Health literacy dimensions among public health service users with chronic diseases in Piracicaba, Brazil, 2019

Aim: This study analyzes factors associated with dimensions of health literacy (HL) functional, communicative and critical among public health service users with chronic non-communicable diseases. Methods: A cross-sectional analytical research was carried out in Piracicaba, São Paulo, Brazil, with adults and older adults attending Family Health Units (FHU). Data were collected by oral exam (CPOD and CPI) and a questionnaire on systemic conditions, sociodemographic factors, health behaviors and HLS (HLS-14). The outcomes consisted of functional, communicative, and critical HL dimensions dichotomized by median (high and low), which were analyzed by chi-square test (p<0.05) to find associations with the variables studied. Results: The study sample comprised 238 FHU users with 62.7 (± 10.55) mean age, of which 47.5% (n=113) showed high functional HL, 50.0% (n=119) high communicative HL, and 46.2% (n=110) high critical HL. High functional HL was associated with men (p<0.05). Functional and communicative HL were associated with having higher education (p<0.001 and p=0.018, respectively). High communicative and critical HL were associated with regular use of dental and medical services (p<0.05). Individuals with low functional HL were more likely to present poor tooth brushing (p=0.020). High HL (in all three dimensions) was associated with regular flossing and having more teeth (p<0.05). Conclusion: Functional, communicative and critical HL were associated with health behaviors and clinical outcomes, whereas the functional dimension was also associated with sociodemographic factors. HL dimensions allowed to differentiate health-related factors.


Introduction
The demographic and epidemiological transition has widened the age pyramid and increased the prevalence of chronic non-communicable diseases (NCD) in the world population and, subsequently, in the Brazilian population 1 . Of strong behavioral character, these morbidities require co-responsibility between health professionals and patients to control their consequences. In this regard, health literacy (HL) has been considered a key to health promotion and to improve health decision-making 2 .
Health Literacy refers to personal knowledge, motivation, and skills to make health decisions throughout life 2 . According to Nutbeam's concept, HL comprises three dimensions: functional, communicative, and critical literacy 3 . Functional HL consists of sufficient basic reading and writing skills to be used in everyday situations. In this dimension, one's role is passive. In the communicative dimension, one seeks information through direct communication with reliable sources, such as health professionals, thus playing an active role. The critical dimension requires more advanced cognitive skills, such as critical analysis to judge whether a health information is appropriate and represents a greater control over one's own health, requiring a proactive role 3,4 .
A low HL can have an impact on people's health 5 , representing difficulties in making health-related decisions. Studies suggest that adults and older individuals with low HL have less access to and understanding of health information, use medications inappropriately, have less disease prevention and control, with higher rates of morbidity and hospitalization 3 . Conversely, a high HL means being able to take responsibility for the collective health and one's own 6 .
Measuring health literacy remains a challenge for health professionals and managers, especially regarding the elaboration of strategies for developing critical thinking 6 . Most of the existing instruments for HL measurement target specific health conditions such as oral health 7 or diabetes 8 , and few of them take on a multidimensional approach 9,10 . Most measure only functional HL 11 and one more dimension 12 .
Using an instrument able to assess the three HL dimensions, as described by Nutbeam, would thus allow us to identify the different aspects that might interfere in how people manage their health. In a community approach, information on the associated factors of HL dimensions contribute to assist and to plan health strategies in different health contexts, health conditions and/or age groups 13 . The Health Literacy Scale (HLS-14) 10 , for example, is a validated instrument that presents three dimensions 14 .
Given this context, this study sought to analyze the socio-demographic variables, health behaviors and clinical factors associated with HL dimensions among public health service users with chronic non-communicable diseases.

Study design and location
A cross-sectional analytical study was carried out with users of the Unified Health System (SUS), with follow up at Family Health Units (FHU) in the municipality of Pira-cicaba, São Paulo, Brazil, using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines 15 .

Location
According to the last census (2010), Piracicaba has a population of 364,571 inhabitants in the urban area, with an adult and older population of 261,567 16 .
In 2018, the municipality's health network had 71 Basic Health Units, of which 51 were FHUs. This study included only adult and older adult hypertensive and/or diabetic users.

Sample
We performed a sample calculation considering the prevalence of low HL as 50%, based on Puello (2018) 17 , with a margin of error of 0.1 and design effect (deff)=2. Predicting probable losses, we added 20%, totaling 298 participants.

