Factors associated with diabetes-related distress among Asian patients with poorly controlled type-2 diabetes mellitus: a cross-sectional study in primary care

Type-2 diabetes mellitus (T2DM) is a challenging disease that requires commitment and adherence to a complex set of self-management tasks in order to get the best possible outcome. As a result, persons with diabetes (PWD) often feel frustrated, burnt out or overwhelmed with worries related to the current management or future implications of their disease [1]. Diabetes-related distress (DRD) refers to the negative emotional state arising from the burden of living with the disease [1]. The likelihood of DRD is higher among those with poor glycaemic control compared to those who are well-controlled [2]. Its relationship with glycaemic control is time-concordant [3]. Its presence at baseline, in one study by Aikens (2012), has been linked to future glycaemic control [4]. DRD has been associated with poorer medication adherence and lower frequencies of self-care behaviours. Both medication adherence and self-care behaviours are well-established determinants of glycaemic control, which are in turn associated with future complications and lower quality of life.

The prevalence of DRD was reported as 36%, in a meta-analysis of 55 studies, with gender and comorbid depressive symptoms as significant factors affecting prevalence [5]. However, a high level of heterogeneity was noted in these studies, and more than half of them were from the United States of America, limiting its global generalizability Distress levels and content are influenced by psychosocial, socio-economic, cultural and demographic variables, which vary across different populations. Even within the same country, prevalence of DRD differs between study sites [6] and level of care, with primary care having a lower prevalence compared to secondary care [7, 8].

Apart from gender, depressive symptoms and glycaemic control, several other factors have been identified to be associated with DRD, including younger age [6, 9‐11], certain ethnicities [6, 10], living alone, lower education [9], greater number of diabetes-related complications [9, 12], frequency and severity of hypoglycaemic episodes, diabetes-related family arguments and diabetes support gap [9]. Conversely, lower levels of DRD were found with the older adults [13], married [10, 14], employed [10], higher family and social support and better patient-physician relationship [13]. Conflicting evidence was found for the association of DRD with duration of T2DM. Several studies showed its association with DRD [8, 15, 16], while one large, multinational study showed the reverse [9].

Understanding the associated factors is key to design appropriate interventions to mitigate DRD in primary care where most PWDs are managed. A systematic review by Sturt (2015) alluded that psycho-education, when conducted by general practitioners or practice nurses for six sessions or more and for three months or more, was effective in reducing DRD [17]. A chronic disease management programme in Sydney, Australia, reported the effectiveness of a diabetes health-coaching programme in reducing DRD for those with the highest level of baseline distress. It was also effective in improving the glycaemic control in those with a baseline HbA1c above the recommended level of 7% [18]. This would suggest that intervention measures would likely benefit those with a high level of distress and suboptimal glycaemic control.

Singapore has a highly urbanised multi-ethnic Asian population with the prevalence of T2DM doubling from 7.3% in 1992 to 13.7% in 2018. DRD had been identified in PWDs of a tertiary hospital but the majority of PWDs in Singapore are managed in primary care. Hence, this study aimed to assess the prevalence of DRD and its associated factors in PWDs with poor glycaemic control of HbA1c ≥ 8% in primary care. The factors included age, duration of T2DM, hypoglycaemic frequency, financial difficulties, social support and encounters with healthcare providers.

A total of 370 patients were recruited and consented to participate in the study. 13 participants were excluded from analysis because they did not fulfil the inclusion criteria. One person dropped out mid-way through the questionnaire because of time constraint. Eventually, completed data of 356 subjects were included in the final analysis.

Patient characteristics are presented in Table 1, while disease and management data are presented in Table 2. The mean age was 58.6 years, with 50.3% female. The frequency of Chinese, Malay, Indian and Others were 62.1%, 22.5%, 13.5% and 2.0% respectively. Mean HbA1c was 9.5%, while mean BMI was 28.1 kg/m². The mean duration of diabetes was 10.8 years, while the mean PAID score was 15.4. The prevalence of DRD was 17.4%.

