Diabetes Self-Care Behaviors in Singapore and Their Associations With Patients’ Characteristics and Health Literacy

Abstract

Purpose

The purpose of this study was to examine the relationship between self-management behaviors (eg, healthy eating, being active, medication taking, glucose monitoring, feet check), sociodemographic factors, disease-related characteristics, and health literacy among patients with type 2 diabetes in Singapore.

Methods

Data were analyzed from a nationwide survey conducted between 2019 and 2020 (n = 387). Self-management behaviors were assessed using the Dietary Approaches to Stop Hypertension questionnaire, the Global Physical Activity Questionnaire, and a diabetes care questionnaire. A linear regression model was generated to examine the association of healthy eating with the variables of interest (sociodemographic factors, disease-related characteristics, and health literacy), and logistic regression models were generated to investigate the significant correlates of the remaining self-care behaviors.

Results

Regression models showed that the 5 self-care behaviors have different correlates. Nonetheless, compared to individuals aged 50 to 64 years, those aged 65 years and above were less likely to be active, adhere to their medication prescription, and check their feet. Individuals with a higher number of diabetes-related complications were less likely to be sufficiently active but more likely to monitor their glucose level and check their feet. Moreover, individuals with poor health literacy were more likely to eat healthily and be sufficiently active.

Conclusions

Programs related to self-care behaviors can be tailored to specific demographics to improve their uptake in the population. Furthermore, encouraging comprehensive self-care behaviors in those aged 65 years and above is crucial for effective diabetes management.

Diabetes mellitus is a chronic condition that can lead to complications, such as neuropathy, retinopathy, and cardiovascular diseases.^1,2 Although the rates of diabetes-related complications have decreased over the years, the burden of diabetes remains high due to its increasing prevalence.³ According to the International Diabetes Federation, approximately 1 in 10 adults aged 20 to 79 years were diagnosed with diabetes in 2021.⁴ Furthermore, projections indicated that this figure could reach 643 million by 2030 and surge to 783 million by 2045.⁴ The increasing prevalence of diabetes implies that examining interventions that can slow disease progression, such as reducing diabetes-related complications, remains relevant.

A cost-effective method of reducing diabetes-related complications is to promote self-care behaviors among individuals with diabetes.⁵ Seven components of self-care behaviors have been identified by the Association of Diabetes Care and Education Specialists: healthy coping, healthy eating, being active, taking medication, monitoring, reducing risks, and problem-solving.⁶ Studies have shown that interventions promoting self-care behaviors among patients with diabetes can lead to better health outcomes. A meta-analysis using data from controlled clinical trials showed that exercise was associated with a lower A1C levels.⁷ Another systematic review found that self-management training among patients with diabetes could improve blood glucose levels, promote weight loss, and increase quality of life.⁸ Nonetheless, it should be noted that the priority of these self-care behaviors should be tailored to the specific diabetes type and the patient’s current condition. For instance, a literature review has shown that self-monitoring of glucose levels may not be beneficial among patients with type 2 diabetes mellitus (T2DM) who were not on insulin.⁹

The prevalence and correlates of self-care behaviors can vary depending on the population of interest and the questionnaire used. A systematic review on the participation of self-care behaviors among patients with diabetes from low- and middle-income countries showed that the prevalence ranged from 29.9% to 91.7% for healthy eating, 26.0% to 97.0% for medication taking, 26.7% to 69.0% for being physically active, 13.0% to 79.0% for self-monitoring of glucose level, and 17.0% to 77.4% for feet care.¹⁰ Although the associations between certain factors, such as health literacy and education level,^11-13 and self-care behaviors are more established in the literature, the association between characteristics such as age and socioeconomic status and self-care behaviors shows varying results.¹⁴ Hence, it is pertinent to examine the prevalence and correlates of self-care behaviors within the population of interest. The findings can demonstrate the effectiveness of diabetes management interventions and enable healthcare providers to identify patients who are more or less inclined to engage in specific self-care behaviors.

Diabetes related studies hold value in Singapore, where the prevalence of diabetes is rising. According to Phan et al,¹⁵ T2DM in Singapore is projected to increase to 15% by 2050. Given this expected increase in prevalence, the government has declared a “War on Diabetes,” including various initiatives to prevent and manage the disease.¹⁶ Therefore, there is a need to understand actions that can slow the progression of diabetes among patients, such as self-care behaviors, and to identify individuals who may be less involved in these behaviors. Although several qualitative studies have explored self-care behaviors among individuals with diabetes in Singapore,^17,18 the study design limits the generalizability of these findings. Hence, this study aimed to examine self-care behaviors using scores and percentages and their associations with patients’ characteristics and health literacy among individuals with T2DM in Singapore.

