A New Consistent Concept Between the I-change Model and the Transtheoretical Outcome to Imply the Leading Guideline of Depression Behavior Health Change Using Social Media

2.1. Social Media with Depression

Social media is a digital tool used for a variety of activities, such as communicating, sharing news, exchanging opinions [51, 52], and doing business. Through research on popular platforms (e.g., Facebook, Line, TikTok, Instagram, Twitter) [52, 53] it has been found that social media have their benefits and disadvantages. In a study on health effects, it was found that activities on social media were an important predictor of individuals’ mental state and living in society [54] that led to depression [55-57], which may, in turn, result in suicide [26, 58]. As a result, additional research was conducted to validate these results. Similarly, the current study was divided into two components: (a) the correlation between social media use and addictive behavior, which, in turn, affects the course of depression, and (b) the association between social media use and the health behavior of individuals diagnosed with depression (Table 1).

These studies show that information on the impact of social media use or lifestyle interference on psychiatric patients is increasingly presented today. This leads to social media relationships with patients with depression, which influence and lead to depression [16, 17, 20, 61] and influences the behavioral modification of depression patients [58, 62]. Research has confirmed that social media has entered a role and a major role in the daily life of patients with depression [63, 64], influencing intentions to act or modify health behavior on various issues [29, 65, 66]. Social media have been used as a medium to engage in groups and share health information [67, 68], and social media has transformed popular methods to promote health effectively, interacting through various channels [69, 70]. In addition, there has been an in-depth study of data, including content type, channel, and source, that can confirm that content received on social media affects the changing health behaviors of patients with depression [30]. Therefore, this research focuses on how information received from social media interventions will affect the process of changing health behaviors.

2.2. Health Behavior Theories

Health behavior encompasses health, actions, or expressions of human actions or inactions for health [71, 72] in the physical, psychological, and social aspects [73, 74] expressed to oneself, family, and community [75]. Health behavior can be divided into five categories: disease treatment, disease prevention, health promotion, health care participation, and consumer health protection [76, 77]. However, if any action is detrimental to one's health, one must find a way to change health behavior further [78, 79]. Presently, various concepts and theories are employed to elucidate behavior in order to understand healthy behaviors and health behavioral modifications [77, 80], and influential social psychological factors have been identified in individual health behavior [80, 81]. It can be divided into three levels: individual health behavior theory and model, interpersonal health behavior theory and model, and community and group health theories and models [77, 82]. The research focuses on the theory of personality that emphasizes the understanding of individuals that influence behavioral change, such as the health belief model (HBM) [38], which is a theory focused on various aspects of perception. It is a factor of belief that arises from the perception of something that influences healthy behavior. The results focus on disease prevention and treatment. The theory of reasoned action (TRA) [83] is a theory that studies healthcare intentions based on a person's behavior as the result of thinking and reasoning. The theory of planned behavior (TPB) [84] is a theory that studies behavioral change by intent and behavior control. A person's healthy behavior is the result of an intention to act that is influenced by beliefs about attitudes to behaviors, the influence of reference groups, and the perception of the ability to control behaviors. The Protection Motivation Theory (PMT) [39] uses a study of risk-prevention behavior. The ASE model [85] is utilized to explain health behavior by considering the factors of attitude, social influences, and self-efficacy. In addition, this theory considers perceived self-efficacy through behavioral intentions and actual behavior. The above theory focuses mainly on explaining factors [81], thus, the argument is that a factor focuses on the cognitive perspective and affects the health behavior of the person. However, there are limitations in each of these theories because they are unable to explain factors from the perspective of the health behavior change process. Furthermore, because of this, theories that are elected to be used should be in line with the goals of the studies and have sufficient complexities to explicate health behavior problems in a study on behavior modification processes. The concepts for explaining health behavior are summarized in Table 2.

