Association of health literacy and diabetes self-management: a systematic review

Additional keywords: randomised control trial, type 2 diabetes mellitus.

The International Diabetes Federation (IDF) recently estimated that ~425 million people aged 20–79 years were living with type 2 diabetes worldwide in 2017, and this is expected to increase to 629 million by 2045 (Suvi et al. 2017). In 2017, diabetes caused ~4 million deaths and cost over USD 727 billion for people with diabetes, which is equivalent to 12.5% of the total healthcare budget worldwide (Suvi et al. 2017). Type 2 diabetes mellitus (T2DM) accounts for almost all diabetes (Suvi et al. 2017), which is commonly associated with ageing and low physical activity (Schwarz et al. 2012; Edwards and Hosseinzadeh 2018).

Self-management is the primary means for controlling T2DM and its burden on healthcare systems and patients (Funnell et al. 2011; Ansari et al. 2019). T2DM self-management includes physical activity, healthy eating, medication adherence, blood glucose monitoring and diabetes self-care-related problem solving (Funnell et al. 2011; Almutairi et al. 2019). Health literacy (HL) is one of the major drivers of self-management behaviours and refers to the ‘cognitive and social skills which determine the motivation and ability of people to gain access to, understand and use information in ways which promote and maintain good health’ (Nutbeam 1998). HL is integral to patient empowerment, knowledge and self-management skills (Yadav et al. 2019). Low HL is associated with poor health outcomes and inability to use available healthcare services (Kim and Lee 2016; Ho et al. 2018). Inadequate HL is independently associated with worse glycaemic control, low medication adherence and higher rates of retinopathy (Schillinger et al. 2002). More specifically, T2DM patients with high levels of HL are more likely to cope with diabetes self-management challenges compared to those with low HL (Powell et al. 2007; Niknami et al. 2018). As such, exploring this association in more depth to find more efficient HL interventions to improve T2DM outcomes is needed. This systematic review aims to provide evidence to inform T2DM self-management planners and policy decision-makers. To achieve this, this review analyses the existing evidence about the effect of HL on T2DM self-management behaviours by assessing the most current randomised controlled trials.

Literature searches were conducted in PubMed, CINHAL full-text plus, Scopus, Medline and Web of Science databases. Grey literatures such as government reports, conference proceedings and dissertations were also searched. Search terms included ‘Health Literacy’ AND ‘Diabetes’ OR ‘Type 2 Diabetes Mellitus (T2DM)’ AND ‘Self-Management’ OR ‘Self-Care’.

To review the most current studies, papers published between 2009 and 2018 in the English language were assessed. Only randomised control trials (RCTs) were included. Articles with a health education focus were also included if health education was interchangeably used with HL and the measurement of health education was similar to HL.

Studies were excluded if they were not peer reviewed or did not measure HL. The Preferred Reporting Item for Systematic Review and Meta-analysis (PRISMA) flow diagram was used to select eligible articles (Fig. 1). The authors (P.K. Dahal and H. Hosseinzadeh) independently reviewed all of the retrieved abstracts and selected eligible papers. Any disagreements were resolved by discussion. Data extraction from the selected papers was carried out by P.K. Dahal and checked by the lead author (H. Hosseinzadeh). Cochrane Back Review Group assessment criteria were used to assess the quality of the trials (Furlan et al. 2009).

Fig. 1.  Preferred Reporting Item for Systematic Review and Meta-analysis (PRISMA) flow diagram.

Initially, 6517 articles were identified through searching for the key search words from the five databases. After twice scanning, 44 full-text articles were assessed for eligibility. Fourteen of them met the inclusion criteria (Fig. 1).