Methods

Research Design

This cross-sectional analysis utilized data from a nationwide study that examined the knowledge, attitudes, and practices of diabetes in Singapore.¹⁹ Previous publications related to the study examined individuals with and without diabetes,¹⁹ but this study focused on only individuals with diabetes. The study was conducted between February 2019 and September 2020. Participants were included if they were Singapore citizens or permanent residents who were ≥18 years, resided in Singapore during the study, and literate in English, Malay, Chinese, or Tamil. Those participants who could not be reached due to incomplete addresses, residing outside of Singapore, or being institutionalized during the study were ineligible for participation. Those who were unable to understand the questionnaire due to cognitive difficulties and those with severe physical or mental disorders who were unable to answer the questionnaire on their own were also excluded from the study. In this analysis, only participants with self-reported type 2 diabetes mellitus were included.

Sampling Method

The study sample was obtained from a national administrative database comprising all Singapore residents. The study employed a disproportionate stratified sampling technique whereby the proportion of each ethnic group (Chinese, Malay, Indian) was fixed at 30% and each age bracket (18-34 years, 35-49 years, 50-64 years, and 65 years and above) was fixed at 20%. Based on a previous study showing that the prevalence of diabetes knowledge in Singapore was 60%,²⁰ the sample size calculated was 3000, accounting for a power of 0.8, type I error of 0.05, and an adjustment for the design effect.

Data Collection

The fieldwork was conducted by a survey firm, which was selected based on institutional guidelines.¹⁹ The interviewers from the survey firm underwent 2 weeks of training before they conducted the interviews. One to 2 weeks before the interviewers conducted house visits, an invitation letter was sent to the participants. The invitation letter included information about the study and a contact number if they have any inquiries. During the house visit, data were collected in real-time via a computer-assisted personal interview.

Written informed consent was obtained from all participants. Parental consent was also obtained for participants aged 18 to 20 years. Approval of the study was given by the Institute of Mental Health’s Institutional Research Review Committee and the National Healthcare Group’s Domain Specific Review Board (NHG DSRB 2018/00430).

Outcome: Self-Care Behaviors

The outcome examined in this analysis were the following self-care behaviors: healthy eating, being active, medication taking, glucose monitoring, and feet check. Although self-monitoring of glucose was only recommended for patients with T2DM who were on insulin, it was included in this analysis because previous studies have demonstrated that a notable proportion of individuals with T2DM without insulin prescription monitored their glucose level independently.^21,22 To ensure that the survey questions accurately capture the key aspects of self-care behaviors among individuals with diabetes, also known as content validity, the study team consulted experts (policymakers, clinicians, and epidemiologists) in the diabetes-related domain and performed cognitive interviews to ensure understanding with participants belonging to different sociodemographic groups.

Healthy eating was assessed using Dietary Approaches to Stop Hypertension (DASH) questionnaire. It evaluates the consumption frequency of 30 food and beverages for the preceding year using a 10-point scale, ranging from “never/rarely” to “6 or more times per day.”²³ The 30 food and beverages were categorized into 7 groups: fruits, vegetables, nuts/legumes, white grains, red and processed meat, low-fat dairy, and sweetened beverages.²³ Within each group, the scores were categorized into quintiles and reassigned values from 1 to 5 according to the quintiles that they fall in.²³ The scores from the 7 groups were aggregated to obtain the overall DASH score.²³ The questionnaire was validated by Whitton et al ²³ in the Singapore context, showing good correlation (Spearman’s rank-correlation coefficients: 0.50-0.51) with other established questionnaires: the Alternative Healthy Eating Index-2010 and the alternate Mediterranean Diet.

Being active was assessed using the Global Physical Activity Questionnaire (GPAQ).²⁴ It examines 3 domains of physical activity (work, transport, and leisure), considering both moderate and vigorous levels of intensity.²⁴ The energy expenditure, in metabolic equivalent (MET), MET/min, for each domain was calculated by multiplying the time spent with the MET. Moderate activities were assigned a MET value of 4, and vigorous activities were assigned a MET value of 8.²⁴ The scores from the 3 domains were summed to form the overall energy expenditure.²⁴ Individuals with ≥600 MET/min were considered to be active, as recommended by World Health Organization.²⁵ GPAQ was validated by Chu et al²⁶ in the Singapore context, showing a moderate correlation (Spearman’s rank correlation: 0.30-0.46) with accelerometer-measured physical activity.

