process and outcome variables (Havas et al., 2003). However, caution needs to be used when interpreting these findings or using such a classification to predict outcomes or to tailor interventions since not all studies have found stage to be predictive and some have found stage assessment difficult and challenging (Bridle et al., 2005; DeNooijer et al., 2005; Plotnikoff et al., 2009). Stages are not boxes or fixed traits but represent tasks and states that are rather fluid and can be difficult to assess and use (DiClemente, 2005). Strategies to assess stage should be piloted and clear distinctions need to be made between what constitutes action and pre-action status in order for this framework to be useful.

There are several important limitations to this study. Stage status is self-reported and there were no additional assessments that could be used to validate stage status (like interview or alternate measures). Women completed the stage assignments for overall goal and specific subcomponent at the same time which could have affected consistency of reporting. The study was cross-sectional in nature so causality cannot be determined. However, the parent study did find stage status to be related to changes during the intervention and post-test. Stage status was not completely consistent with dietary behaviors highlighting the difficulty of operationalizing and measuring stages. The specific dietary component behaviors were only a subset of those needed to make diets low in fat. Finally, these findings were in a low-income multiethnic group of women with children so may not be generalizable to other populations.

Despite limitations, results of this study indicate that self-reported diet-specific process variables and dietary behaviors are an integral part of how women evaluate their stage status. Thus, additional questions or micro assessments and strategies may be needed to more sensitively evaluate dietary attitudes and behaviors. For example, giving objective feedback to individuals about their actual dietary consumption and whether they are meeting goals before they are asked about attitudes and intentions for change may make for more accurate assessment.

The findings of this study highlight the variability in motivation regarding overall and behavior-specific dietary goals and support the suggestion that tailored interventions may be useful and necessary (Finckenor and Byrd-Bredbenner, 2000; Johnson et al., 2008; Kavookjian et al., 2005; Prochaska et al., 2004, 2005), albeit complicated; tailoring to someone in precontemplation who is not doing much to eat a low-fat diet but whose normal diet is one low in fat should differ from tailoring messages to those who see themselves in precontemplation and are consuming a high-fat diet. To be useful for assessment and tailoring, self-reported stage should be used as one diagnostic data point to be used in conjunction with other data and not simply as a label or grouping variable. Outcome assessments may need adjustment based on a multidimensional set of variables to adequately understand the impact of motivation on interventions especially with dietary behaviors.

Footnotes

Acknowledgements

The authors would like to thank their collaborators on the main study, Jean Anliker, PhD, RD, Deborah Greenberg, MA, Gladys Block, PhD, Torin Block, Cheryl Blik, MS, and Patricia Langenberg, PhD as well as Dorothy Damron, Project Manager for the first 3 years of the NCI grant. Additionally, the authors would like to thank Meagan Graydon and Anusri Patel, students in the HABITS lab at the University of Maryland Baltimore County, for conducting additional literature reviews in response to reviewer comments.

Funding

This original project was supported by the National Cancer Institute (grant no. RO1CA59725).

References

Armitage

(2006) Evidence that implementation intentions promote transitions between the stages of change. Journal of Consulting and Clinical Psychology 74(1): 141–151.

Armitage

Arden

(2007) Felt and potential ambivalence across the stages of change. Journal of Health Psychology 12(1): 149–158.

Armitage

Sheeran

Conner

. (2004) Stages of change or changes of stage? Predicting transitions in transtheoretical model stages in relation to healthy food choice. Journal of Consulting and Clinical Psychology 72(3): 491–499.

Bandura

(1977) Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review 84(2): 191–215.

Bock

Marcus

Rossi

. (1998) Motivational readiness for change: Diet, exercise, and smoking. American Journal of Health Behavior 22(4): 248–258.

Bridle

Pattenden

Mather

. (2005) Systematic review of the effectiveness of health behavior interventions based on the transtheoretical model. Psychology & Health 20(3): 283–301.

Curry

Kristal

Bowen

(1992) An application of the stage model of behavior change to dietary fat reduction. Health Education Research 7(1): 97–105.

Dammann

Smith

(2009) Factors affecting low-income women’s food choices and the perceived impact of dietary intake and socioeconomic status on their health and weight. Journal of Nutrition Education and Behavior 41(4): 242–253.

