A cross-sectional survey was conducted among ultrasound diagnosed adult (above the age of 18) NAFLD patients who visited the fatty liver clinic at the Tel Aviv Medical Center during 2011 and 2012. Eighty-seven of these individuals participated in a clinical trial testing the effect of resistance training (RT) vs stretching on NAFLD[15]. Exclusion criteria included positivity for serum HBsAg or anti-HCV antibodies; fatty liver suspected to be secondary to hepatotoxic drugs; known diabetes treated by medications other than Metformin; cancer; renal failure; heart disease; non-compensated cirrhosis; inflammatory bowel disease; excessive alcohol consumption (≥ 30 g/d in men or ≥ 20 g/d in women)[2]; positive markers of genetic or immune-mediated liver diseases.

The study conformed to the ethical guidelines of the 1975 Declaration of Helsinki, and was approved by the Tel-Aviv Medical Center ethics committee. All patients signed an informed consent. The clinical trial was pre-registered in the NIH registration website (TRIAL no. NCT01264198).

Patients were asked to complete a self-report structured questionnaire. The questionnaire was assembled by the Israeli Center for Disease Control and was used in the first Israeli National Health Survey (MABAT)[16]. Following a 12-h fast, each participant of the clinical trial underwent blood tests, analyzed at the same lab.

Nutrition habits were measured using six yes/no questions (0 = no, 1 = yes) adopted from the national survey questionnaire tailored for the Israeli population[17]. Individuals’ nutritional habits score was calculated based on the patient’s answers (0-6, where 0 represents poor nutritional habits and 6 represents excellent nutritional habits). Participants in the clinical trial completed a detailed semi-quantitative food frequency questionnaire (FFQ) reporting their habitual nutritional intake during the past year. The FFQ was assembled by the Food and Nutrition Administration, Ministry of Health as previously described[18].

The main independent variables were measured by questions adopted from validated questionnaires as follows (the questions included are specified in Table 1): Self-efficacy was measured by adopting the Self-Efficacy Scale questionnaire[11,19], which was modified for the aims of the current study, referring to both physical activity and healthy dietary habits and keeping a food diary. The questions included “how confident you are that you could do things like these consistently, for at least six months. Ratings were made on a 5-point Likert-type scale ranging from “Sure I could not do it” (1) to “Sure I can do it” (5). The patient’s self-efficacy rating was calculated as the mean of 7 questions on a scale from 1 to 5 (α = 0.785, 5 represents high level of self-efficacy).

The questions regarding disease and treatment perceptions were adopted from the brief illness perception questionnaire[20], including: emotional representation (mean of 2 questions number 5, 6), illness understanding degree (a single question number 4), timeline perception (a single statement number 1), treatment perception (a single statement number 2) and symptoms (a single question number 3).

In addition, Illness consequences were measured based on questions adopted from the Personal Models of Diabetes Interview (PMDI) questionnaire[14], which is based on the Health Belief Model (HBM)[21,22] and the Self-Regulation Model of illness (SRM)[22,23] (mean of 4 questions number 7-10, α = 0.722).

All questions were measured using a scale from 1 to 5 (e.g., 5 represents: high level of illness emotional representation, high level of perceived understanding, perceiving NAFLD as a long-term disease, a high level of symptoms experienced severe perceived consequences and high belief in treatment). In addition, the questionnaire included an open-ended question[20], “Please list in rank-order the three most important factors that you believe caused your NAFLD”. Respondents’ answers were categorized into six main reasons (see Figure 1).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Patients’ perceptions regarding the main reasons for non-alcoholic fatty liver disease.

Statistical analyses were performed using SPSS version 22 (SPSS Inc., Chicago, IL, United States). Continuous variables are presented as mean ± SD. Independent samples t-test were performed to test differences in continuous variables between the two groups. The Mann-Whitney test was used if non-parametric tests were required based on data distribution. Pearson’s χ² tests were performed to examine associations between nominal variables. Pearson or Spearmen correlations were used to test correlations between the diet score, self-efficacy and disease perception.

A path analysis was performed, using AMOS version 22 software (IBM Corp., SPSS, 2013), to describe the interrelationships between the study variables and assess the extent to which the data fit a predicted model associated with nutritional habits. This is a useful methodological and statistical tool to assess the fit between a set of variables and a theory. This method estimates the unidirectional relationship between a set of independent predictors and a set of dependent criteria. A model-modification approach is used to estimate existing relationships and to evaluate and improve them. The final result is a model that, on the one hand, is theoretically based and, on the other, fits the specific set of data.