As outlined in Table 1, all of the selected studies were RCTs. For instance, Bowen et al. (2016) conducted a three-arm RCT; Miller et al. (2014) conducted a RCT with two parallel interventions; and Graumlich et al. (2016) conducted a two-arm RCT. Target population of all of the studies were individuals living with T2DM aged ≥18 years. Two studies (Bowen et al. 2016; Grillo et al. 2016) focused on primary healthcare centres (PHCCs); seven studies (Hill-Briggs et al. 2011; Rosal et al. 2011; Kim et al. 2015; Graumlich et al. 2016; Cortez et al. 2017; Lee et al. 2017; Wichit et al. 2017) focused on community centres; four studies (Shi et al. 2010; Beverly et al. 2013; Miller et al. 2014; Cheng et al. 2018) focused on hospital settings and one study focused on different international healthcare systems (Muller et al. 2017).

Table 1.  Summary of selected review articles
RCT, randomised control trial; PHC, primary health care; HL, health literacy; T2DM, type 2 diabetes mellitus; UK, United Kingdom; USA, United State of America; DSME, diabetes self-management education

All of the trials examined the effect of health literacy or education program on T2DM self-management. Eleven studies have emphasis on HL intervention through a diabetes education program (Shi et al. 2010; Hill-Briggs et al. 2011; Rosal et al. 2011; Beverly et al. 2013; Miller et al. 2014; Bowen et al. 2016; Grillo et al. 2016; Cortez et al. 2017; Lee et al. 2017; Wichit et al. 2017; Cheng et al. 2018), whereas Kim et al. (2015) focused on the effect of structural behavioural education; Muller et al. (2017) focused on a web-based intervention and Graumlich et al. (2016) explored a Medtable-based intervention (a medication-planning tool designed to improve patient self-management). The main outcome measures included self-efficacy; glycaemic control; self-medication; diabetes-related quality of life; and self-blood glucose monitoring, knowledge and attitude towards T2DM, physical activities and diet management (Table 1).

As shown in Table 2, the effect of HL on glycaemic control as well as self-monitoring of blood glucose was inconclusive. This is because only one out of six trials measuring blood glucose level showed that HL was associated with significant improvements in glycaemic control (Graumlich et al. 2016). Similarly, only two out of four trials examining self-monitoring of blood glucose (Shi et al. 2010; Rosal et al. 2011; Lee et al. 2017; Cheng et al. 2018) demonstrated that HL is a significant predictor of blood glucose self-monitoring behaviours (Shi et al. 2010; Lee et al. 2017).

Table 2.  Relationship between health literacy or education characteristics and self-management
RCT, randomised control trial; NA, not available; DM, diabetes mellitus; DSM, diabetes self-management

In terms of diabetes knowledge, HL improved diabetes knowledge in all studies measured diabetes knowledge (Hill-Briggs et al. 2011; Rosal et al. 2011; Miller et al. 2014; Kim et al. 2015; Grillo et al. 2016; Cortez et al. 2017; Muller et al. 2017; Wichit et al. 2017), except in one study by Graumlich et al. (2016).

Six studies examining diabetes self-efficacy (Shi et al. 2010; Rosal et al. 2011; Kim et al. 2015; Bowen et al. 2016; Lee et al. 2017; Wichit et al. 2017) showed HL significantly improved diabetes-related self-efficacy. Quality of life measured by three studies (Beverly et al. 2013; Kim et al. 2015; Wichit et al. 2017) was also positively associated with HL. Physical activities tested by four studies had significant association with HL (Beverly et al. 2013; Lee et al. 2017; Muller et al. 2017; Cheng et al. 2018). Treatment satisfaction examined by Bowen et al. (2016) was improved significantly by a HL intervention program. One of the two trials examining diet management (Grillo et al. 2016; Lee et al. 2017) indicated that HL was able to generate significant improvements in diet behaviours (Grillo et al. 2016). The associations of HL with foot care and medication adherence were also inconclusive. A study by Lee et al. (2017) reported that HL did not improve foot care and medication adherence, whereas Cheng et al. (2018) demonstrated that HL led to significant improvements in both foot care and medication adherence.