DeNooijer

Van Assema

De Vet

. (2005) How stable are stages of change for nutrition behaviors in the Netherlands? Health Promotion International 20(1): 27–32.

10.

DiClemente

(2003) Addiction and Change: How Addictions Develop and Addicted People Recover (1st edn). New York: The Guilford Press.

11.

DiClemente

(2005) Conceptual models and applied research: The ongoing contribution of the Transtheoretical model. Journal of Addictions Nursing 16: 5–12.

12.

DiClemente

(2007) Mechanisms, determinants and processes of change in the modification of drinking behavior. Alcoholism: Clinical and Experimental Research 31(S3): 13s–20s.

13.

DiClemente

Prochaska

(1998) Toward a comprehensive, transtheoretical model of change: Stages of change and addictive behaviors. In: Miller

Heather

(eds) Treating Addictive Behaviors (2nd edn). New York: Plenum Press, pp. 3–24.

14.

Di Noia

Thompson

(2012) Processes of change for increasing fruit and vegetable consumption among economically disadvantaged African American adolescents. Eating Behaviors 13(1): 58–61.

15.

Eckel

Jakicic

Ard

. (2014) 2013 AHA/ACC guideline on lifestyle management to reduce cardiovascular risk: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 129(25 Suppl. 2): S76–99.

16.

Emmons

Marcus

Linnan

. (1994) Mechanisms in multiple risk factor interventions: Smoking, physical activity, and dietary fat intake among manufacturing workers. Working Well Research Group. Preventive Medicine 23(4): 481–489.

17.

Feldman

Damron

Anliker

. (2000) The effect of the Maryland WIC 5-A-Day promotion program on participants’ stages of change for fruit and vegetable consumption. Health Education & Behavior 27(5): 649–663.

18.

Finckenor

Byrd-Bredbenner

(2000) Nutrition intervention group program based on preaction-stage-oriented change processes of the transtheoretical model promotes long-term reduction in dietary fat intake. Journal of the American Dietetic Association 100(3): 335–342.

19.

French

Story

Fulkerson

. (2004) An environmental intervention to promote lower-fat food choices in secondary schools: Outcomes of the TACOS Study. American Journal of Public Health 94(9): 1507–1512.

20.

Glanz

Kristal

Sorensen

. (1993) Development and validation of measures of psychosocial factors influencing fat- and fiber-related dietary behavior. Preventive Medicine 22(3): 373–387.

21.

Glanz

Patterson

Kristal

. (1994) Stages of change in adopting healthy diets: Fat, fiber, and correlates of nutrient intake. Health Education Quarterly 21(4): 499–519.

22.

Greene

Rossi

(1998) Stages of change for reducing dietary fat intake over 18 months. Journal of the American Dietetic Association 98(5): 529–534; quiz 535–536.

23.

Greene

Fey-Yensan

Padula

. (2008) Change in fruit and vegetable intake over 24 months in older adults: Results of the SENIOR project intervention. Gerontologist 48(3): 378–387.

24.

Hall

Rossi

(2008) Meta-analytic examination of the strong and weak principles across 48 health behaviors. Preventive Medicine 46(3): 266–274.

25.

Havas

Anliker

Damron

. (1998) Final results of the Maryland WIC 5-A-day promotion program. American Journal of Public Health 88(8): 1161–1167.

26.

Havas

Anliker

Greenberg

. (2003) Final results of the Maryland WIC food for life program. Preventive Medicine 37(5): 406–416.

27.

Havas

Heimendinger

Reynolds

. (1994) 5 a day for better health: A new research initiative. Journal of the American Dietetic Association 94(1): 32–36.

28.

Heimendinger

Feng

Emmons

. (1995) The Working Well Trial: Baseline dietary and smoking behaviors of employees and related worksite characteristics. The Working Well Research Group. Preventive Medicine 24(2): 180–193.

29.

Janis

Mann

(1979) Decision Making: A Psychological Analysis of Conflict, Choice, and Commitment. New York: Free Press.

30.

Johnson

Paiva

Cummins

. (2008) Transtheoretical model-based multiple behavior intervention for weight management: Effectiveness on a population basis. Preventive Medicine 46(3): 238–246.

31.