2.3. I-change Model

The I-change model (ICM) is a health behavioral theory that places emphasis on the consequences of intentions that lead to behavior changes [41, 86] with the relevant factors presented in each component. Part one is: 1) information factors (health-related information) evaluated from personal, message, channel, and source. Part two is: 2) about predisposing factors. Next, part three is represented in the form of processes but not as a practical process. It consists of three main factors: 1) awareness factors, assessed from knowledge, cues to action, and risk perceptions; 2) motivation factors, evaluated by attitude, social influence, self-efficiency, and action intention; and 3) action, planning, skill, and barrier until the result is in the behavior [29, 41, 42], as shown in Fig. (1). Within the theoretical framework, it illustrates the interconnections among each factor. The ICM theory is an extensively recognized concept utilized to influence health behaviors such as exercise, drinking alcohol [87], predictive smoking behavior [88], sleeping behavior [89], weight loss [90], eating behavior [91], and cyberbullying [92]. In each of the various health behaviors mentioned above, factors from the I-change model, including predisposing, information, awareness, motivation, intention, action, and behavior, were used for analysis. Different factors may be used in the analysis depending on the issues and goals of the work. The analysis is conducted in the form of logistic regression, multiple linear regression analyses, and structural equation modeling (SEM). Furthermore, ICM has been used for patients suffering from depression, which is a group with complex behavior. Moreover, by studying the intention to change health behavior, we can see how it is affected by accepting social media interference [29]. The results of the study revealed the factors that led to the intention to modify health behavior, which resulted in an explanation of only the relations between factors of the I-change model. Additionally, there was also a study using the I-change model that focused on information issues (e.g., content type, channel, source, perceived privacy risk) that influence changes in health behavior affected by social media risk factors among patients with depression [30]. The study found associations between content type, channel, and source and perceived privacy risk only under the theoretical framework of the I-Change model. In many studies, associations between factors within this theoretical framework were analyzed, which is a non-practical nature in explaining concepts. There has not been any research that has analyzed the elements of the I-change model, or each factor separately. Determine the relationship between each factor that influences the outcome of the health behavior change process.

2.4. The Transtheoretical Model

The transtheoretical model that describes the process of change in individual behavior [43, 94] passed through a six-stage implementation [44]. The first stage is re-contemplation, which means they are not planning to change their behavior in the next six months. Contemplation is confusion or thinking about change; preparation is moving towards change; action is attempting to change; and maintenance is a behavior change that can be maintained for more than 6 months and works to prevent relapse, leading to the termination phase. This is the stage where people are fully confident that they will not return to their old behavior [94-96]. Achieving success at each stage depends on individual variations. It involves emotions, cognition, behavior, social support, information received, etc [97]. Each step can change up and down over time based on a variety of factors. Many people change their desired behavior and soon return to their original behavior again. By changing successful behaviors, new behaviors should be sustainable [44]. TTM has become one of the most referenced and widely used models for health behavioral change interventions, it is clinically the most widely utilized in the field of mental health [95] and is effective in solving many problems, such as smoking [98-100] drinking alcohol [101] use of drugs, physical activity [49, 102] accepting and maintaining PA [103] weight control [104] medication adherence [105, 106] and contraceptive [107]. Several earlier studies have shown that the Transtheoretical Model (TTM) can be used to change behavior. Each work will focus on a different analysis model based on the main concepts of the TTM model, which include state of change, process of change, decision balance, and self-efficacy [44]. In the analysis, statistical principles will be used, such as correlation analysis, one-way ANOVA, and path analysis. Descriptive statistics and logistic regression analyses, etc, are also used to analyse health behavior. This model focuses on creating behavioral interventions at each stage [108]. In addition, TTM has been introduced for patients with behavioral complications, such as patients with combined depression and other disorders [108-110]. However, there are currently few studies for people suffering from depression. Moreover, by showing behavioral interventions through activities or methods in each step to treat and adjust patient behavior on issues such as eating, exercising, medication adherence, etc., and discovering research into how to use TTM as a framework to examine the structural relationships of the theory has been considered.

The I-Change and transtheoretical models incorporate health behavioral alterations employed in various subjects to target results in behavior change (see Items 2.3 and 2.4). The difference, however, is that the I-Change model is a conceptual theory which does not reflect the process or stage of behavioral modification (Table 3). Research related to the use of the I-change model has only described the factors and the relations between the studied variables. It explains the potential cause and effect between external and internal variables to assess the model's consistency with other relevant theories. It is a proof-of-concept examination of the relationships between each factor within the I-Change model. It has not led to actual implementation in the form of processes. No studies have analyzed each factor separately. In contrast, the transtheoretical model represents the process of behavior change and is characterized by the practice of the processes in five stages. When a person goes through a behavior change phase without a relapse, the person enters the final phase, i.e., termination, at which a person can change behaviors permanently. Research has focused on behavioral interventions using various methods in each step and has examined the structural relations of the theory by looking at behavioral interventions in each phase, wherein processes at each level are the results of the behavioral intervention within the framework of the model.