Medication taking was assessed using related questions adapted from the Summary of Diabetes Self-Care Activities measure (SDSCA), a brief self-reported questionnaire devised by Toobert et al.²⁷ The participants were initially asked whether they were prescribed insulin injections or oral medications.²⁷ Among those who were prescribed insulin injections, participants were considered taking their prescription if they administered the injection for a continuous 7-day period within the past week.²⁷ Similarly, among those prescribed oral medication, participants were considered taking their prescription if they took oral medication for a continuous 7-day span during the preceding week.²⁷

Self-monitoring of glucose was assessed by initially asking the participants to select their primary approach for checking their glucose level from the following 4 methods: undergoing blood or urine test at the doctor’s office only, conducting flash glucose monitoring/continuous glucose monitoring at home, conducting blood glucose tests using blood glucose test strips at home, and using urine glucose strips at home. Subsequently, they were asked the frequency with which they assessed their glucose levels last month, with options spanning from “never” to “2 or more times a day.” Participants were classified as monitoring their glucose level independently if their primary approach of glucose monitoring was in a home setting and if they conducted at least 1 check in the last month. Conversely, participants were considered not to monitor their glucose level independently if their primary method for glucose monitoring was in a clinical setting or if they did not check their glucose level last month.

Feet check was determined by asking, “How often have you examined your feet in the last week?” Responses were categorized as “no” if participants had not performed feet checks. If participants reported conducting feet checks once or more a week, responses were classified as “yes.”

Variables of Interest: Patient’s Characteristics and Health Literacy

The following sociodemographics were included in the analysis to examine their association with self-care behaviors: age, gender, ethnicity, education, marital status, employment status, personal income, body mass index, and number of other chronic conditions. The assessment of chronic conditions was done through a self-reported checklist containing 18 chronic conditions. The association between diabetes-related characteristics (number of diabetes-related complications, duration of diabetes, type of diabetes, on insulin, and on oral medications) and self-care behaviors were also investigated. Regarding diabetes-related complications, the following 6 complications were assessed: retinopathy, nephropathy, neuropathy, peripheral vascular disease, lower limb ulcers, and gangrene.

Health literacy was evaluated using the Brief Health Literacy Screen questionnaire.²⁸ The questionnaire examined 3 dimensions of health literacy: difficulty in understanding written information about medical conditions, ability to read hospital materials, and confidence in completing medical forms. The participants rated the questions using a 5-point scale ranging from “all the time” to “none of the time” for the first 2 dimensions and “extremely” to “not at all” for the last dimension. Reverse-scoring was performed for the last dimension. The total score ranges from 3 to 15, with higher scores representing better literacy. This study categorized the score into “inadequate health literacy” and “adequate health literacy” using a cutoff score of 9, as established in earlier studies.^29,30 The construct validity of the questionnaire was demonstrated in other studies through its correlation with similar scales, such as the Short Test of Functional Health Literacy in Adults²⁸ and Rapid Estimate of Adult Literacy in Medicine.³¹ The Cronbach’s alpha was 0.86, implying that the scale has good internal consistency.

Statistical Analysis

Survey weights were included in the analysis to account for the survey design, nonresponse, and poststratification by age groups and ethnicities. Continuous variables were summarized using mean and standard deviation, and categorical variables were presented in weighted proportions and unweighted counts. Linear regression was performed to examine the associations between patients’ characteristics and health literacy with dietary behavior. For the remaining 4 self-care behaviors, logistic regressions were performed to examine their association with patients’ characteristics and health literacy. Beta coefficient and 95% CI were presented for linear regression, and odds ratio (OR) and 95% CI were presented for logistic regressions. The regression model assessing medication taking as the outcome did not include the covariates “on insulin” and “on oral medications” because these variables are analogous.

Standard errors were estimated using Taylor’s series linearization method to account for the survey design. All analyses were performed using STATA/MP version 18, with 2-sided tests at a 5% significance level.

Results

A total of 387 participants with T2DM were included in the analysis. Table 1 summarizes the characteristics of the sample. Most participants were aged 50 to 64 years (49.9%, n = 169), male (57.2%, n = 198), of Chinese ethnicity (66.4%, n = 62), with primary school or below educational qualification (40.4%, n = 150), employed (52.2%, n = 186), with either no personal income or personal income below SGD2000 (67.0%, n = 261), overweight (42.0%, n = 134), and 2 or more chronic conditions (59.9%, n = 223). For diabetes-related characteristics, most of them had no diabetes-related complications (68.4%, n = 260), had diabetes for <5 years (33.0%, n = 102), and were on oral medication (88.4%, n = 350). A majority of the participants had adequate health literacy (63.7%, n = 266).

Table 1.