Kavookjian

Berger

Grimley

. (2005) Patient decision making: Strategies for diabetes diet adherence intervention. Research in Social & Administrative Pharmacy 1(3): 389–407.

32.

Lechner

Brug

De Vries

. (1998) Stages of change for fruit, vegetable and fat intake: Consequences of misconception. Health Education Research 13(1): 1–11.

33.

Levi

Vinter

St. Laurent

. (2010) F as in Fat: How obesity threatens America’s future. Available at: http://healthyamericans.org/reports/obesity2010/Obesity2010Report.pdf

34.

Mozaffarian

Hao

Rimm

. (2011) Changes in diet and lifestyle and long-term weight gain in women and men. New England Journal of Medicine 364(25): 2392–2404.

35.

Ogden

Carroll

Kit

. (2014) Prevalence of childhood and adult obesity in the United States, 2011–2012. JAMA 311(8): 806–814.

36.

Ogden

Lamb

Carroll

. (2010) Obesity and socioeconomic status in children and adolescents: United States, 2005–2008. NCHS Data Brief 51: 1–8.

37.

Perz

DiClemente

Carbonari

(1996) Doing the right thing at the right time? The interaction of stages and processes of change in successful smoking cessation. Health Psychology 15(6): 462–468.

38.

Plotnikoff

Hotz

Johnson

. (2009) Readiness to shop for low-fat foods: A population study. Journal of the American Dietetic Association 109(8): 1392–1397.

39.

Prochaska

DiClemente

Norcross

(1992) In search of how people change: Applications to addictive behaviors. American Psychologist 47(9): 1102–1114.

40.

Prochaska

Velicer

Redding

. (2005) Stage-based expert systems to guide a population of primary care patients to quit smoking, eat healthier, prevent skin cancer, and receive regular mammograms. Preventive Medicine 41(2): 406–416.

41.

Prochaska

Velicer

Rossi

. (1994) Stages of change and decisional balance for 12 problem behaviors. Health Psychology 13(1): 39–46.

42.

Prochaska

Velicer

Rossi

. (2004) Multiple risk expert systems interventions: Impact of simultaneous stage-matched expert system interventions for smoking, high-fat diet, and sun exposure in a population of parents. Health Psychology 23(5): 503–516.

43.

Reimer

Smith

Reicks

. (2004) Child-feeding strategies of African American women according to stage of change for fruit and vegetable consumption. Public Health Nutrition 7(4): 505–512.

44.

Riebe

Blissmer

Greene

. (2005) Long-term maintenance of exercise and healthy eating behaviors in overweight adults. Preventive Medicine 40(6): 769–778.

45.

Rosen

(2000) Is the sequencing of change processes by stage consistent across health problems? A meta-analysis. Health Psychology 19(6): 593–604.

46.

Shai

Schwarzfuchs

Henkin

. (2008) Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet. New England Journal of Medicine 359(3): 229–241.

47.

Sorensen

Thompson

Glanz

. (1996) Work site-based cancer prevention: Primary results from the Working Well Trial. American Journal of Public Health 86(7): 939–947.

48.

Story

Kaphingst

Robinson-O’Brien

. (2008) Creating healthy food and eating environments: Policy and environmental approaches. Annual Review of Public Health 29: 253–272.

49.

Subar

Thompson

Kipnis

. (2001) Comparative validation of the Block, Willett, and National Cancer Institute food frequency questionnaires: The Eating at America’s Table Study. American Journal of Epidemiology 154(12): 1089–1099.

50.

Tassell

Flett

(2005) Stages of change for fruit and vegetable intake and dietary fat modification in Maori women: Some relationships with body attitudes and eating behaviours. New Zealand Journal of Psychology 34(1): 28–34.

51.

US Department of Agriculture and US Department of Health and Human and Services (2010) Dietary Guidelines for Americans (7th edn). Washington, DC: US Government Printing Office and US Government Printing Office.

52.

Wright

Velicer

Prochaska

(2009) Testing the predictive power of the transtheoretical model of behavior change applied to dietary fat intake. Health Education Research 24(2): 224–236.

53.

Wright

Wang

C-Y

(2010) Trends in intake of energy and macronutrients in adults from 1999–2000 through 2007–2008. NCHS Data Brief 49: 1–8.