These comparisons show clear pictures of the I-change model, a theory integrating the relations between factors, and of the transtheoretical model, a theory that focuses on changing behavior practiced through various steps to lead to permanent and sustainable new behaviors. Therefore, this leads to a new idea that studies the relationship between the two theories. Additionally, by using the I-Change Model, each factor is separated into components to lay out their relation to the overall results of the health behavior change process under the theoretical framework of the transtheoretical model. A case study has been carried out of a group of patientswith depression from using social media. This confirms the relationship between the theory that is the conceptual model and the theory that models the process, leading to an in-depth study of the TTM model separately at each stage in the future.

3.1. Proposed Conceptual Model

The examination of the elements that affect the health behavior change in people with depression is based on the effects of information, awareness, motivation, intention, and action, which are factors that can explain modifying health behaviors through integrating the TTM, which was used as a framework for the study as shown in Fig. (2).

3.2. Hypotheses

Predisposing factors were the basic data of the patients, which were divided into behavioral factors, psychological factors, social factors, and Environment factors [42], and these factors influenced the behavior of patients.(Huang, 2022; Keles et al., 2020, Shensa et al., 2018). For this study, the following baseline data were included: gender, age, education level, social media usage experience (hour), platform of social media use, and total time per day (hour).

Table 3.

Theories	Definitions	Type
Factors and Theory	Process and Framework	Practical Treatment
I-change model [42]	It is a form of health behavior theory that explains changes in health behavior that result from motivation or intention.		-	-
The transtheoretical model [44]	It is a model that outlines the process of altering individual behavior at each stage, which results from a person's goal and preparedness to modify their own conduct.	-
This Paper	A new consistent concept between the I-change model and the transtheoretical outcome to imply the leading guideline of depression behavior health change using social media

An information factor is a description of information obtained from various sources. This focuses on information about health and behavioral modification. It consists of 1) personal factors. 2) message factors related to the content, type of content, frequency, quantity of content received, and quality of information. 3) channel factors: channel for obtaining health information; and 4) source factors: source for receiving health information. Recent research has demonstrated the relationship between the information or content received from social media, describing the relation between the quality of the information and the perception of the reliability of the source, which is important to the content shared on social media that influences individual perceptions [111, 112]. In addition to obtaining online health information, communicating health information on social media can lead to future behavioral changes [113, 114]. Moreover, research has confirmed the importance of information that influences behavior changes in TTM processes, such as eating behaviors and weight gain in young children [115]. Therefore, this study describes the information and advice received from social media about health behavior caused by group membership or obtaining information from various sources. The following is the assumption:

H1: Information factors have a positive effect on the patient's health behavior change process because of the social media use intervention.

Awareness factors are the factors that refer to awareness, which is the perception that self-behavior affects behavior [116, 117] and influences a person's behavioral motivation [42, 118]. These include 1) knowledge about such behaviors that cause harm to health [117], 2) risk perceptions are the perception of the symptoms, severity, and severity of the disease that affects health [119, 120], and 3) cues to action, events or alarms caused by close friends or people that cause a change in behavior or environment [90, 117-119]. Therefore, this study will focus on the advantages, disadvantages, and advice on the awareness of social media use that influences behavioral change caused by the following assumptions:

H2: Awareness factors have a positive effect on the patient's health behavior change process because of the social media use intervention.

Motivation factors are behavior-modification motivations [121] in which motivation factors are influenced by: 1) attitudes as perceptions of the advantages and disadvantages of behaviour [42, 90, 121], 2) social influences as norms or comments on such behaviours from intimate persons, which are important and motivating in behavioural modification [42, 119]); and 3) self-awareness as the perception of a person's level of ability in a situation or at the time of conduct, or confidence in his or her ability to conduct behavioral actions [122, 123]. Motivation influences a person's intention toward changing behavior [29] and is the determinant that causes action in a behavioral modification process. As a result, this study describes the motivation for health behavior change because of social media interference caused by the following assumptions:

H3: Motivation factors have a positive effect on the patient's health behavior change process because of the social media use intervention.