Characteristics of the Sample (N = 387)^a

	Weighted %	Unweighted n
Age group, y
18-49	13.2%	49
50-64	49.9%	169
65 and above	36.9%	169
Gender
Female	42.8%	189
Male	57.2%	198
Ethnicity
Chinese	66.4%	62
Malay	17.7%	144
Indian	14.3%	166
Others	1.5%	15
Education level
Primary and below	40.4%	150
Secondary	27.6%	122
A-level/polytechnic/vocational school/ITE	20.0%	71
Degree and above	12.0%	44
Employment status
Employed	52.2%	186
Economically active	41.9%	184
Unemployed	5.9%	17
Personal income (SGD)
No income/below 2000	67.0%	261
2000-3999	17.2%	63
4000-5999	7.9%	22
6000 and above	7.9%	28
Body mass index (international)
Underweight and normal	37.9%	107
Overweight	42.0%	134
Obese	20.1%	86
No. chronic conditions
0	13.6%	70
1	26.5%	92
2 or more	59.9%	223
No. complications
0	68.4%	260
1	18.4%	74
2 or more	13.2%	53
Duration of diabetes
<5 y	33.0%	102
5-9 y	19.5%	87
10-14 y	17.9%	74
≥15 y	29.6%	122
Treatment with insulin	8.9%	47
Treatment with oral medication	88.4%	350
Health literacy
Poor	36.3%	120
Adequate	63.7%	266

Abbreviation: ITE, Institute of Technical Education.

Missing data: Personal income (n = 13), body mass index (n = 60), number of chronic conditions (n = 2), duration of diabetes (n = 2), health literacy (n = 1).

The mean (SD) DASH score was 20.5 (6.0). The prevalences of self-care behaviors were as follows: 73.0% (n = 280) of participants were physically active, 96.0% (n = 335) reported medication taking, 35.2% (n = 175) monitored their glucose level at home, and 70.0% (n = 290) checked their feet in the past 1 week.

Table 2 presents the association between patients’ characteristics and health literacy with healthy eating. Individuals with diabetes were more likely to eat healthily if they were female (vs male, β = 2.6, 95% CI, 0.8-4.4), of Indian ethnicity (vs Chinese, β = 2.3, 95% CI, 0.8-3.9), had degree and above educational qualification (vs no formal education/primary school, β = 3.6, 95% CI, 0.2-6.9), with 1 chronic condition (vs no chronic condition, β = 3.1, 95% CI, 1.0-5.2), and had poor health literacy (vs adequate, β = 1.8, 95% CI, 0.2-3.4).

Table 2.

Linear Regression Model for the Association Between Patient’s Characteristics and health Literacy With Diet

	Diet (DASH score)
	β (95% CI)	P
Age group, y
18-49	1.7 (−1.1 to 4.6)	0.225
50-64
65 and above (reference)	0.6 (−1.3 to 2.6)	0.534
Gender
Male (reference)
Female	2.6 (0.8 to 4.4)	0.005
Ethnicity
Chinese (reference)
Malay	0.9 (−0.8 to 2.5)	0.308
Indian	2.3 (0.8 to 3.9)	0.003
Others	3.4 (−0.1 to 6.9)	0.056
Education level
Primary and below (reference)
Secondary	0.2 (−1.6 to 2.1)	0.795
A-level/polytechnic/vocational school/ITE	0.8 (−2.4 to 3.9)	0.624
Degree and above	3.6 (0.2 to 6.9)	0.036
Employment status
Employed (reference)
Economically active	1.6 (−0.2 to 3.4)	0.084
Unemployed	0.3 (−3.4 to 4.0)	0.880
Personal income (SGD)
No income/below 2000 (reference)
2000-3999	0.6 (−2.1 to 3.3)	0.661
4000-5999	0.8 (−2.4 to 3.9)	0.626
6000 and above	0.6 (−3.3 to 4.6)	0.750
Body mass index (international)
Underweight & normal	−0.4 (−2.2 to 1.3)	0.644
Overweight (reference)
Obese	−0.1 (−2.2 to 2.0)	0.940
No. chronic conditions
0 (reference)
1	3.1 (1.0 to 5.2)	0.003
2 or more	−0.3 (−2.1 to 1.4)	0.698
No. complications
0 (reference)
1	−0.2 (−2.3 to 1.9)	0.836
2 or more	−1.6 (−3.6 to 0.5)	0.129
Duration of diabetes
<5 y (reference)
5-9 y	0.3 (−2.4 to 3.0)	0.831
10-14 y	1.1 (−0.8 to 3.1)	0.261
≥15 y	0.7 (−1.4 to 2.8)	0.526
Treatment with insulin	2.0 (−0.7 to 4.7)	0.143
Treatment with oral medication	−0.6 (−2.5 to 1.4)	0.558
Poor health literacy vs adequate health literacy	1.8 (0.2 to 3.4)	0.024

Abbreviation: DASH, Dietary Approaches to Stop Hypertension; ITE, Institute of Technical Education.

Bold values indicate statistical significance (P < 0.05).

Table 3 shows the regression models for the remaining 4 self-care behaviors. Individuals with diabetes were more likely to be active if they were of other ethnicities (vs Chinese, OR: 12.2, 95% CI, 1.0-146.7), had diabetes for 5 to 9 years (vs <5 years, OR: 3.8, 95% CI, 1.1-13.2) or ≥15 years (vs <5 years, OR: 6.7, 95% CI, 1.8-24.6), and had poor health literacy (vs adequate health literacy, OR: 3.9, 95% CI, 1.1-13.6). They were less likely to be active if they were aged 65 years and above (vs 50-64 years, OR: 0.1, 95% CI, 0.03-0.3), had 1 or more complications (vs no chronic condition, OR for 1 complication: 0.3, 95% CI, 0.1-1.0; OR for 2 or more complications: 0.1, 95% CI, 0.02-0.5).

