Given the current health situation characterized by a remarkable prevalence of inadequate eating practices, the Transtheoretical Model (TTM) stands out as one of the most promising theoretical models for understanding and predicting behavior change [1]. The TTM features a core concept used to delineate more specific and adapted nutritional interventions termed the stages of change (SOC), which refer to when the behavioral change occurs and represent the different degrees of individuals’ motivation to modify their health behaviors. The stages are well described and include precontemplation, contemplation, preparation, action, and maintenance [2]. However, there are limitations and methodological challenges involving the measurement of the SOC and hence developing appropriate interventions [3].

The SOC of individuals is classified by applying an algorithm that refers to a specific questionnaire for one or more food items (e.g., fat consumption), which may be based on the perception of food and/or the assessment of food consumption [4]. Regarding the algorithm that is based solely on food perception, it is important to consider that individuals usually have an extremely optimistic view of their consumption, especially fat consumption [5]. Consequently, the misperception of food intake can lead to misclassification of the SOC, hindering interventions.

Accordingly, in 1996, Steptoe et al. [6] proposed reclassifying individuals classified in the maintenance stage who had a perception of adequate fat intake but who actually had high fat consumption into the “pseudo-maintenance (PM)” stage. Similarly, in 2003, Ma et al. [7] proposed reclassifying individuals who perceive their fruit and vegetable intake is inadequate despite having adequate consumption in the “non-reflective action” stage. It is unknown whether specific intervention strategies need to be developed for individuals in these new SOC [7].

As a substantial percentage of individuals usually misperceive their eating behavior, it is challenging to define the stages of behavioral change. Classification errors are reportedly persistent and can reduce the effectiveness of interventions [5]. Nevertheless, interventions based on the TTM usually do not investigate the concordance between perceived and actual food intake.

The following questions and knowledge gaps presented by the scientific community prompted this study: considering eating practices related to fats, what is the concordance between perceived and actual food consumption? What is the proportion of individuals in the PM stage, i.e., those who have an overly optimistic perception of their fat consumption? What is the proportion of individuals in the non-reflective action stage, i.e., those who have adequate fat intake but do not recognize it? Is it necessary to develop specific strategies for individuals reclassified in these stages? Therefore, the present study aimed to identify the proportion of individuals classified in the PM and non-reflective action stages for fat consumption as well as to explore subgroup effects by PM classification of a previously conducted, effective intervention. The main aim of the TTM-based intervention was reduction in the consumption of high-fat foods [8, 9].

The hypotheses were the following: (1) the proportion of individuals in the PM stage is high, (2) the proportion of individuals in the non-reflective action stage is low, and (3) individuals in the PM stage are less sensitive to interventions.

Of the 97 individuals who started the intervention, 79% completed the study (n = 77). Individuals who attended less than 50% of all workshops were excluded from the analysis [18]. In addition, male participants (n = 6) were excluded from the analysis in order to obtain a homogeneous sample. Thus, 71 women remained in the final analysis. Attrition data, comparisons of individuals who dropped out with those who completed the study, and intervention effectiveness have been reported previously [8].

Regarding the post-hoc reclassification of these participants, approximately half were in the PM stage. Figure 2 presents the classification of PM and participant allocation to the pre-action and action intervention subgroups. Regarding the non-reflective action stages, only one individual each in the UCG and TM-IG was reclassified; both individuals were classified into the preparation stage according to their perception of lipid intake although they had adequate fat consumption. Given the number of individuals classified in the PM stage, the effect of the intervention with respect to the classification at this stage was analyzed.

At baseline, participants classified into the PM stage had similar sociodemographic characteristics to those not classified into the PM stage (Table 1).

Participants in the PM stage who were reclassified into the action group had decreased caloric intake (pre-intervention: 2265.7 kcal [1504.5–2846.1]; post-intervention: 1679.7 kcal [1253.6–2073.1]), weight (pre-intervention: 71.3 ± 13.7 kg; post-intervention: 70.0 ± 13.3 kg), and BMI (pre-intervention: 28.5 ± 5.4 kg/m²; post-intervention: 28.0 ± 5.2 kg/m²).