Model fit was assessed with NFI, NNFI, CFI, and RMSEA. Usually, two statistical criteria are applied: a measure of the model fit (represented by χ² and its derivates) and significant estimates of the interrelationship of the model variables (represented by Bs) (Arbuckle, 2013). The most common measures of model fit, χ², should be non-significant in order to represent a non-significant departure from the best model possible; the Fit Indexes (NFI - normal fit index, NNFI - non-normal fit index, CFI - comparative fit index) should be close to a value of 1. RMSEA (Root mean standard error of approximation) should be lower than 0.05. Estimates given by the model (β) should be significant[24]. P < 0.05 was considered statistically significant for all analyses.

The study sample included 146 NAFLD patients, 54.1% men, with a mean age of 47.76 ± 11.68 years (range 19-72), mean body mass index (BMI) of 31.56 ± 4.62, 25 patients (28.7% out of the 87 subjects with blood tests) had the metabolic syndrome according to the accepted criteria[25] and 13 patients (8.9%) had type-2 diabetes, all were treated solely by Metformin. The average nutrition habits score was 4.73 ± 1.45 on a scale between 0-6, where 6 represents the healthiest eating habits. Table 2 describes the study sample characteristics.

Table 1 presents the patients’ perceptions regarding different aspects of having NAFLD. Most of the study participants (57.2%) did not feel they fully understood what having a NAFLD really meant. Only a little more than half of the patients (53.6%) perceived NAFLD as a long-term condition. The majority (72.6%) believed there was some form of treatment that could help healing or reducing the severity of NAFLD.

With regard to emotional representation, 60% of the patients reported high levels of concern regarding having a NAFLD, but only 20% reported that having a NAFLD greatly affected them emotionally. In addition, with regard to illness consequences, almost 60% believed that NAFLD is a severe health problem, but only half of the patients believed NAFLD is a disease. Furthermore, only a little more than half of the patients (53.4%) anticipated having medical complications due to NAFLD. Only 13.7% felt that having NAFLD severely affected their everyday life.

The majority of the patients reported relatively high self-efficacy regarding their ability to maintain a healthy diet. For instance, 72.6% of participants reported high levels of self-efficacy to maintain a low sodium diet, 71.9% reported high levels of self-efficacy to maintain a low fat diet, and 67.1% reported high levels of self-efficacy to maintain a low sugar diet. The overall self-efficacy mean rating of the patients was 3.76 ± 0.72 on a scale of 1-5, which indicates medium-high self-efficacy rating.

The majority of the patients reported maintaining healthy eating habits. Most patients reported making efforts to reduce dietary intake of fat (88.8%), calories (67.5%), sugar (83.3%), soft drinks (89.5%), and red meat (65.1%), and to increase intake of vegetables (79.4%).

Figure 1 presents the distribution of the perceived main reasons for having NAFLD. Most patients (73.7%) believed that behavioral factors were the most important reasons for NAFLD; approximately 54% reporting dietary habits as the main reason for fatty liver, followed by 13.9% who reported overweight or obesity and 6.3% who reported lack of physical activity. The second most prevalent perceived reason was genetics at 16%. Medication use was cited as the main reason for fatty liver by 4.2% of respondents.

Table 3 presents bivariate correlations between the study variables. Maintaining healthy nutritional habits was significantly positively correlated with the patients’ perceptions regarding their illness consequences (r_s = 0.19; P = 0.035), the degree of understanding their illness (r_s = 0.27; P = 0.003) and their self-efficacy (r_s = 0.20; P = 0.028). Self-efficacy was significantly negatively correlated with timeline perceptions (r_s = -0.27; P = 0.001).

There was a significant, strong positive correlation between the perception of illness consequences and high emotional representation (reflecting the extent by which the patients were afraid or concerned about having NAFLD; r_s = 0.58; P≤ 0.001), timeline perceptions (reflecting the extent by which the patients perceived NAFLD as a long term illness, r_s = 0.21; P = 0.011), symptoms perception (r_s = 0.45; P≤ 0.001) and perceived Illness understanding degree (r_s = 0.38; P≤ 0.001).

A path model was developed to describe the interrelationships between the study variables and assess the extent by which nutritional habits can be predicted by them.

The model was found to fit the data: NFI = 0.870, NNFI = 0.985, CFI = 0.992, RMSEA = 0.018, and χ²(29) = 30.413, P = 0.394, as shown in Figure 2 and in Supplementary Table.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Path model for the study variables predicting nutritional habits. Values attached to the arrows represent regression standardized coefficients (β values), with significance levels in asterisks. Values within the rectangles represent the explained variance by the variable (R²). NFI = 0.870, NNFI = 0.985, CFI = 0.992, RMSEA = 0.018, χ²(29) = 30.413, P = 0.394. ^aP < 0.05, ^bP < 0.01, ^eP < 0.001 NFI: Normal fit index; NNFI: Non-normal fit index; CFI: Comparative fit index; RMSEA: Root mean standard error of approximation.

Better nutritional habits were positively and directly predicted by illness understanding degree (β = 0.26; P = 0.002), self-efficacy (β = 0.25; P = 0.003) and timeline perception (longer-term illness) (β = 0.17; P = 0.036).