Table 3.

Regression Models for the Association Between Patient’s Characteristics and Health Literacy With Physical Activity, Medication Taking, Self-Monitoring of Glucose Level, and Feet Check

	Physical activity (GPAQ score)		Medication taking		Self-monitoring of glucose level		Feet check
	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Age group, years
18 to 49	4.9(0.9-27.2)	0.072	0.1(0.01-0.9)	0.039	0.7(0.2-1.8)	0.426	0.4(0.1-1.3)	0.113
50-64	Ref		Ref		Ref		Ref
65 and above	0.1(0.03-0.3)	<0.001	0.2(0.1-0.9)	0.029	0.8(0.2-2.7)	0.750	0.1(0.04-0.4)	<0.001
Gender
Male	Ref		Ref		Ref		Ref
Female	0.9(0.3-2.2)	0.773	0.8(0.3-2.1)	0.617	1.3(0.5-3.5)	0.561	1.1(0.4-3.3)	0.874
Ethnicity
Chinese	Ref		Ref		Ref		Ref
Malay	1.7(0.5-5.4)	0.357	0.2(0.04-1.2)	0.081	4.0(1.5-10.8)	0.006	2.0(0.8-5.0)	0.121
Indian	0.7(0.3-1.8)	0.446	0.3(0.1-1.4)	0.122	4.6(2.0-10.3)	<0.001	1.7(0.7-4.1)	0.232
Others	12.2(1.0-146.7)	0.048	0.1(0.02-5.0)	0.254	2.9(0.5-15.6)	0.208	1.5(0.3-7.9)	0.601
Education level
Primary and below	Ref		Ref		Ref		Ref
Secondary	2.1(0.7-6.4)	0.218	2.7(0.3-21.4)	0.338	1.3(0.4-4.2)	0.704	0.7(0.2-2.7)	0.633
A-level/polytechnic/vocational school/ITE	2.0(0.3-11.2)	0.446	1.8(0.3-13.0)	0.538	2.1(0.6-8.3)	0.269	0.8(0.2-4.0)	0.775
Degree and above	0.9(0.1-6.1)	0.881	2.5(0.1-89.4)	0.622	1.3(0.2-9.9)	0.831	4.1(0.5-31.1)	0.178
Employment status
Employed	Ref		Ref		Ref		Ref
Economically active	1.0(0.3-3.4)	1.000	0.4(0.1-1.4)	0.161	0.8(0.2-2.6)	0.704	1.3(0.4-4.1)	0.622
Unemployed	1.8(0.1-25.5)	0.667	0.3(0.01-6.0)	0.397	1.0(0.2-4.2)	0.977	4.2(0.6-29.0)	0.145
Personal income (SGD)
No income/below 2000	Ref		Ref		Ref		Ref
2000-3999	0.6(0.1-2.4)	0.472	0.7(0.1-3.0)	0.586	0.6(0.2-1.7)	0.352	0.8(0.2-2.8)	0.706
4000-5999	1.2(0.2-7.6)	0.867	0.8(0.04-16.1)	0.885	3.6(0.7-17.8)	0.118	0.6(0.1-2.6)	0.476
6000 and above	1.4(0.2-10.6)	0.767	10.0(0.5-203.7)	0.134	0.8(0.2-3.9)	0.773	0.9(0.1-6.3)	0.879
Body mass index (international)
Underweight & normal	2.0(0.7-5.3)	0.180	2.5(0.3-18.0)	0.365	1.2(0.5-2.8)	0.735	0.6(0.2-1.6)	0.300
Overweight	Ref		Ref		Ref		Ref
Obese	1.2(0.3-4.0)	0.797	4.6(0.6-35.1)	0.143	0.9(0.4-2.2)	0.868	1.9(0.6-5.8)	0.242
No. chronic conditions
0	Ref		Ref		Ref		Ref
1	0.5(0.1-2.1)	0.314	2.7(0.5-14.4)	0.253	0.9(0.2-3.0)	0.803	2.1(0.4-10.8)	0.388
2 or more	0.5(0.1-1.6)	0.221	6.0(1.5-23.6)	0.010	1.3(0.4-3.6)	0.670	0.6(0.2-2.2)	0.452
No. complications
0	Ref		Ref		Ref		Ref
1	0.3(0.1-1.0)	0.041	0.4(0.1-1.3)	0.118	7.5(2.3-24.3)	0.001	1.0(0.4-2.6)	0.953
2 or more	0.1(0.02-0.5)	0.003	—^a		1.0(0.3-3.1)	0.959	5.6(1.1-30.0)	0.043
Duration of diabetes
<5 years	Ref		Ref		Ref		Ref
5-9 years	3.8(1.1-13.2)	0.039	1.3(0.2-9.6)	0.782	2.5(0.9-7.2)	0.090	1.5(0.4-5.2)	0.517
10-14 years	2.8(0.8-10.1)	0.123	2.3(0.2-28.3)	0.517	0.5(0.1-2.5)	0.410	2.0(0.5-7.4)	0.308
≥15 years	6.7(1.8-24.6)	0.004	0.4(0.1-2.5)	0.345	1.8(0.6-5.5)	0.269	4.8(1.5-15.3)	0.008
Treatment with insulin	7.9(0.8-74.2)	0.070	—^b		6.6(1.6-27.4)	0.010	1.3(0.3-6.1)	0.729
Treatment with oral medication	2.0(0.5-7.5)	0.328	—^b		8.2(1.9-34.6)	0.004	0.7(0.1-3.9)	0.692
Health literacy
Adequate health literacy	Ref		Ref		Ref		Ref
Poor health literacy	3.9(1.1-13.6)	0.034	0.8(0.2-2.9)	0.737	1.2(0.4-3.8)	0.742	1.3(0.4-4.0)	0.680

