Depression is a prevalent mental health condition that affects more than 280 million individuals worldwide (World Health Organization, 2021). According to the International Diabetes Federation (2021), around 537 million adults aged 20–79 years are living with diabetes worldwide, and the proportion of individuals diagnosed with type 2 diabetes is increasing in most countries. Golden et al (2017) found that, globally, 43 million people who have diabetes also have symptoms of depression. Similarly, Public Health England (2018) noted that depression is more common among people living with type 2 diabetes, compared with those who are not.
In addition, depression may have negative implications for those with type 2 diabetes in relation to diabetes management. Schmitz et al (2014) suggested that adults with diabetes and depression have been shown to have more problems with self-management of their diabetes—in terms of following a healthy diet, exercising, adherence to medication, blood glucose monitoring and not smoking—which can increase the risk of macrovascular and microvascular complications. Ma et al (2018) argued that people with depression are prone to have a poor quality of life, poor treatment compliance and suboptimal glycaemic control. Furthermore, Lunghi et al (2016) identified that establishing the risk factors for developing depression in this patient group could help health professionals to identify those diabetic patients at high risk of developing depression and thus prevent or treat depression in people with type 2 diabetes.
This integrative literature review explored the prevalence of depression in individuals with type 2 diabetes. An integrative literature review is a systematic way of carrying out research that synthesises the findings from different types of previously conducted studies, to gain a deeper theoretical understanding of the research topic (Santos and Eslabao, 2019). An integrative review is also useful as it helps to summarise the literature that is available on a particular topic (Aveyard, 2014).
Methods
This review used the framework of population, intervention/issue and outcome (PIO) to generate the research question (Table 1). Whiffin (2020) suggested that a well-designed research question can be framed using a PIO model because it allows key concepts in a question to be identified and a search strategy developed that will generate as many search terms as possible. In addition, key words were searched in combination rather than alone to develop a focused search strategy using Boolean operators such as ‘AND’ to narrow down the search by combing different concepts as well as using ‘OR’ to broaden the search by including alternative words (Roberts et al, 2014) as shown in Table 1.
Table 1. PIO framework and key words
| PIO: | Key words | |
|---|---|---|
| Population | Adults (aged over 18 years) | Young adults, middle-aged adults, older adults |
| Intervention/issue | Diagnosed with type 2 diabetes | Type 2 diabetes mellitus, T2DM, adult-onset diabetes, non-insulin dependent diabetes, diagnosis, diagnosed |
| Outcome | Risk of developing depression | Depressive symptoms, depressive disorders, depressive status, likelihood, possibility, prevalence of depression, associations |
Inclusion and exclusion criteria were used to help select relevant articles for the integrative literature review. This is reinforced by Hiebl (2021) who argued that these criteria should help to reveal the precise explanations of why a specific piece of research was included or not included within the review. Table 2 shows the inclusion and exclusion criteria.
Table 2. Inclusion and exclusion criteria for search
| Inclusion criteria | Exclusion criteria |
|---|---|
| Individuals aged over 18 years | Individuals under the age of 18 |
| Type 2 diabetes | Type 1 diabetes |
| Primary research | Secondary research |
| Articles published from 2011 to 2021 | Articles published more than 10 years ago |
| Articles published in the English language | Articles not published in the English language |
| Depression | Other mental health conditions (such as anxiety, schizophrenia, and post-traumatic distress disorder) |
| Full text articles | Reviews, commentaries, articles only displaying the abstract and discussion papers, opinions and editorials |
| Research articles carried out in Europe | Research articles not carried out within Europe |
Electronic databases that were used to retrieve the articles were CINAHL (Cumulative Index of Nursing and Allied Health Literature) (871 hits), Medline (168 hits) and PsycInfo (34 hits), applying the inclusion and exclusion criteria to all three databases. The most relevant articles that met the criteria for this review were then selected. In total, there were seven quantitative studies, three of which were retrieved from CINAHL and four from Medline, with no articles retrieved from PsycInfo.