Subjects with DRD were younger than those without DRD (55.2 years old vs 59.4 years old, p = 0.02) (Table 1). No difference in duration of illness was noted between the two groups (Table 2).

Those who experienced symptoms of hypoglycaemia were more likely to have DRD compared to those without such symptoms (OR (95% CI) = 2.18 (1.22–3.90), p = 0.008). Non-smokers and ex-smokers were more likely to have DRD compared to current smokers (OR (95% CI) = 5.02 (1.18–21.36), p = 0.029 and OR (95% CI) = 8.00 (1.56–40.91), p = 0.013). Subjects with kidney disease were also more likely to have DRD (OR (95% CI) = 2.13 (1.10–4.10), p = 0.024). No difference was associated with other co-morbidities such as hypertension, dyslipidaemia, ischemic heart disease, cerebrovascular disease, diabetic eye disease, peripheral vascular disease, neuropathy, diabetic foot disease and anaemia. Financial assistance (Table 1), or previous diabetes-related nurse encounters or previous dietician encounters were not associated with DRD (Table 2).

Those screened positive for anxiety and depression using GAD-7 and PHQ-9 with scores of 10 or higher were more likely to have DRD (OR (95% CI) = 16.77 (6.62–42.46) and OR (95% CI) = 23.14 (7.36–72.72), both p < 0.001) (Table 3).

Subjects with DRD had a lower mean EQ-VAS score compared to those without DRD (66.6 vs 75.4, p < 0.001), and a lower EQ-5D index score compared to those without DRD (0.8 vs 0,9, p < 0.001) (Table 3).

Subjects who felt anxious about their weight, felt discriminated against, found difficulties paying for medications, worried about their financial future and felt that their medications interfered with their normal life were more likely to have DRD (Table 4).

After adjusting for confounding factors, younger age (AOR (95% CI) = 0.93 (0.89–0.97), p = 0.004), ex-smokers (AOR (95% CI) = 22.30 (2.43–204.71), p = 0.006), history of kidney disease (AOR (95% CI) = 3.41 (1.39–8.35), p = 0.007) and those who felt that medications interfered with their normal lives (AOR (95% CI) = 2.89 (1.38–6.08), p = 0.005) remained significantly associated with DRD. A positive PHQ-9 screen (AOR (95% CI) = 4.98 (1.19–20.86), p = 0.028) as well as a lower EQ5D index score (AOR (95% CI) = 0.11 (0.01–0.97), p = 0.047) continued to be significantly associated with DRD (Table 5).

The DRD prevalence of 17.4% is lower than the 32% previously reported when the PAID questionnaire was administered in a tertiary endocrinology clinic [20]. It is compatible with two other studies where the prevalence in primary care was lower compared to secondary care [7, 8]. PWDs in secondary and tertiary centres tend to be more complex and with more complications compared to those who are treated in the primary care setting. No difference was noted in the number of comorbidities or number of medications between those with and those without DRD in our study. Although the prevalence of DRD seems to be low in primary care setting, the absolute numbers are still significant. Effective interventions to address DRD can potentially result in marked reduction in T2DM-related disease burden.

Subjects with DRD were younger compared to subjects without DRD. This result is consistent with local [32] as well as overseas [6, 9‐11] studies. Younger working PWDs face financial stressors, challenges at work and family responsibilities which may increase their difficulties of living with T2DM. They may perceive their illness as a threat or a loss at a time in their lives when they expect themselves to be able-bodied to perform their role as providers or caregivers for their families, hence reacting more negatively to the stressor. Clinicians need to recognise and address the challenges unique to these younger PWDs. A good doctor-patient relationship, based on empathic communication and person-centred approaches, such as motivational interviewing, has been found to reduce diabetes distress, and also improve self-care [17, 33, 34].