Abbreviation: GPAQ, Global Physical Activity Questionnaire; ITE, ; Ref, reference.Bold values indicate statistical significance (P < 0.05).

Insufficient sample size.

Not included in the model as these variables are analogous to the outcome variable.

Medication taking was less likely among individuals with diabetes who were aged 18 to 49 years (vs 50 to 64, OR: 0.1, 95% CI, 0.01-0.9) or 65 years and above (vs 50 to 64, OR: 0.2, 95% CI, 0.1-0.9) and more likely among those with 2 or more chronic conditions (vs no chronic condition, OR: 6.0, 95% CI, 1.5-23.6). Self-monitoring of glucose was more likely among individuals with diabetes who were of Malay (vs Chinese, OR: 4.0, 95% CI, 1.5-10.8) or Indian ethnicity (vs Chinese, OR: 4.6, 95% CI, 2.0-10.3), had 1 complication (vs no complication, OR: 7.5, 95% CI, 2.3-24.3), and treated with insulin (OR: 6.6, 95% CI: 1.6-27.4) or oral medication (OR: 8.2, 95% CI, 1.9-34.6). Individuals were more likely to examine their feet if they had 2 or more complications (vs no complication, OR: 5.6, 95% CI, 1.1-30.0) and had diabetes for ≥15 years (vs <5 years, OR: 4.8, 95% CI, 1.5-15.3). They were less likely to check their feet if they were aged 65 years and above (vs 18-49 years, OR: 0.1, 95% CI, 0.04-0.4).

Discussion

This study found that except for self-monitoring of glucose, over half of the patients with T2DM engaged in the corresponding self-care behaviors. Moreover, among the 4 self-care behaviors examined (excluding healthy eating), medication taking had the highest prevalence (96.1%), and self-monitoring of glucose had the lowest prevalence (34.7%). For self-monitoring of glucose, its lower prevalence is not surprising because it is only recommended to patients with T2DM if they have poorly managed diabetes or are on insulin.³²

The observation that medication taking was the most prevalent among the self-care behaviors examined is consistent with several studies.^22,33 For instance, a study on the Arab population utilized the SDSCA questionnaire and found that medication taking had the highest prevalence (92.9%) and that exercise had the lowest prevalence (27.1%).²² However, the absolute percentage of medication taking should be interpreted with caution. First, the prevalence of medication taking may vary depending on the self-reported scale used and the duration over which patients were required to recall. A systematic review on medication taking among patients with diabetes found that the prevalence of medication taking ranged from 38.5% to 93.1%.³⁴ Second, questions related to medication taking in this study may have a ceiling effect due to possible reasons such as respondent bias.²⁷

This study assessed healthy eating using DASH, which is suitable because this type of eating plan has been shown to improve body mass index, reduce blood pressure, and lower insulin resistance among patients with diabetes.³⁵ However, a limitation of DASH is that it lacks a cutoff score that defines healthy eating. Furthermore, other studies have different ways of calculating DASH scores, thus making it difficult to compare with these findings.^36,37 Nevertheless, a previous publication examining the dietary patterns of the Singapore population, stratified by the number of chronic conditions, revealed that the DASH scores among those without chronic conditions (score: 18.5) are similar to those observed in this study.³⁸ Additional analysis done in this study also showed no significant difference between those with diabetes and those without chronic conditions when adjusted with sociodemographic characteristics (P = 0.133). These results imply that individuals with diabetes may have a similar dietary pattern to those without chronic conditions. These findings can serve as a reference for future local studies to make comparative assessments of the scores.