A data extraction table was used to help systematically extract relevant information (Collaboration for Environmental Evidence, 2020) (Table 3), and a data extraction table also helped to draw out themes. Sandelowski et al (2012) discussed how aggregating findings into themes is used to confirm findings and enable the researcher to answer the research question.
Table 3. Data extraction table
| Author/Year | Study aim(s) | Study design method | Data collection method | Research results/findings |
|---|---|---|---|---|
| Lloyd et al (2018) | To evaluate the incidence as well as management of depressive disorders in individuals with type 2 diabetes across 14 countries |
|
|
|
| Alonso-Moran et al (2014) | To evaluate the occurrence of depression in adults with a diagnosis of type 2 diabetes, and to analyse the hypothesis of whether depression is linked with poor glycaemic management and increased healthcare costs |
|
Database of the population stratification programme of the Basque Health Service known as Osakidetza. All individuals within the Basque Country aged 35 years and above were included in the study (sample size 126 894 participants) |
|
| Salinero-Fort et al (2018) | To evaluate the occurrence of depression in those with a diagnosis of type 2 diabetes, as well as to recognise sociodemographic, psychological and clinical determinants |
|
|
|
| Lopez-de-Andres et al (2015) | Describe the trends in the occurrence of depression among patients in hospital diagnosed with type 2 diabetes within Spain from 2001 to 2011 |
|
Sampling method:
|
|
| Foran et al (2015) | To establish whether the incidence of depression is more prominent in patients aged over 50 with type 2 diabetes within the West of Ireland and if depression is an independent predictor of diabetes management |
|
|
|
| Jacob and Kostev (2016) | To investigate depression in German patients with type 2 diabetes (with or without diabetes complications). |
|
|
|
| Bo et al (2020) | To demonstrate the incidence of depressive symptoms, diabetes distress and perceived stress among adults with early-onset type 2 diabetes, as well as to assess their link with sociodemographic and clinical traits |
|
|
|
Results
Based on the seven articles selected, the themes gender, age, and socio-economic circumstances were chosen.
Gender
Foran et al (2015) carried out a quantitative study using a cross-sectional design to find out the prevalence of depression in individuals with type 2 diabetes within Irish primary care centres. They found that gender was not a statistically significant predictor of having depression (P>0.05) among this cohort. A critique of this study could be that the study sample only included 283 patients (34% response rate). Therefore, the results may be at risk of non-response bias, as Cheung et al (2017) have argued that non-response bias may affect the validity of prevalence estimates within a population. Hence, the results of this study may not be reliable or valid to draw concrete conclusions about the relationship between type 2 diabetes, gender and depression. The use of a cross-sectional design may also be a weakness of the study, as it hinders any conclusions regarding causal relationships (Bo et al, 2020). However, Setia (2016a) has argued that cross-sectional studies are useful for designing cohort studies and can help to provide information about the prevalence of outcomes or exposures.
Lloyd et al (2018) carried out a collaborative study across 14 countries using quantitative data for 2013–2015 to find out the occurrence of depressive disorders in adults with type 2 diabetes. Unlike Foran et al (2015), Lloyd et al discovered that individuals with type 2 diabetes, current major depressive disorder and moderate or severe depressive symptomatology were more likely to be linked to being female (P<0.0001). The P value, used to determine statistical significance, reflects the reliability and choice of analytic measures, trustworthiness of the protocol within the study, quality of design and, most importantly, the sample size (Kraemer et al, 2020). Lloyd et al (2018) further uncovered that more women participated in the study compared with men (93.7 vs 90.8%; P=0.003), which could lead to participation bias. Participation bias may cause the sample to not be representative of the population being studied, which can affect the overall findings and conclusions drawn (Keeble et al, 2013). Deischinger et al (2020) argued that men may under-report their depressive symptoms and are less likely to seek help, thus depression in men may not be as recognised by health professionals compared with that in women.
Age
Alonso-Moran et al (2014) in their cross-sectional study using quantitative data found that for each individual aged under 65 years, 1.5 individuals aged 65 or over with type 2 diabetes suffered with depression (OR-1.5; CI 95%: 1.4-1.6). However, all participants who took part in the study were aged 35 years and over, which may mean that the results cannot be generalised to the entire type 2 diabetes population, potentially affecting the external validity of the results, according to Bhandari (2021).