Sample selection
Selection took place in two stages: first, we chose the FHU and then the participants. We performed a probabilistic drawing of eight Family Health Units and then four alternates, considering the number of hypertensive and/or diabetic users in the population registered at the FHU, according to a study by Morgan (2013) 18 . After two FHU refused to participate, two of the alternates were included. We had to include the remaining two alternate FHU to reach the sample size, thus totaling a final sample of 10 participating FHU (Figure 1). Estimating possible losses and refusals, we added 10 participants for each selected FHU, with 40 users taken from the list of hypertensive and/or diabetic patients registered at each health unit. 8 Figure 1. Distribution of the sample of NCD patients (type 2 diabetes and hypertension) and the FHUs selected for the study, adapted from Morgan (2013).
The health teams of each selected FHU randomly distributed 40 invitations to registered users with type 2 diabetes and/or SAH in attendance on the scheduled dates to participate in the study. Data collection took place at the FHU during its opening hours.
Inclusion criteria consisted of patients registered and monitored at the FHU in Piracicaba for type 2 diabetes and/or SAH, who attend the FHU on the scheduled day and time. Exclusion criteria included presence of abscesses or emergency oral health care on the day of collection, refusal to undergo clinical dental examination, and being unable to answer the questionnaire due to physical and/or psychological status (informed by the respective FHU).

Data collection
Clinical data were collected by a dental surgeon (DS) after an 8-hour theoretical and practical training with an experienced examiner, with intra-examiner agreement which, considered within reliability standards, ranged from 90.6% to 100.0% for caries and periodontal disease 19,20 .
Clinical oral examinations were performed by the examining board, properly dressed, and under World Health Organization (WHO) criteria, using a sterile periodontal probe and a clinical mirror, with the participant sitting in a chair, under natural light 21 , at the FHU offices. The clinical conditions evaluated were visible dental biofilm 22 , index of decayed, lost and filled permanent teeth (DMFT) and Community Periodontal Index (CPI) 21 .
Interviews were conducted with the participants following a questionnaire with 66 objective questions about behaviors, oral and general health determinants 20,23 .
Subsequently, we applied the Health Literacy Scale (HLS-14) 10 validated in Brazilian Portuguese 14 . This instrument presents 14 questions (5 for the functional and communicative dimensions and 4 for the critical dimension), answered by a 5-point Likert-type scale, with the following categories: "strongly disagree," "disagree," "neither agree nor disagree," "agree" and "strongly agree." Total score ranges from 14 to 70 points, with higher scores indicating better HL. In the functional dimension (questions 1 to 5) the score is reversed, where agreeing means having low HL, whereas the questions related to communicative (questions 6 to 10) and critical literacy (questions 11 to 14) refer to high HL 10 .
Data on blood pressure and glycemic indexes were collected from the current information in the medical records.
Application of the questionnaire and HL instrument and the clinical oral examination were performed on the same day.

Study variables
Our variable of interested was HL, presented, in each dimension, at two levels: low and high, dichotomized by the median. Cutoff points for high and low levels were 11.0 for the functional dimension, 16.5 points for the communicative dimension, and 14.0 for critical literacy. Figure 2 summarizes the three dimensions 3 and roles 4 , namely: functional HLpassive role, communicative HL -active role, and critical HL -proactive role.  The study variables were grouped into sociodemographic, behavioral, and clinical data. Sociodemographic data consisted of age (considered continuously), gender (man or woman), and schooling level (less than 4 years, 4 full years, or 5 years or more), the cutoff point being elementary school 20,23 .

Results
A total of 238 users with chronic diseases participated in the six-month data collection period. Two users refused to undergo clinical oral examination, and a sample loss characterized by the non-attendance of 162 invited users, which was expected and calculated in the sample size and selection method.
Mean age was 62.7 (±10.55) years old, and 78.5% (n=187) had lower schooling level. Regarding health behaviors, 68.1% (n=162) of the patients flossed regularly, and 74.8% (n=172) made irregular use of dental services (+1 year). As for the oral clinical exams, 57.6% (n=147) of participants presented a periodontal pocket > 4mm (Table 1).   HL dimensions were associated with the sociodemographic, behavioral, and clinical variables. The bivariate analysis showed that having more than 20 teeth and regular flossing were associated with high HL in all three dimensions (Table 3).

