Diabetes distress as mediators of loneliness and health promotion behaviour: a cross-sectional study


  • The study employed face-to-face data collection methods instead of online consultations to mitigate potential biases associated with online surveying.

  • This study indicated attention to diabetes distress among older adults with diabetes.

  • The study relies on self-reported outcomes, which may introduce measurement bias.

  • This was not a multicentre study not representative of a pan-global population.


Diabetes mellitus (DM) is one of the fastest-growing diseases worldwide, projected to affect 693 million adults by 2045.1 The past 50 years have seen a growing ageing population with an increasing prevalence of DM. Now, nearly half of all individuals with DM are older adults (aged ≥65 years).2 Older adults with DM face a higher risk of premature death, functional disability and coexisting conditions than older adults without DM.3 DM increases the vulnerability of older adults to several common geriatric syndromes, such as polypharmacy, cognitive impairment, depression, urinary incontinence, injurious falls, persistent pain and frailty.4 Moreover, the complexity of disease management in older adults is compounded by the presence of comorbidities, the higher susceptibility to hypoglycaemic events and the individualised care needs.5 Improving the health promotion behaviours of older adults with DM is an important goal in disease management.

Health promotion behaviour can effectively delay the development of diabetic complications and cause the body, biochemical and other indicators to return to normal levels.6 Health promotion behaviours for people with DM include blood glucose monitoring, use of insulin and medications, use of diabetes technology, physical activity and nutritional changes.7 Health promotion behaviour is vital for older adults with DM. However, psychological factors are proven to influence the behaviour of patients with DM.8

One such psychological factor is loneliness. Loneliness is the feeling of isolation regardless of objective social network size.9 In industrialised countries, around one-third of people are affected by this condition, with 1 person in 12 affected severely, and these proportions are increasing.10 A longitudinal cohort study showed loneliness had a negative impact on health promotion behaviours.11 The result of a phenomenological study showed that loneliness in people with DM is related to diabetes distress.12 Higher loneliness was also associated consistently with worse health outcomes among older adults.13

Another negative psychological characteristic reported in patients with DM is diabetes distress.14 Diabetes distress refers to the worries, concerns, fears and threats that are associated with struggling with a demanding chronic disease such as diabetes over time, including its management, threats of complications, potential loss of functioning and concerns about access to care.15 Many studies have found that diabetes distress is negatively associated with health promotion behaviour in older patients with DM.16 17 In summary, loneliness and diabetes distress are independently negative characteristics that seem to be associated with health promotion behaviour in older patients with DM.

Considering all this evidence, the relationship among these three important factors has not been fully explored. Up to now, previous studies remain narrow in focus dealing only with unidirectional relationship. No previous study has investigated the mediating role of diabetes distress on loneliness to health promotion behaviour in older patients with DM.

The theoretical framework of this study is the Health Belief Model (HBM).18 The HBM interprets an individual’s health-promoting behaviour as the result of perceptions of being exposure to a certain health threat. Loneliness is seen as an exposure risk, diabetes distress is seen as a health threat, and lonely individuals form beliefs about health-promoting behaviours through their perceptions of diabetes distress. If individuals believe that distress increases health risk, they may be more inclined to adopt positive health behaviours. Based on the theoretical framework, we propose the hypothesis: loneliness directly affects health promotion behaviour and indirectly influences health promotion behaviour through diabetes distress as a mediator.

Hence, this study was designed to investigate the mediating association of diabetes distress on loneliness to health promotion behaviour and to clarify the direct/indirect relationships in older patients with DM.



This study aimed to examine the mediating role of diabetes distress on the relationship between loneliness and health promotion behaviours. The hypothesis was diabetes distress plays a mediation role in the relationship between loneliness and health promotion behaviours in older adults with DM.


This study employed a cross-sectional design.


Participants were recruited from three third-class hospitals in Hunan by convenience sampling. The inclusion criteria for recruitment included participants aged above 65 years having been diagnosed with DM. Exclusion criteria included patients with psychiatric problems and blindness. G power V. software was used to calculate the adequate sample size. Based on an α level of 0.05, a power of 0.80 and an effect size of 0.2 for the correlation coefficient, 84 samples were required. Finally, 140 patients finished this study, and the sample size was adequate to test the hypotheses of this study.

Data collection

Participants complete the questionnaire independently. For those with blurred vision or a low knowledge level, investigators read out the content and options of the article without implication and complete it on their behalf. After the questionnaire is completed, the investigator will take it back on the spot and check the completeness of the questionnaire to ensure its completeness and validity.

Demographic and disease characteristics

Age, gender, medical payment method, comorbidities (no/yes), course of disease and use of insulin(no/yes) were collected.