Lopez-de-Andres et al (2015) used quantitative data for their retrospective observational study on the trends in the occurrence of depression among patients in hospital with a diagnosis of type 2 diabetes in Spain during 2001-2011. In contrast with Alonso-Moran et al (2014), Lopez-de-Andres et al found that depression was most prevalent among young adults and found that the lower expectation of experiencing chronic disease at an early age negatively affected their ability to deal with health problems and life stressors compared with older adults. A strength recognised by the authors was the large sample size. Roessner (2014) pointed out that a large sample size allows for an easier assessment of representativeness of a sample and the generalisability of results achieved—how confidently the results from the sample of the study can be extended to the rest of the research population (Murad et al, 2018).
However, a study by Jacob and Kostev (2016), using a longitudinal design over a 10-year period (2004-2013), found that age had no effect on the incidence of depression in younger or older people. Caruana et al (2015) have emphasised that longitudinal studies are beneficial for evaluating relationships between risk factors and the development of disease, along with the results of treatments over different periods of time. Jacob and Kostev (2016) revealed that the relationship between depression and age is not well understood and that numerous studies have produced contradictory results, which suggests the results may be unreliable. This implies it may be difficult to estimate the specific age range within which depression is more likely to develop in those with type 2 diabetes.
Socio-economic circumstances
A cross-sectional study using quantitative data by Bo et al (2020) found that those with a medium to high level of education had lower levels of depressive symptoms compared with those with a low education level, and that levels of depression were higher among unemployed people by 5.69 compared with people in employment (95% CI 3.32 to 8.06). Similarly, Salinero-Fort et al (2018) conducted a prospective cohort study using quantitative data and uncovered that type 2 diabetes patients with depression were more likely to have a lower educational level compared with those without depression (P<0.001). Salinero-Fort et al further found that those with type 2 diabetes and depression were also less likely to be categorised as employed compared with those without depression (P<0.001). Unemployment was linked with depression, which suggests that working may have a protective role due to social support from colleagues (Salinero-Fort et al, 2018). Therefore, socio-economic circumstances such as a lower level of education and unemployment may increase the risk of depression in those with type 2 diabetes.
Within the study by Bo et al (2020) the participants were invited to be a part of the Danish Centre for Strategic Research in Type 2 Diabetes (DD2) cohort study once a GP or hospital physician diagnosed them with type 2 diabetes as part of routine clinical practice. This suggests the study promoted objectivity: the researcher did not choose which patients were invited to the study, so it was based purely on the diagnosis of type 2 diabetes. However, achieving absolute objectivity in research might not be possible. Karimova (2014) argued that objectivity is an unreachable goal that can never truly be achieved as all knowledge is subjective or partial in some way—everything is observed and reported by humans, therefore human error is inevitable. Consequently, it might not be possible to have absolute objectivity, but the participants involved in the study had not been selected by the researcher, which may have helped promote objectivity within this study.
A disadvantage for the design of the study by Salinero-Fort et al (2018) was the short time between telephone interviews (12 months), meaning it could be hard to compare cumulative incident rates with other studies. This point is supported by Setia (2016b), who stated that these types of studies help to estimate cumulative incidence and incident rates, and that a benefit of a cohort study is its longitudinal nature. However, Salinero-Fort et al argued that a prospective design for their study helped to reduce the risk of selection bias.
Discussion
This integrative literature review has shown that individuals with type 2 diabetes are at a greater risk of developing depression, and factors such as age, gender and socio-economics also play a role in predicting whether a person with type 2 diabetes will develop depression. This demonstrates the importance of identifying depression in these higher risk groups. Therefore, the findings indicate that it would be beneficial to implement screening tools for depression within GP surgeries to screen for depression at the time of diagnosis of type 2 diabetes, and ensure that patients have annual follow-up appointments. This is because screening for depression in those with type 2 diabetes may help to ensure that risk factors are not overlooked. Habtewold et al (2016) found that a major barrier in diagnosing depression in those with type 2 diabetes was lack of screening tools. Depression is linked with a higher risk of mortality, absenteeism, poor disease management, and poor health outcomes (Owens-Gary et al, 2019). Hence, it seems logical that early recognition through screening for depression in those with type 2 diabetes could have a positive impact by ensuring that an appropriate treatment/management plan is put in place.