Self-care behaviors were associated with specific respondents’ characteristics. Individuals aged 65 years and above were less inclined to be active, take their medication, and check their feet than those aged 50 to 65 years. This finding corroborates with a population-based study in Germany showing that older age was associated with a lower likelihood of being active and feet check. A qualitative study among older patients with diabetes in Singapore also demonstrated that although they acknowledged the importance of physical activity and medication taking, they struggled to find time for exercise and tended to forget their medication.¹⁷ This finding implies that those who are aged 65 years and above should be prioritized for interventions that encourage self-care behaviors.

Furthermore, the number of complications was associated with lower odds of being active, but higher odds for self-management of glucose and feet check. Given that patients with more complications may face constraints in participating in physical activity, they may compensate by engaging in other self-care behaviors, such as self-monitoring of glucose and feet check. However, no significant association was observed between having 2 or more complications and self-management of glucose. Although this study was a secondary analysis and may have limited statistical power to detect differences, it is recommended that future studies in Singapore verify this finding and examine the causes of the lack of glucose monitoring among individuals with more severe diabetes.

The current study found that individuals with lower health literacy were more likely to eat healthily and be sufficiently active. In the context of diabetes, previous studies have reported a positive association between health literacy and self-care behaviors.^11,12 For example, a study on older people with T2DM in Indonesia revealed that individuals with better diabetes literacy were 2 times more likely to practice self-management behavior.¹² Nonetheless, in the context of other medical conditions, a study on patients with chronic kidney disease (CKD) found a similar association as this study. It was observed that among individuals with CKD, those with lower health literacy were less likely to consume fast food.³⁹ The study suggests that patients with lower health literacy may be given more opportunities to participate in educational interventions that promote healthy lifestyles.³⁹ Moreover, healthcare providers may emphasize lifestyle changes among these patients, especially when lifestyle modification is known to manage the progression of chronic conditions effectively.³⁹ Because several explanations are possible for this finding, including analysis constraints such as residual confounding, future related studies may consider validating this result.

This study has several implications. First, healthcare professionals can explore strategies to motivate individuals aged 65 years and above to engage in self-care behaviors. This implication is particularly important to Singapore due to its aging population, with an estimated 25% of the population projected to be aged 65 years and above by 2030.⁴⁰ Second, except for age, different correlates were observed for the 5 self-care behaviors. Interventions related to self-care behaviors can tailor programs for particular demographics related to the specific self-care behavior under consideration. Third, the results imply that healthcare providers should refrain from assuming that patients with adequate health literacy are more likely to eat healthily and be active. Hence, discussions concerning healthy eating and being active should be initiated for all patients with diabetes regardless of their health literacy levels.

Because this study adopts a population-wide approach instead of recruiting from clinical settings, the results are representative of the Singapore population. However, as mentioned, causality cannot be inferred from the findings because it is a cross-sectional analysis. Additionally, this study is a secondary analysis, which implies that potential confounders not collected in the study, such as self-efficacy of self-care behaviors, could not be adjusted for. Because the study did not use existing instruments on self-care behaviors, it is challenging to compare with research using the same instrument. Subsequent studies in Singapore can consider validating existing questionnaires on self-care behaviors to facilitate cross-study comparisons.

Conclusion

Among individuals with T2DM, more than half of them engaged in the corresponding self-care behaviors except for self-monitoring of glucose levels. Moreover, those aged 65 years and above were associated with lower odds of being active, medication taking, and feet check. Hence, it is salient to focus on this group and encourage them to engage in self-care behaviors.

Footnotes

Author Contributions

YSK and MS developed the analysis conceptually. YSK conducted the analysis and wrote the manuscript. EA provided guidance on the statistical analysis. PVA, FD, KR, PW, EA, CFS, ESL, SAC, and MS participated in the study planning and conduct and reviewing of the study. All authors contributed to the substantial revision of the manuscript and offered intellectual insights. All authors reviewed and endorsed the final manuscript.

Declaration of Conflicting Interests

The authors declare that there is no conflict of interest

Funding

This work was supported by the National Medical Research Council of Singapore (NMRC/HSRG/0085/2018).

