Depression is a major global public health problem both because of its high prevalence, with an estimated 300 million people of all ages affected (Friedrich, 2017), and because it carries the highest burden of disability among all mental disorders (Üstün et al., 2001). It is also responsible for the largest proportion of the burden of comorbidity and morbidity attributable to non-fatal health outcomes (Moussavi et al., 2007). The comorbidity between depression and chronic disease, especially diabetes has been widely reported in the literature (Khaledi et al., 2019). The risk of depression in a person with diabetes is approximately double that of a person without this disease (Anderson et al., 2001; Semenkovich et al., 2015), generating a very negative impact on emotional well-being, quality of life, and control of the disease, resulting in poorer diabetes outcomes (Egede and Ellis, 2010). A recent review (Khaledi et al., 2019) found that one in four adults with diabetes had depression and concluded that, given the high prevalence of depressive disorders in patients with diabetes, screening for comorbid depression and its prevalent risk factors in this population is recommended.

Scientific evidence on treatment for depression shows that depression can be successfully treated with a variety of psychological and pharmacological interventions, often implemented through collaborative and stepped care approaches (Cuijpers et al., 2008; Petrak and Herpertz, 2009; Baumeister, 2012). However, despite evidence recommending the combination of both treatments, pharmacological treatments remain the most common option in routine clinical practice. Regarding psychological interventions for depression, those that implement behavioral therapy and cognitive techniques have the highest efficacy and scientific evidence (Hofmann et al., 2012). On the other hand, treatment with antidepressant drugs is shown to be effective, but they differ substantially with respect to short- and long-term side effects (Hackett et al., 2008; Rayner et al., 2010; Moncrieff, 2011; Baumeister, 2012). For this reason, the National Institute for Health and Clinical Evidence (NICE) guideline on depression recommends psychological interventions as first-line treatments for treating depressive symptoms (National Institute for Clinical Excellence, 2009; Baumeister, 2012).

The effective management of diabetes requires a complicated and demanding treatment regimen where the patient must take an active role in his or her self-care, involving a high degree of responsibility for his or her disease and constant decision making about treatment. In addition, each type of diabetes has certain distinct clinical and treatment characteristics. Daily life with diabetes differentiates this disease from other chronic conditions and can generate routines associated with high levels of stress, which can lead to the appearance of depressive symptoms and the need for treatment. The appearance of such depressive symptomatology can also affect glycemic control; however, the data reported to date suggest a mainly indirect effect of depression on glycemic control due to poor self-care behaviors (Snoek et al., 2015). Therefore, treatment of depression in people with diabetes should be oriented toward improving both psychological and medical outcomes, according to the recommendations of the American Diabetes Association (Americam Diabetes Association, 2021).

Psychological interventions aimed at the treatment of depression in patients with diabetes are well-documented as being effective in treating depressive symptoms (Van der Feltz-Cornelis et al., 2010; Markowitz et al., 2011; Petrak et al., 2015). In contrast, studies on pharmacological treatment for depression in people with diabetes report inconclusive results (Baumeister, 2012; Baumeister et al., 2012; Petrak et al., 2015) and sometimes negative results (Lustman et al., 1997). However, reviews on the efficacy of both interventions on improving glycemic control obtained unsatisfactory results (Van der Feltz-Cornelis et al., 2010; Markowitz et al., 2011; Petrak et al., 2015). It is essential to take these data into account for the development of future research and interventions.

Despite the availability of effective therapeutic interventions for the treatment of depressive symptomatology, not all patients can be treated with the resources available. It has been found that 50% of patients are not being treated (Egede and Ellis, 2010), and pharmacological treatment remains the treatment of choice, due to the high cost of face-to-face delivery of these treatments. As a result, healthcare professionals are demanding alternatives for their patients. Recently, in response to the need to improve this situation, non-face-to-face models of alternative psychological interventions have been proposed for implementation in medical care using new information and communication technologies (ICTs), known as eHealth. However, although this is a rapidly advancing field, the scientific evidence is not yet abundant.

Effective online interventions for depression have been designed for the general population (Andrews et al., 2010; Montero-Marín et al., 2016; Karyotaki et al., 2021). However, although eHealth programs to address depressive symptoms in the population with diabetes appear to show improvement in depressive symptomatology and diabetes related distress (Franco et al., 2018), they are scarce and methodologically diverse, offering no data on which aspects are the most effective. Accordingly, this review aims to examine the information published to date on the efficacy of psychological interventions delivered through eHealth to improve depressive symptoms in patients with Type 1 diabetes (T1DM) or Type 2 Diabetes (T2DM) and to analyze the characteristics of each, in order to contribute empirical evidence useful to professionals in their decision-making when developing, designing, or selecting future ICT-based interventions for depression in people with diabetes.