Intention factors are the intention to do something if referred to in terms of behaviour, it is the intent to modify a person's behavior [41], which is a post-motivation phase [29] that leads to those behaviors and involves the planning and implementation of a plan for targeted behavior [117, 124]. Furthermore, it mostly defines intention as the outcome of the work. This study examines the impact of social media interference on the patient's intention to change their health behavior.

H4: Intention factors have a positive effect on the patient's health behavior change process because of the social media use intervention.

Action factorss mention barriers, which are problems and constraints in behavioral modification, and are factors that may interfere with behavior adaptation, occurring between intentions and behaviors [125], which affects the achievement of health behavior goals [126, 127]. Therefore, this study will evaluate the involvement, problems, and barriers that affect behavioral modification caused by social media usage.

H5: Action factors have a positive effect on the patient's health behavior change process because of the social media use intervention.

4.1. Study Participants and Setting

Moreover, to determine the sample group used for testing the hypothesis of the research, a total of 180 individuals were chosen for the sample group using the method of simple random sampling. This is how everyone in the population has the same probability of being randomly selected into a sample by using calculations from statistical formulas. In this research, the formula for calculating the size of the population according to the formula for finding the number of samples using the table of Krejcie & Morgan will select the sample group that meets the study's objectives, which were to primarily include patients who were diagnosed with mild to moderate depression. The sample utilized for this study was 180 patients with mild and moderate depression who were evaluated for depression symptoms using the 9-question (9Q) Depression Scale to determine the severity of symptoms at all 3 levels: mild, moderate, and severe in the 2 weeks preceding study participation. The results were evaluated by a psychiatrist, clinical psychologist, or psychiatric nurse. During the months of November 2022 and February 2023, data were collected from a facility in Thailand. All participants provided written informed consent before answering the question, and the study approach was authorized by Mahidol University's Centre for Ethical Reinforcement for Human Study (CIRB) with the reference number COA. No.MU-CIRB 2022/119.0711.

4.2. Data Collection

The questionnaire was developed in accordance with the Transtheoretical model and the I-change model. The questionnaire consists of three sections, from which thirty questions were selected to comprise the questionnaire. Section 1 outlines the information in general and social media use; Section 2 outlines the information, awareness, motivation, intention, and action factors; and Section 3 outlines the recommendations. The questionnaire replies were graded on a five-point Likert scale, with 1 being strongly disagreeing and 5 being strongly agreeing. The third section contains open-ended inquiries, including suggestions and recommendations.

Prior to commencing data collection, a clinical psychologist or psychiatric nurse conducted a Mental Status Examination (MSE) to evaluate the patient's level of awareness and ascertain their preparedness to respond to questions. After that, a clinical psychologist or psychiatric nurse read the questionnaire questions to the patient in a personal conversation with the patient in a manner like the therapy process to prevent possible questioning from provoking the patient's symptoms. In addition, it was also about observing and evaluating the situation. The clinical psychologist or mental nurse wrote down the patients' answers after talking with them to evaluate them and made sure they do not have to answer questions that could make their condition worse. This is also performed to evaluate the situation before the questionnaire phase. Each conversation lasts no more than 15 minutes (Table 4).

4.3. Data Analysis

In research, statistics are employed. It is a study of causal links using SmartPLS 4.0.9 [129] is utilized to analyze the partial least squares structural equation model (Partial Least Square Structural Equation Modeling). There are two types of analysis shown: measurement model analysis and equation model analysis structure [130].

5.1. Descriptive Statistics

A random sample of 200 persons who have been diagnosed with depression was recruited, and complete data were obtained from 180 participants. Table 5 displays the characteristics of the participants' demographics.

Table 5 shows the demographic characteristics, revealing that 58.59% of participants were female, 41.11% were between the ages of 26 – 35 years, and most of them had a bachelor's degree. Most of the samples, which made up 47.22 percent, had used social media for more than 3 years. Facebook was the most used social media, with 37.22 percent, and 47.22 percent of the time was spent on social media for 1–3 hours, which accounts for 35%.