HL interventions delivered in community settings generated more promising results in knowledge acquisition, self-empowerment and clinical outcomes. In a Baltimore HL study involving patients from community practice sites in underserved areas, significant improvements were evident post-intervention in terms of problem solving and self-care behaviour (Hill-Briggs et al. 2011). The intervention group in a HL RCT in a community setting experienced statistically significant improvement in haemoglobin A1c (HbA1c) compared with the control group (Cortez et al. 2017). Even secondary outcomes such as self-care, attitude, knowledge and empowerment improved significantly in the intervention group (Cortez et al. 2017). Community settings also appeared to be more supportive of improving diabetes self-related quality of life, self-efficacy and adherence to a diabetes self-management regime and health literacy (Rosal et al. 2011; Kim et al. 2015; Lee et al. 2017; Wichit et al. 2017). Kim et al. (2015) concluded that community-partnered interventions have great potential to achieve promising clinical and psycho-behavioural outcomes among underserved communities. Further, a randomised HL trial among outpatients in a community showed that participants who received Medtable had greater knowledge about diabetes medications versus the usual care group regardless of the HL status (Graumlich et al. 2016). Similarly, a community-based HL trial among a Latino community reported that the HbA1c level, diabetes knowledge, self-efficacy and blood glucose self-monitoring improved significantly among the intervention group compared to the control group after 12 months across five health centres located in community settings (Rosal et al. 2011).

This study aims to summarise the existing evidence about the effects of HL on T2DM self-management behaviours using recent RCTs.

Our findings suggest that HL leads to significant improvements in T2DM self-management behaviours, including diabetes knowledge (Grillo et al. 2016; Muller et al. 2017), physical activity (Beverly et al. 2013), self-efficacy (Shi et al. 2010) and diabetes-related quality of life (Wichit et al. 2017). However, the associations of HL with glycaemic control, self-monitoring of blood glucose, foot care, diet management and medication adherence was inconclusive. Only one out of six trials measuring blood glucose level showed that HL was associated with the glycaemic control (Graumlich et al. 2016). Two out of four trials examining self-monitoring of blood glucose demonstrated that HL was associated with blood glucose self-monitoring behaviours (Shi et al. 2010; Lee et al. 2017). One of the two trials examining diet management (Grillo et al. 2016; Lee et al. 2017) indicated that HL was able to generate significant improvements in diet behaviours (Grillo et al. 2016). Similarly, one of the two trials testing diet management (Lee et al. 2017; Cheng et al. 2018) showed that HL significantly improved both foot care and medication adherence (Cheng et al. 2018). This is consistent with previous systematic reviews, which suggested that there is discrepancy among studies in terms of the association of HL and T2DM health outcomes (DeWalt et al. 2007; Al Sayah et al. 2013; Hosseinzadeh and Shnaigat 2019). Our finding suggests that customised, structured (Beverly et al. 2013) and community-based HL interventions are more likely to empower patients and yield significant improvements in T2DM self-management behaviours (Rosal et al. 2011; Hill-Briggs et al. 2011; Kim et al. 2015; Graumlich et al. 2016; Cortez et al. 2017; Lee et al. 2017; Wichit et al. 2017). This might be explained by the fact that tailored and community-oriented HL interventions are more likely to address patients’ real needs, learning styles and cultural values, which are essential for providing integrated care (Crengle et al. 2018).

This study has several limitations. The study findings are based on heterogeneous trials with varying target populations, study settings, intervention types and outcomes. Further, the long-term effect of the interventions cannot be verified. This is because the study outcomes are often reported for less than 1 year. In some cases, it was very difficult to determine whether the significant outcomes in T2DM self-management behaviours caused by HL are due to using multiple-component interventions. Therefore, these findings should be generalised with caution.

Our review shows that HL leads to significant improvements in T2DM; however, its associations with glycaemic control, self-monitoring of blood glucose, foot care and medication adherence was inconclusive. Given the significance of HL in self-management, further studies are required to explain and address any inconsistency in relation to the role of HL in T2DM self-management behaviours.

More importantly, this review strongly suggests that customized and community-based HL interventions are more likely to yield promising T2DM self-management outcomes. This warrants more studies to understand further why these types of HL programs are more effective to inform T2DM self-management planners, patient educators, clinicians and policy decision-makers.

The authors declare that they have no conflicts of interests.