Participants in the UCG who were not classified into the PM stage exhibited significantly increased lipid consumption (pre-intervention: 27.2 [25.3–30.8] %kcal; post-intervention: 33.1 [29.5–35.7] %kcal; p = 0.015), including saturated fatty acids (pre-intervention: 7.8 [6.7–8.4] %kcal; post-intervention: 9.3 [7.0–12.1] %kcal; p = 0.034). There were no statistically significant differences in lipid consumption for the TM-IG non-PM and UCG PM groups and in body perception for the other groups.

Starting from a baseline sample composed mostly of women in the action and maintenance stages, the post-hoc reclassification revealed a very small proportion of individuals in the non-reflective action stage and a large proportion in the PM stage. After participating in the TTM-based interventions, individuals in the PM stage evolved differently from others, exhibiting results that were more satisfactory than what was expected.

Regarding the non-reflective action stages, very few individuals did not recognize that they had adequate fat consumption. In contrast, Toral & Slater [4] found greater percentages of individuals in non-reflective action regarding fat consumption. However, their study involved teenagers, and the non-reflective action stage was originally developed for fruit and vegetable consumption [7]. Therefore, further investigations are required to confirm if the non-reflective action stage is not relevant for fat consumption.

On the other hand, approximately half of all participants were classified into the PM stage. This may be because most of the women were classified in the action or maintenance stage at baseline despite poor eating habits and the high prevalence of overweight and diseases [19–21].

Although studies do not usually evaluate the proportion of individuals in PM, the present results, specifically the large proportion of individuals in the later stages of the TTM at baseline, are corroborated by other investigations [20, 21]. Given the small sample size for these results, other studies should be conducted to explore whether the baseline classification accurately reflects actual eating perception and motivation to change.

Misconceptions regarding the perception of fat intake adequacy and consequently the design of interventions if individuals are not reclassified have been investigated previously [5, 22]. Misclassification based on perception can be due to an individual’s difficulty estimating fat consumption [23], which arises from difficulty visually assessing the presence of fat in food, because fat is usually added as an ingredient [4].

In addition, Plotnikoff et al. [5] suggest individuals may have exaggerated beliefs regarding their actions to reduce fat intake that arise from cognitive biases activated when analyzing this behavior. Making one or two changes such as reducing the consumption of fried snacks for example, may be sufficient to create an impression of having adequate fat consumption. Accordingly, the only question of intention used in the algorithm (i.e., “do you consistently avoid eating high-fat foods?”) may favor individuals who remember such changes in their eating behavior, resulting in a consideration of adequate fat intake in their diet.

The high percentage of individuals classified in the PM stage in the present study is corroborated by other studies evaluating this stage [4, 23]. However, other studies that evaluated the behavior of individuals in PM after participating in an intervention were not found. In the present study, women in PM exhibited better improvements in anthropometric parameters and food consumption as well as SOC progression.

To better understand the results exhibited by the women in PM in the TM-IG, it is important to consider that they participated in different intervention workshops (i.e., pre-action and action groups; Fig. 2).

Given the small group sizes, descriptive and exploratory statistics were used, and all participants in the TM-IG exhibited reduced calorie consumption, weight, and BMI (see Additional file 1) independent of the subgroups, i.e., non-PM in pre-action or action and PM in pre-action or action. However, the degree of improvement differed. Thus, this study is exploratory and pioneering, by comparing these groups of participants. Furthermore, the initial analyses are essential for formulating additional hypotheses with greater precision to guide future studies.

Interventions were more effective for women in PM reclassified into the action group; their median caloric intake decreased 503.3 kcal after the intervention, with post-intervention differences in mean weight (−1.26 kg) and BMI (−0.55 kg/m²). It is possible that these women reported healthy eating behaviors because they had already initiated healthy changes [7] but were still incipient to significantly reflect on the adequacy of their lipid intake. As a result, this classification into the PM stage would have occurred not because of a poor perception of their diet but rather because they have not achieved the recommended values for lipid intake (i.e., ≤30%) [4].