5.2. Measurement Model

The examination of variable quality was conducted based on the criteria outlined in a study [130] and the questionnaire's measurement list of questions. According to the findings, all composite reliability (CR) values fall within the range of 0.884–0.942, and this fits the criteria of not being less than 0.7. Similarly, Cronbach's α values range between 0.882 and 0.938, also meeting the criterion of not being less than 0.7. Additionally, the average variance extracted (AVE) values ranged from 0.711 to 0.809, which is greater than the 0.5 minimum criteria, as shown in Table 6.

The following is a compilation of inquiries utilized in quality assessment surveys, satisfying all requirements pertaining to weight values. The dependability of the question list used in educational measurement is demonstrated by considering a value that is equal to or greater than 0.7 based on specific criteria. The set of questionnaires exhibited a weight outer loading ranging from 0.800 to 0.926. Additionally, the Fornell-Larcker criteria were employed [35], for evaluating the relationship between variables in the form of a diagonal matrix, it was found that the square roots of AVEs in each variable (bold letters) are greater than the corresponding horizontal and vertical values. This shows that the interpreter had discriminant validity and can be used for structural equation model analysis, as shown in Table 7.

5.3. Structural Model

Subsequently, the structural model should be validated by employing the bootstrap approach on resampled patient data with 5,000 lists, to enhance the reliability of examining the associations among components. Multicollinearity tests were conducted using VIF values, revealing that the causative variables exhibit no correlation over the threshold of 5 [131]. When evaluating the importance of path coefficients, p-values and t-values are examined based on predetermined thresholds. These thresholds are set at t-values greater than 1.96 (at a significance level of 5%), 2.58 (at a significance level of 1%), and 3.29 (at a significance level of 0.1%) [131]. Furthermore, it was found that accepting hypotheses H1: Information Factors (INF) had a positive effect on the transtheoretical model (TTM) outcome at the significant level 0.05 (β = 0.156, t=2.3). Accepting H2: Awareness (AWN) had a positive effect on that the transtheoretical model (TTM) outcome at the significant level 0.05 (β = 0.112, t=2.091). Accepting H3: Motivation (MOT) had a positive effect on that the transtheoretical model (TTM) outcome at the significant level 0.05 (β = 0.236, t=2.421). Accepting H4: Intention (INT) had a positive effect on the transtheoretical model (TTM) outcome at the significant level 0.05 (β = 0.047, t=1.983). Accepting H5: Action (ACT) had a positive effect on the transtheoretical model (TTM) outcome at a significant level 0.01 (β = 0.368, t=3.532). The results are shown in Table 8 and Fig. (3).

5.4. Model Fit

Moreover, to determine model fit, all model structures proposed were evaluated. The outcomes of the model's goodness of fit analysis can be categorized into three components [132]. The first component is the determination coefficient (R²), with the criterion that the model is unacceptable if the determination coefficient (R²) is less than 0.19. If the value falls between 0.19 and 0.33, the low-level performance of the model is acceptable. If the value is within the interval of 0.33 to 0.67, the model's acceptability can be deemed moderate. On the contrary, when the value surpasses 0.67, the model's acceptability can be deemed to be at a good level. The present study revealed that the endogenous variables, namely the transtheoretical model (TTM), exhibited an R-squared value of 0.776. It is noteworthy that the R-squared values surpassed the required threshold, therefore showing that the model adequately reflected the gathered data. This observation suggests a good fit for the model, as depicted in Table 9. The second component is concerned with the Standardized Root Mean Square Residual (SRMR) measure, which should not exceed 0.08 [133, 134], with an SRMR value of 0.050 being acceptable. The third component is the Goodness of Fit (GoF) value, which is derived by taking the square root of the product of the average determination coefficient (R²) and the average variance extracted (AVE). The acceptable bounds for the GoF value are as follows: less than 0.1, signifying no fit; between 0.25 and 0.25, signifying a moderate fit; and greater than 0.36, signifying a high fit [135]. Exploiting the same measurement technique is based on defined methods and results [136-138]. The outcome of this experiment yields a substantial Goodness-of-Fit (GoF) rating of 0.769. The result of the goodness-of-fit (GoF) may be computed using the equation shown as Eq. (1).

(1)