Several indirect significant associations were observed. First, male gender was negatively related with illness understanding (β = -0.46; P = 0.014) and self-efficacy (β = -0.24; P = 0.031). Meaning that, women perceived higher illness understanding and higher self-efficacy. Second, BMI was positively related with a perception of more severe illness consequences (β = 0.27; P = 0.002). The latter was related with a higher perception of illness understanding and a perception of a longer term illness, both of which were positively predictive of better nutrition habits. However, perceptions of more severe illness consequences were strongly related with higher emotional representation (β = 0.55; P < 0.001), which in turn was related with lower self-efficacy (β = -0.17; P = 0.034), predicting worse nutritional habits. Third, although the perceptions of a longer term illness were directly related with better nutritional habits, these perceptions were also negatively related with the perception of effective treatment (β = -0.31; P < 0.001). The perception of treatment effectiveness was positively related with self-efficacy (β = 0.32; P < 0.001), which predicts better nutritional habits. Fourth, experiencing many symptoms was related with lower perceptions of illness understanding (β = -0.19; P = 0.026) and may lead indirectly to a higher illness emotional representation. However, only 9% of the sample experienced significant NAFLD-related symptoms.

A sub-sample of the study population (n = 84) completed FFQ. There was a positive correlation between healthy eating habits scores and fiber intake (r = 0.26, P = 0.027), and there was a negative correlation between healthy eating habits scores and saturated fat intake (percent of saturated fat calories from total calories) (r = -0.43, P < 0.001).

In accordance with the correlation between self-efficacy and the perceived diet score, self-efficacy was also correlated with nutrient intake evaluated by the FFQ; negatively with saturated fat (percent of saturated fat calories from total calories) (r = -0.28, P = 0.010) and positively with fiber (r = 0.22, P = 0.047) and vitamin C intake (r = 0.34, P = 0.002).

In a sub analysis of RCT participants who in the intervention arm (resistance training) who had regular training sessions at the gym (n = 34). The number of training sessions at the gym was automatically recorded every time a patient entered the gym with his/her personal chip. Compliance was defined as the number of recorded training sessions divided by the recommended number of sessions (i.e., 3 times a week during the 12-wk trial). This objective compliance measure was positively correlated with the self-efficacy level (r = 0.34, P = 0.046).

Non-alcoholic fatty liver disease (NAFLD) is emerging globally as the most prevalent liver disease. Poor dietary habits represent a main modifiable target for the primary prevention and treatment of NAFLD. Thus, efficient and sustainable lifestyle modification programs are needed for NAFLD patients. However, building and implementing such programs may be difficult without adequate knowledge about disease and treatment perceptions of NAFLD patients and their association with self-efficacy to execute lifestyle changes.

The factors associated with self-efficacy among NAFLD patients have never been tested. Furthermore, the association between self-efficacy, along with disease and treatment perceptions, and keeping a healthy lifestyle among NAFLD patients has not been studied.

In the present study, the authors demonstrated that nutritional habits and lifestyle modification among NAFLD patients may be associated with the patients’ perceptions concerning their illness, and with their self-efficacy regarding their ability to change their lifestyle. Current results highlight the importance and the complexity of achieving NAFLD patient’s adherence to healthy nutritional habits.

Understanding NAFLD patients’ cognitive representations of their disease can help in developing NAFLD-tailored lifestyle interventions. The model suggests ways by which clinicians can improve NAFLD patient’s eating habits, or, on the other hand, may unintentionally discourage them from trying. The main conclusions that can be drawn are that higher perceptions of understanding the illness and a higher self-efficacy are positively related to better nutritional habits, and therefore its enhancement should be part of the behavioral treatment. Conversely, “scaring” the patients and leading them to believe that NAFLD has severe consequences may lead to the undesirable outcome of lower self-efficacy. In addition, it may be important to emphasize to patients that although NAFLD is a chronic condition, it is still treatable and even reversible if diet is maintained.

Self-efficacy, defined as beliefs in one’s capabilities to organize and execute the courses of action required for producing given attainments. The health Belief Model (HBM), indicates that perceived risk, severity, or threat of disease among patients as well as their confidence to make lifestyle changes may be critical to motivate patients into chancing their eating and physical activity habits.

The topic is quite interesting and the paper is very well written, with clear language and discussion consistent with the results obtained. The study by Zelber-Sagi et al has investigated the relation between illness perception and dietary modification among 146 patients with NAFLD. The study has shown significant correlations between disease perception and dietary habits among NAFLD patients, suggesting self-efficacy enhancement as an effective tool for improving illness understanding, thus treatment effectiveness. Aim of the study has been clearly stated. Data relevant to the topic have been precisely presented and discussed in detail. Statistical methods have been meticulously mentioned in the text.