In addition, individuals with high consumption of lipids need to change multiple dietary habits, such as methods of food preparation, avoiding high-fat foods, and changing the type and amount of vegetable oil used. All these factors (high fat intake at baseline, misperception of their own consumption of lipids, and difficulty in changing dietary habits) may have influenced the evolution through the stages. Thus, the proposed intervention for the action group, which aimed to facilitate nutritional knowledge and suggested strategies for action, may have encouraged new attitudes among participants who were already in the process of change.

In contrast, women in the PM stage who were allocated to the pre-action group only initially believed they had appropriate lipid intake (i.e., step 1 of the algorithm). However, when asked about their food perception again, they revised their perception and considered themselves in the pre-action stage (i.e., step 3 of the algorithm). These women, who previously believed they had adequate fat intake, may have felt more motivated to change when they became aware of their true behavior. This subsequently prompted better evolution through the stages than the women not classified in the PM stage.

It is worth noting that, although some authors report the need to develop specific strategies for individuals in the PM stage [7], the present study is pioneering in the analysis of participants in the PM stage after an intervention. Furthermore, the participants exhibited positive changes even though the intervention was not developed specifically for their group. These preliminary results indicate that allocation of individuals in the PM stage to workshops for pre-action or action groups seems appropriate, showing the importance of valuing food perception in the design of TTM-based interventions [23]. These novel hypotheses must be investigated by further studies with larger samples to verify the consistency of intervention effects on pseudo-respondents.

Women in the action group who were not classified in the PM stage exhibited less evolution post-intervention than the others, as expected (see Additional file 1). The intervention is less novel for individuals in action stages who are already more engaged in healthy behaviors. Therefore, the intervention for this group aimed to prevent relapse and maintain behaviors [2].

Regarding the evolution of individuals who did not participate in the intervention, the usual care group (i.e., non-PM) exhibited increased lipid consumption. This result is concordant with global surveys showing increased fat intake among the general population [24], which reinforces the need and importance for interventions.

The main strength of this study is the validation of the algorithm used to classify individuals in the SOC for fat consumption. Furthermore, the interventions were based on the four pillars of the TTM. The cognitive and behavioral processes of change were strategically selected to promote motivation and progress through the SOC.

Despite this study’s strengths, some challenges and limitations are also present. The implementation of interventions for the pre-action and action subgroups may have limited the effects, because broad approaches that addressed all stages within in each group were required. However, 54 workshops were still held for these two groups. It is questionable whether workshops specific to the five SOC would be feasible in the context of health services.

It is difficult to compare this study with others because few intervention studies have evaluated the effectiveness of reclassifying individuals in the PM stage. Furthermore, the small sample size of the present study precluded identifying the statistical significance of other positive changes in the participants.

Future TTM-based interventions should assess the others pillars in changing eating behaviors (such as improved self-efficacy and reduced perceived barriers), which would expand knowledge about the cognitive and emotional dimensions of individuals. Evaluating these parameters will be important for advancing the understanding of the ability of interventions to positively affect individuals.

The present results regarding the PM stage must be verified by additional studies, which should also address the assumptions and questions raised herein. There was a high proportion of participants in the PM stage and a low proportion of participants in the non-reflective stages, and the individuals in the PM stage evolved distinctly despite participating in a nutritional intervention that was not tailored to their specific behaviors.

Women attending public health promotion services may misperceive their lipid intake and not recognize the related dietary habits to a greater extent, especially with a high lipid intake. However, women in the PM stage who participated in TTM-based interventions for dietary fat consumption exhibited distinct positive changes compared with the other participants, indicating the importance of considering this classification in the evaluation of TTM-based interventions. Regardless, there is a paucity of research assessing the existence of the PM stage.