The aim of this study was to conduct a systematic review of the effectiveness of eHealth programs designed to reduce depressive symptomatology in people with type 1 diabetes mellitus (T1DM) and type 2 diabetes mellitus (T2DM) compared to control groups (treatment as usual [TAU] or other modalities). To do this, the main outcome evaluated was the change in depressive symptoms assessed by validated psychometric instruments that evaluate depressive symptomatology after the treatment and in the follow up. Secondary objectives were to analyze the effectiveness of the treatments on glycemic control through glycosylated hemoglobin (HbA1c) or other measures of diabetes monitoring.

This systematic review was carried out according to the recommendations of the PRISMA statement (Urrútia and Bonfilll, 2013), and the protocol followed to develop this systematic review is available in PROSPERO (Reg: CRD42020180405).

The aim of this study was to conduct a systematic review of the current evidence on eHealth programs available for the treatment of depression in people with diabetes, and to discuss the procedures and findings extracted, since this is a current and rapidly growing topic in the field of mental health and chronic diseases. Our main findings indicate that, although psychological intervention programs are effective in reducing depressive symptomatology in patients with diabetes, the evidence reported thus far on their delivery through eHealth formats is scarce despite the high incidence of depression in this population (Anderson et al., 2001) and the difficulty in accessing face-to-face interventions (Lehtinen et al., 1997). We also found a great deal of heterogeneity in the methodology used by the studies analyzed: type of diabetes, severity of depressive symptomatology, recruitment and intervention context, treatment content, type of eHealth delivery, comparison group, and follow-up, which makes it difficult to generalize the results and draw conclusions. However, this finding is consistent with that reported by other reviews (Petrak and Herpertz, 2009; Markowitz et al., 2011) on the implementation of these treatments in the usual format. Nevertheless, their methodological quality was very good, a basic result that is very favorable for scientific quality.

An important consideration is that three of the reviewed studies (Van Bastelaar et al., 2011; Nobis et al., 2015; Newby et al., 2017) included patients with T1DM and T2DM. This is relevant because although both conditions belong to the same endocrine disease, they are categorized by the American Diabetes Association (Americam Diabetes Association, 2021), as different etiopathogenetic categories with distinct characteristics and treatments. In the case of T2DM, the main objectives of medical treatment focus on lifestyle modification and administration of oral antidiabetic drugs (Knowler et al., 2009; Americam Diabetes Association, 2021). In contrast, T1DM involves a complicated treatment regimen that additionally requires daily self-monitoring of blood glucose, insulin administration, carbohydrate counting, management of hyperglycemia and hypoglycemia, etc. (Americam Diabetes Association, 2021). These differences between the two types of diabetes have important implications for the individual (differentiated sources of distress), as well as for the design and content of the treatment programs implemented. We therefore recommend that this aspect be considered in future research when designing intervention programs, as it may influence their effectiveness.

There was higher percentage of women than men was observed, which again demonstrates the higher prevalence rate of women with depressive symptomatology (Carreira et al., 2010; Snoek et al., 2015). It was also noted that the different studies report high percentages of the population with a high educational level (university studies), but no data is analyzed or reported regarding this variable, which could influence the effectiveness of and adherence to web-based treatment programs for depression. Therefore, it would be of interest for future research to take this variable into account and report efficacy data comparing groups with different educational levels, in order to advance our understanding of this type of intervention.

Regarding the analysis of the results that were the main focus of this review, we found that treatment programs for depression in people with diabetes implemented using eHealth technology appear to be effective in reducing depressive symptomatology. Nevertheless, it is not possible to draw conclusions regarding which format of eHealth technology is most effective in treating depression for the following reasons. The first is the paucity of studies (only seven included studies) and variability of the formats used (web, mobile phone and video call). Second, no studies have been found comparing different eHealth formats implementing the same treatment program. Only the study by Egede et al. (2018) compared an eHealth format (video call) with face-to-face interventions, finding no significant differences between the two formats. Third, and although it was not the subject under study in this review, none of the studies reported economic assessments that indicate whether these programs are cost-effective, possibly because they were not performed in routine clinical practice. This would be an important aspect to include in future research.

As a positive feature, we found that the treatment programs had in common the use of cognitive-behavioral tools, which constitute the psychological treatment for depression that has been shown to be the most effective in the scientific literature (Markowitz et al., 2011; Petrak et al., 2015) and clinical guidelines (National Institute for Clinical Excellence, 2009), with good results. None of the studies included pharmacological therapy, and the psychoeducational content varied among the different studies.