Smoking is well known to increase cardiovascular morbidity and mortality among PWDs, and most, if not all, guidelines would recommend that PWDs do not smoke. Our study found that ex-smokers, defined as those who have stopped smoking for at least one year, were at risk for DRD. While most trials take six months to one year as a proxy for life-time smoking cessation, a meta-analysis estimates that annual relapse rates after the first year is approximately 10% [35]. Quitting smoking and remaining smoke-free is challenging, even for those without diabetes. Many who smoke do so as a response to stress. PWDs with DRD who are ex-smokers may benefit from frequent check-ins on their smoking remission status so as to keep them in the maintenance phase of smoking cessation, and to identify and reduce stressors that may trigger a relapse.

The result revealed that PWDs who experienced hypoglycaemia symptoms in the past one year were more likely to have DRD. The perception that diabetes medications interfere with normal life may have been a confounder in this relationship. Hypoglycaemia has been identified as a factor associated with DRD in other multi-centre studies [9, 12]. The DAWN-2 study examined the frequency and severity of symptomatic hypoglycaemia on DRD, and found an association not only with DRD, but also with well-being (WHO-5 Wellbeing Index) and quality of life (WHOQOL-BREF). Hypoglycaemia assessment and management is an integral part of good diabetes care [36]. Managing and educating PWDs about hypoglycaemia, with a focus on addressing issues and concerns related to the PWD’s diabetes medications, not only mitigate this life-threatening risk but may also alleviate their DRD.

DRD, depression and anxiety are often viewed as overlapping concepts, or part of a continuum. However, depression is a separate entity from DRD. Any PWD may have depression or DRD or both. Co et al. reported that DRD was a mediator between poor glycaemic control and health related quality of life [32]. It has also been suggested that the higher rates of depression among PWDs may be mediated by DRD [37]. Thus, psychotherapy and counselling strategies to alleviate DRD in PWD may concurrently prevent and manage their depression and improve their quality of life.

The results identified anxiety around weight and perception that diabetic medications interfered with normal life as significant factors associated with DRD. Younger PWDs could be more image conscious, with higher propensity to be more troubled by weight (and appearance) compared to older PWDs. In a study on self-perception of weight of Korean women aged 20 to 79 years old, it was reported that older women tended to underestimate their weight relative to actual body mass index, compared to younger women [36]. This misperception could reduce incidence of DRD in older PWD but would potentially dampen their activation to lose weight.

The results did not show a significant association between number or type of medications and DRD. While the medications were not free to the PWDs, they were dispensed at highly subsidised prices at polyclinics due to prudent national healthcare finance policies [38], which ensure that local residents will not be deprived of healthcare services due to cost.

DRD is linked to lower health-related QOL [32], and social support has been shown to moderate this relationship [39]. The results of this study support the association between DRD and lower QOL. However, after adjusting for confounders, there was no significant association found between perceived social support and DRD in this study population. In 2016, Singapore’s Ministry of Health declared War on Diabetes, a whole-nation effort to tackle diabetes through various channels, such as public awareness and health promotion campaigns, patient education, and healthcare financing strategies. At the institution level, a multidisciplinary team involving care coordinators, nurses, doctors, pharmacists and medical social workers, provide holistic care through various programmes, including health coaching and telecare services. In the community, organisations such as Diabetes Singapore, provide peer support, counselling and screening for diabetes and diabetes-related complications. While the healthcare structures and social service structures are largely in place, the interface between the two continues to be a focus in the delivery of population health in Singapore.

One in six (17.4%) PWDs with poor glycaemic control had DRD in this study. Younger age, ex-smoker status, kidney disease, depressive symptoms and the perception that medications interfere with normal life are associated with their DRD. The results highlight the need for primary care physicians to proactively screen PWDs for DRD, especially when their glycaemic control is suboptimal, and to direct them for evidence-based therapy. Beyond glycaemic control, the mental and emotional health of PWDs (and their families) must not be overlooked, as it will impact their health outcomes and quality of life.