Diabetes distress

We measured the diabetes distress levels in participants using the Chinese version of the Diabetes Distress Scale.19 This scale has 17 items and uses a Likert 6-point scoring method, ranging from 1 (not a problem) to 6 (serious problem). The scale evaluates the results based on the average score of the item. A higher score indicated a higher level of diabetes distress. An average score of ≥3 on the scales suggests more than moderate distress and requires clinical attention. The Cronbach’s α value of this scale was 0.92.


We measured loneliness using the Chinese version of the University of California at Los Angeles (UCLA) Loneliness Scale.20 This scale has 20 items and uses a 4-point Likert scoring method, ranging from 1 (never) to 4 (always). The total scores on this scale vary from 20 (least lonely) to 80 (most lonely). A higher score indicated a higher level of loneliness. A score of ≥34 on the scales represents a patient with more than moderate loneliness. The Cronbach’s α value of this scale was 0.86.

Health promotion behaviour

We used a 20-item Chinese version of the Elderly Health Promotion Scale to measure health promotion behaviour.21 This scale has a subscale that assesses health habits, community participation, health responsibilities, health diet, regular movement and oral healthcare. A higher score indicated a higher level of health promotion behaviour. The Cronbach’s α value of this scale was 0.90.

Data analysis

The statistical analyses were performed using SPSS V.26.0 and PROCESS macro for SPSS V.3.3.22 The independent t-test and analysis of variance test were used to assess the difference of the means of normally distributed variables. A multiple linear regression and Spearman correlation analyses were performed to examine the correlations among variables. PROCESS macro with model 4 for SPSS using 5000 bootstrap samples was conducted to examine the mediating role of diabetes distress on the relationship between loneliness and health promotion behaviours.

Patient and public involvement

No patient involved.


We distributed 150 questionnaires and collected 140 valid questionnaires finally. Table 1 shows the demographic and disease features of participants and their association with health promotion behaviours. Participants were equally male (50.0%) and female (50.0%), with a mean age of 72.57 years. All demographic and disease characteristics except the use of insulin did not significantly associate with health promotion behaviours. Thus, we did not consider these characteristics in the mediating tests.

Table 1

Distributions of demographic and disease characteristics and the association with health promotion behaviours in participants (N=140)

Table 2 shows the distribution of the subscale and total scale of health promotion behaviours, loneliness and diabetes distress. We used Spearman’s correction to test the correlations among the variables. As shown in table 2, diabetes distress and loneliness were both significantly and negatively related to health promotion behaviours. Loneliness and diabetes distress also had a significantly positive relationship.

Table 2

Distribution and correlation among health promotion behaviours, loneliness and diabetes distress (N=140)

Table 3 reports the results of the regression analysis of health promotion behaviours, loneliness and diabetes distress. We performed a multiple linear regression analysis with health promotion behaviours of patients with diabetes as the dependent variable and loneliness and diabetes distress scores as independent variables. The results indicated that both loneliness and diabetes distress entered the regression equation. Loneliness and diabetes distress negatively predicted the health promotion behaviour scores of patients.

Table 3

The multiple linear regression analysis results of health promotion behaviours, loneliness and diabetes distress (N=140)

Figure 1 and table 4 show the mediating role of diabetes distress on the relationship between loneliness and health promotion behaviours. The total effect of loneliness on health promotion behaviour was significantly negative (c path, β=−0.312, SE=0.055, 95% CI (−0.204 to –0.436), p=0.006). Loneliness significantly and negatively correlated with diabetes distress (a path, β=−0.043, SE=0.005, 95% CI (0.035, 0.052), p<0.001), while diabetes distress significantly and negatively correlated with health promotion behaviours (b path, β=−2.724, SE=1.019, 95% CI (−4.739 to –0.710), p=0.008). According to the results of bootstrapping, the indirect effect of loneliness on health promotion behaviour was significant (β=−0.118, SE=0.046, 95% CI (−0.211 to –0.027)). The direct effect of loneliness on health promotion behaviour was significant (c’ path, β=−0.194, SE=0.07, 95% CI (−0.331 o –0.056), p=0.006). This result suggested that diabetes distress partially mediated the association of loneliness with health promotion behaviours.

Table 4

Total, direct and indirect effect of loneliness on health promotion behaviour (N=140)

Figure 1
Figure 1

The mediated model represents the total direct effect (path c), direct effect (path c’) and indirect effect (paths a and b) from loneliness to health promotion behaviours. **p<0.01; ***p<0.001.


To the best of our knowledge, this is the first study to explore the mediating role of diabetes distress in the relationship between loneliness and health promotion behaviours among older patients with DM. The result indicated that diabetes distress partially mediates the association between loneliness and health promotion behaviours.