Concerning the trial setting and recruitment of participants, we found great diversity in the studies analyzed. Recruitment ranged from the use of advertisements in social networks to healthcare settings based on patient lists or only brochures. However, no studies were identified that recruited the sample from primary care (PC), carrying out the intervention in the manner most similar to how such interventions would be carried out in the health care setting. In PC, the prevalence of depression is very high at around 29% (Roca et al., 2009), has a high comorbidity with chronic diseases such as diabetes (Anderson et al., 2001), and is associated with poorer glycemic control (Egede and Ellis, 2010). Moreover, pharmacotherapy remains the treatment of choice in PC for this population, despite scientific evidence that psychotherapy achieves superior long-term results and lower relapse rates (Cuijpers et al., 2019). However, due to this high prevalence, the economic resources required to meet the psychological treatment needs of this population in PC are not feasible (Bower and Gilbody, 2005). For this reason, innovative cost-effective alternatives using ICTs for the treatment of depression in PC (Whiteside et al., 2014; Castro et al., 2015; Montero-Marín et al., 2016; Rodriguez-Pulido et al., 2020) that minimally involve mental health services are being proposed. Nonetheless, it is not possible to draw conclusions from this review with respect to the intervention settings because, although ICTs were used for the treatment of depression in people with diabetes, none of the studies were carried out directly in the PC environment. These data are important factors to consider for future interventions in order to study the possible barriers to implementation of these treatments in PC health systems.

Stepped care models have also been proposed in PC (Bower and Gilbody, 2005), whereby a large proportion of patients are treated first with low-intensity interventions, with significant clinical benefits (García-Herrera et al., 2011). These interventions involve a simpler and easier approach than formal psychotherapies. The contact with patients is shorter, and methods such as the Internet or mobile telephony can be used. In the case of depression, low-intensity interventions are offered to those patients who present mild or moderate depressive symptomatology. In addition, interventions that require less interaction time with the therapist than face-to-face psychotherapy (guided self-help approach) or even no interaction at all (unguided self-help approach) seem to provide very positive results at low cost (Spek et al., 2007; Tate et al., 2009). In this review, we found that only two of the studies analyzed focused on the treatment of mild-moderate depression (Piette et al., 2011; Clarke et al., 2019), two focused on the treatment of moderate-severe symptoms (Van Bastelaar et al., 2011; Naik et al., 2019), one on severe depressive symptoms (Nobis et al., 2015), and three were directed at treatment for MDD (Van Bastelaar et al., 2012; Newby et al., 2017; Egede et al., 2018). However, the different intervention studies examined by this review report good results following the intervention, regardless of the severity of depressive symptomatology. Accordingly, the findings appear to indicate that treatment programs for depression in people with diabetes implemented through eHealth formats are effective in improving depressive symptomatology regardless of severity. These results are very promising because many more patients could benefit. Nevertheless, further research is needed in this regard. In addition, it will be important to include diagnostic interviews based on DSM-5 criteria and not only psychometric instruments, since clinical guidelines recommend that the assessment of depression should not be based only on a mere symptom count (National Institute for Clinical Excellence, 2009).

These results therefore suggest that this technology is effective for the treatment of depression in people with diabetes and has the benefit of providing greater reach and care to a broader patient population. These are very important findings given the scarcity of mental health resources (Bower and Gilbody, 2005). However, scientific evidence indicates that depression in people with diabetes not only has an adverse effect on the person's well-being but also on the clinical outcomes of the disease (Petrak et al., 2015), so treatment should be geared toward improving both psychological and medical outcomes (Petrak and Herpertz, 2009). Nonetheless, the results of this review indicate that these treatments are not effective for improving control of diabetes. These results are similar to those reported by other reviews (Van der Feltz-Cornelis et al., 2010; Markowitz et al., 2011; Baumeister et al., 2012; Petrak et al., 2015), so it may be necessary to review the treatments used in order to provide comprehensive patient management. Considering these results, we asked ourselves the following question: what do these treatments bring to the diabetes setting and how do they differ from those developed for people with depression without diabetes? This answer is key for future research to advance existing knowledge. The scientific literature reports that depression appears to exert its effect on glycemic control in an indirect manner, through poor adherence and self-care behaviors in diabetes (Snoek et al., 2015). However, none of the studies included interventions addressed this aspect in their programs. Therefore, it may be essential for the treatment of depression in people with diabetes to include tools aimed at improving adherence and diabetes self-management together with cognitive-behavioral strategies for the reduction of depressive symptomatology.

eHealth interventions have great potential to impact public health. The rising use of the Internet and mobile devices across the world has made these interventions increasingly common. However, the scientific evidence in this field is very limited and recent. In order to draw conclusions, further studies that integrate these treatments into clinical practice are needed, as well as economic analyses of this type of intervention versus the traditional face-to-face model.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

EV-M and MC performed the independently searching process and organized the databases. MA-O intervened in order to resolve the discrepancies found after the search in the databases. EV-M wrote the first draft of the manuscript and was reviewed by MA-O and MC. All authors contributed to conception, design of the study, revised the manuscript, read, and approved the submitted version.

The translation of this article was carried out with funding from the Puente Project (B.4) of the University of Malaga, entitled Efficacy and Cost-Effectiveness of a Web Application to treat Depressive symptoms in adults with type 1 Diabetes: A controlled clinical trial (20-02-2021 /28-02-2022).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.