Discussion
Our study highlighted different associations between the dimensions of health literacy (HL) and sociodemographic factors, health behaviors, and clinical outcomes. A multidimensional evaluation of HL provides a broader approach that can deepen our understanding regarding HL levels and enhance one's health autonomy. Hence, the differential of a multidimensional instrument used to increase measurement sensitivity is evident, allowing more variables associated with the construct to be identified. Despite the research on validated HL tools, few studies have assessed HL dimensions and associated factors 27 .
The Health Literacy Scale (HLS-14), validated in Brazilian Portuguese, showed good internal consistency, which is considered adequate when greater than or equal to 0.70. Its psychometrics properties were satisfactory to evaluate health literacy, as showed by Batista et al. 14 .
Recent studies using HL instruments associated with NCDs, including oral diseases 27 , have assessed mainly reading and writing skills 28 , that is, only the functional dimension, disregarding communication and/or broad interaction with health care systems.
In our study, therefore, we chose to use the HLS-14 instrument, a pioneering tool for measuring the three HL dimensions (functional, communicative, and critical), according to Nutbeam (2000) 3 . Rapidly applicable, with reliable psychometric indexes not restricted to a specific area or health condition 10 , it can serve both to define clinical protocols more consistent with reality, thus improving people's level of understanding of health information, and to carry out interventions capable of improving health literacy 29 . The questions with the greatest impact on literacy inquired about the difficulty in reading and finding information when needed, and in communicating one's opinion about a health condition and being able to judge whether the information is reliable.
Regarding sociodemographic factors, men showed greater functional literacy, result not found in other studies 30 . Gender inequity is an important social marker in Brazil, especially in a sample of predominantly older adults. This finding may indicate a lack of study opportunities in a generation where these opportunities, including decision-making, were restricted for women. Today, as observed in the 2010 census, women have a high level of schooling, with female school attendance increasing 9.8% in high school compared to men 16 . Studies also highlight that older adults may have limited understanding of health information 31 and greater participation of women due to the feminization of the aging process 32 . But even with this limited functional literacy, the literature points to greater self-care among women, including regular use of health services 33 . Consequently, HL needs to go beyond the functional level.
Our findings showed that high functional and communicative HL were associated with high schooling level. These HL dimensions are related to passive and more active attitudes, such as communication. However, we must consider the cognitive differences, skills, and roles between people with the same educational level 34 . As such, research that exclude illiterate individuals from its sample 30 may lose heterogeneity of results and restrict the understanding of literacy dimensions after all, literacy is one of and not the only aspect analyzed by HL dimensions. Studies show that functional literacy focuses on reading skills, in which the people act more passively in health-related issues 4 . Better reading and comprehension skills are associated with better formative education, which is related to schooling level, a marker and social determinant of health 35 . HL is thus related to one's schooling, reflecting on their health behaviors; consequently, developing health literacy can reduce health inequalities 36 .
Regarding oral health behaviors, our results revealed that regular flossing was associated with high levels of all HL dimensions. Lower frequency of tooth brushing was associated with low functional HL, corroborating a recent study 37 . Oral health care and use of dental services can have an impact on clinical health conditions 13 .
In our study, regular use of dental services was also associated with communicative and critical HL. The literature points out that, besides greater use of services, individuals with higher HL seek preventive consultations, showing a more active role in the pursuit of health 4 . An unexpected finding in the present research was the association between low communicative and high critical HL and irregular use of medical services. This result can be explained by the sample characteristic of patients with chronic disease, who need continuous medical follow-up.
The presence of a periodontal pocket was associated with a low critical HL. Oral hygiene is associated with HL and with the risk of developing periodontal disease 38 , which can lead to tooth loss.
Tooth loss the worst oral health outcome was associated with all HL dimensions, but remains inconclusive 13 . When associated with risk behaviors for oral diseases, HL becomes relevant as a measure to reduce and control tooth loss, as it can help promote oral and general health, and studies exploring this topic have been performed 39 . Thus, an in-depth knowledge of one's HL level can be an important differentiator in the health-disease process 40 .
SAH was associated with communicative and critical HL, as shown by Borges et al. (2019) 30 . Considering that such dimensions of HL are associated with people's proactive abilities 4 , blood pressure indices may, in this case, be influenced by aspects that interfere with their discharge, such as: frequency, type, and access to health services, interaction with health professionals, and others 37 .
Since the outcomes of oral and general health diseases and aggravations, such as periodontal disease, tooth loss and SAH, are associated with more advanced dimensions of literacy, such as communicative and critical HL, it becomes clear that inequality negatively impacts health.
As for the limitations, we can cite the restricted sample of the study. Nonetheless, it was representative of Unified Health System users with SAH and diabetes, where important associations between the HL dimensions and aspects involving the integral health of these users, often neglected in research 2 , were contemplated. Measuring health literacy by a self-report instrument is always challenging, but using a validated questionnaire and proper analysis can control bias, thus improving the quality of the study. Despite the limitations, our exploratory study presents unprecedent results that show a new perspective regarding the application and analysis of health literacy dimensions, reaffirming the need for greater research interest in exploring and improving on this topic in future studies.
Improving population HL can reduce the prevalence of chronic health conditions and the individual and collective impacts of these morbidities. Our results showed that using instruments that cover only functional literacy may be insufficient to assess health literacy, and that the analyzes need to incorporate all three dimensions to formulate safer and more accurate strategies for professionals, managers, and users.
The present work contributes to greater attention to the complexity and challenges involved in advancing the topic, serving as a starting point for future studies and as an aid to evidence-based public health policies that seek to improve the health of SUS users. Thus, future studies should consider HL using a multidimensional approach for public health policies and health promotion strategies.