In this study, the mean age of participants was 72.57±4.60, which is older than 57.9±5.60 and 66.9±5.80 in previous large sample studies.23 24 This result may be explained by the fact that advanced age is one of the risk factors for admission in patients with diabetes, and our study included hospitalised patients. Comparing a previous study that used the same scale but was not limited to older patients with DM, the health promotion behaviours were at the middle level.25 Another important finding is that older adults with DM performed poorly in community participation and health responsibilities. Health responsibilities are relevant to health policy development.26 Community participation is associated with gaining health-related information, reducing physical and mental stress, and maintaining motor function.27 Further studies are needed to investigate the effects of these two factors on older adults with DM.

Loneliness has a negative and significant association with health promotion. This may be because individuals who experience heightened levels of loneliness tend to also experience elevated levels of stress. And stress maladaptation is significantly associated with poor health promotion behaviours.28 Loneliness correlates with heightened sedentary behaviour and decreased levels of physical activity.29 The correlation between loneliness and activity levels could, therefore, play a crucial role in the inverse connection between loneliness and health promotion behaviours.

Diabetes distress also has a negative and significant association with health promotion behaviours. This finding was also reported by a previous study.30 One potential explanation is that diabetes distress, as a negative psychological factor, compromises their ability to self-manage their condition. Narrower self-management abilities are associated with poorer health-promoting behaviours.30 Evidence supports that the level of diabetes distress can significantly affect glucose control in older patients with DM.31 Many patients with DM experience a psychosocial burden and mental health problems related to the disease. Diabetes distress has distinct effects on behaviours and disease control.32 These findings imply that the assessment and intervention of psychological dimensions in diabetes management are essential.

In this study, diabetes distress partially mediates the relationship between loneliness and health promotion behaviours among older patients with DM. Previous studies have explored the mediating relationship between negative psychology and behaviour in diabetes distress.33 Older patients with DM experiencing high levels of diabetes distress exhibit more pronounced negative psychological factors associated with their disease compared with those with significantly lower levels of diabetes distress. These individuals tend to employ ineffective coping strategies in response to feelings of loneliness and engage in suboptimal health promotion behaviours. This study explores the mediating relationship between diabetes distress and loneliness and health promotion behaviours. This suggests that loneliness not only directly affects health promotion behaviours in people with DM, but also indirectly affects health promotion behaviours through the level of diabetes distress. HBM theory provides a framework for understanding the networks that are responsible for health promotion behaviours. When older adults with DM experience loneliness, they are at risk for worsening health promotion behaviours that may lead to a worsening of their disease condition. This study validates the key role of diabetes distress in this process of change. The detection rate of diabetic distress in China is 50%,34 which surpasses that of Perrin et al (39%).35 This implies that there is a greater need for increased consideration of diabetes distress in China.

Traditionally, the focus has often been on improving health promotion behaviours by increasing disease knowledge among people with DM, while neglecting psychological factors. Findings from our research are intriguing because diabetes distress is important as an intervenable mediator in influencing health promotion behaviours during periods of loneliness caused by irresistible factors (eg, the COVID-19 pandemic). This study and previous studies indicated that diabetes distress can be used as a window for the intervention of psychological state and behaviour change in patients with DM. Future research and clinical work need to evaluate and intervene in disease distress in patients with DM.

Some limitations need to be considered in this study. First, self-report questionnaires which may result in recall and reporting bias. Second, this study did not control confounding factors. Third, the factors included in this study did not fully explain the network relationship between loneliness, diabetes distress and health promotion behaviours. Additional research is needed to determine how other negative psychological traits impact the health promotion behaviour of elderly patients with DM.


The present research aimed to examine the mediating role of diabetes distress in the relationship between loneliness and health promotion behaviour. When the loneliness of the elderly cannot be changed in a short period, improving diabetes distress can be used as a way to change their health promotion behaviours for disease management. Another broad recommendation is to pay more attention to the negative psychological states in the elderly.

Data availability statement

Data are available on reasonable request. The data supporting the findings of this study are available from the corresponding author on reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

The study was approved by the Ethics Committee of Xiangya School of Nursing, Central South University (E202283). All participants were informed of the purpose and process of this study before the study began. Our study obtained informed consent from each participant in writing prior to completing the questionnaire. For participants who were not literate enough to comprehend the informed consent form, explanations of each entry were provided on an individual basis. And their legally authorised representatives were also given informed consent. Participation in this study was voluntary. Participants may withdraw from this study at any time. All the participants were informed that they would bear no penalty for refusal to participate in the study and would be allowed to withdraw at any time without affecting their treatments. All information collected here is strictly confidential. Confidentiality and anonymity were maintained by not asking for names and by numbering the questionnaires (each participant received a number on the debriefing form). Any student wishing to withdraw data before data analysis can use this number to contact the researcher. All data were kept in a locked cabinet and on a password-protected computer to ensure privacy. All methods follow the relevant guidelines and regulations.

This post was originally published on https://bmjopen.bmj.com