Practice nurses have a crucial role in supporting, screening, and maintaining those with diabetes, as well as having an awareness of how mental health can affect individuals with diabetes (Royal College of Nursing, 2020). Therefore, practice nurses are in an ideal position to implement regular depression screening in patients diagnosed with diabetes within GP surgeries. Implementing this change also helps nurses to adhere to the Nursing and Midwifery Council (2018)Code, which states that registered nurses must encourage the wellbeing of individuals, prevent ill health, as well as keep up with the changing healthcare needs of individuals through all stages of their life.
Alonso-Moran et al (2014) highlighted the need for more accurate identification of depression among the diabetic population and periodical screening and monitoring for depression among those with type 2 diabetes. Furthermore, the International Diabetes Federation (2017) clinical practice recommendations for primary care, which took into account a wide range of guidelines from other sources, state that some of these guidelines recommend using validated screening tools for depression such as the Patient Health Questionnaire-2 (PHQ-2) (Kroenke et al, 2003): if the probability of a depressive disorder is 75% or above, then the primary care physician should consider referral to a specialist for evaluation. Thus, the evidence suggests that validated screening tools are beneficial for patients with type 2 diabetes by recognising early signs of depression to ensure patients are referred to the appropriate specialists to ensure the condition is managed or treated early.
There could be various facilitators and barriers to implementing the PHQ-2 screening tool in order to improve the recognition of depression in those with diabetes. One these barriers could be staff attitudes (Health Foundation, 2015), for example, they may be resistant to change, not agree that the change is required, or not agree that a particular change is the best way to improve care. This may be because staff members may not understand why a change is needed. A facilitator to overcome this could be education and training, for example, Maxwell et al (2013) suggested that, if screening for mental health conditions is to be encouraged, then primary care nurses need training to improve their confidence in dealing with mental health conditions.
Conclusion
This integrative literature review has revealed that those with type 2 diabetes are at a greater risk of developing depression, and factors such as age, gender, and socio-economic status also play a role in predicting whether a person with diabetes will develop depression. One suggestion emerging from the findings of the review is that it would be beneficial for practice nurses to use screening tools such as PHQ-2 to assess for depression within GP surgeries at the time of diagnosis of type 2 diabetes and ensure patients have annual follow-up appointments.
A strength of this suggestion is that there is supporting evidence from Alonso-Moran et al (2014) for a need for more accurate identification of depression in the diabetic population, as well as periodical screening and monitoring for depression among those with type 2 diabetes. However, a potential weakness of the suggestion may be the barriers to implementing it into practice, such as staff attitudes, which may need to be overcome, possibly through training to increase confidence in mental health screening in order to achieve this change in practice.
Even though age was identified as a potential risk factor of depression in those with type 2 diabetes, there was still contradictory evidence in relation to what influence age has on depression in individuals with type 2 diabetes. More research is needed to fully understand the association between different age cohorts in relation to the development of depression in people with diabetes.
KEY POINTS
- People with type 2 diabetes are at a greater risk of developing depression
- Gender, age, and socioeconomic factors may increase the risk of adults with type 2 diabetes developing depression
- A major barrier in diagnosing depression in those with type 2 diabetes is a lack of screening tools
- Screening tool such as the Patient Health Questionnaire-2 (PHQ-2) may be used to identify depression within GP surgeries
CPD reflective questions
- Is depression considered one of the risk prognoses that develops in people with type 2 diabetes? Why?
- Are there any screening tools in your clinical area used to identify depression in people with type 2 diabetes?
- Where do you think people with type 2 diabetes could have this screening, and who would be best placed to undertake it?