ORCID iD

Yen Sin Koh

References

American Diabetes Association. Implications of the United Kingdom prospective diabetes study. Diabetes Care. 2003;26(suppl 1):s28-s32. doi:10.2337/diacare.26.2007.S28

Deshpande

Harris-Hayes

Schootman

Epidemiology of diabetes and diabetes-related complications. Phys Ther. 2008;88(11):1254-1264. doi:10.2522/ptj.20080020

Gregg

Wang

, et al. Changes in diabetes-related complications in the United States, 1990–2010. N Engl J Med. 2014;370(16):1514-1523. doi:10.1056/NEJMoa1310799

International Diabetes Federation. IDF Diabetes Atlas. 10th ed. 2021. https://www.diabetesatlas.org

Ausili

Bulgheroni

Ballatore

, et al. Self-care, quality of life and clinical outcomes of type 2 diabetes patients: an observational cross-sectional study. Acta Diabetol. 2017;54(11):1001-1008. doi:10.1007/s00592-017-1035-5

Association of Diabetes Care and Education Specialists, Kolb L. An effective model of diabetes care and education: the ADCES7 self-care behaviors^TM. Sci Diabetes Self Manag Care. 2021;47(1):30-53. doi:10.1177/0145721720978154

Boulé

Haddad

Kenny

Wells

Sigal

RJ.

Effects of exercise on glycemic control and body mass in type 2 diabetes mellitus: a meta-analysis of controlled clinical trials. JAMA. 2001;286(10):1218. doi:10.1001/jama.286.10.1218

Norris

Engelgau

Venkat Narayan

KM.

Effectiveness of self-management training in type 2 diabetes. Diabetes Care. 2001;24(3):561-587. doi:10.2337/diacare.24.3.561

Faas

Schellevis

Eijk

JTMV

. The efficacy of self-monitoring of blood glucose in NIDDM subjects: a criteria-based literature review. Diabetes Care. 1997;20(9):1482-1486. doi:10.2337/diacare.20.9.1482

10.

Mogre

Johnson

Tzelepis

Shaw

Paul

A systematic review of adherence to diabetes self-care behaviours: evidence from low- and middle-income countries. J Adv Nurs. 2019;75(12):3374-3389. doi:10.1111/jan.14190

11.

van Der Heide

Uiters

Rademakers

Struijs

Schuit

Baan

CA.

Associations among health literacy, diabetes knowledge, and self-management behavior in adults with diabetes: results of a Dutch cross-sectional study. J Health Commun. 2014;19 (suppl 1):115-131. doi:10.1080/10810730.2014.936989

12.

Rachmawati

Sahar

Wati

DNK

. The association of diabetes literacy with self-management among older people with type 2 diabetes mellitus: a cross-sectional study. BMC Nurs. 2019;18(S1):34. doi:10.1186/s12912-019-0354-y

13.

Chowdhury

Fatema

Hossain

, et al. Knowledge and self-care practice regarding diabetes among type 2 diabetics: experience from a non-profit hospital chain in Bangladesh. Int J Diabetes Dev Ctries. 2018;38(4):478-485. doi:10.1007/s13410-017-0597-7

14.

Zarei

Saleh

Arziev

, et al. Investigating the factors affecting self-care behaviors in diabetic patients: a systematic review. J Biol Res. 2022;95(1). doi:10.4081/jbr.2022.10261

15.

Phan

Alkema

Tai

, et al. Forecasting the burden of type 2 diabetes in Singapore using a demographic epidemiological model of Singapore. BMJ Open Diabetes Res Care. 2014;2(1):e000012. doi:10.1136/bmjdrc-2013-000012

16.

Ministry of Health Singapore. War on diabetes. 2021. Accessed August 23, 2023. https://www.moh.gov.sg/wodcj

17.

Tan

CCL

Cheng

KKF

Sum

Shew

JSH

Holydard

Wang

. Perceptions of diabetes self-care management among older Singaporeans with type 2 diabetes: a qualitative study. J Nurs Res. 2018;26(4):242-249. doi:10.1097/jnr.0000000000000226

18.

Ong

Lee

Low

Vasanwala

Attitudes and perceptions of people with diabetes mellitus on patient self-management in diabetes mellitus: a Singapore hospital’s perspective. Singapore Med J. 2023;64(7):467-474. doi:10.11622/smedj.2022006

19.

AshaRani

Abdin

Kumarasan

, et al. Study protocol for a nationwide Knowledge, Attitudes and Practices (KAP) survey on diabetes in Singapore’s general population. BMJ Open. 2020;10(6):e037125. doi:10.1136/bmjopen-2020-037125

20.

Wee

SC.

Public awareness of diabetes mellitus in Singapore. Singapore Med J. 2002;43(3):128-134.

21.

Harris

MI.

Frequency of blood glucose monitoring in relation to glycemic control in patients with type 2 diabetes. Diabetes Care. 2001;24(6):979-982. doi:10.2337/diacare.24.6.979

22.

Al-Ozairi

Taghadom

Irshad

Al-Ozairi

Association between depression, diabetes self-care activity and glycemic control in an Arab population with type 2 diabetes. Diabetes Metab Syndr Obes. 2023;16:321-329. doi:10.2147/DMSO.S377166

23.

Whitton

JCY

Rebello

van Dam

RM.

Relative validity and reproducibility of dietary quality scores from a short diet screener in a multi-ethnic Asian population. Public Health Nutr. 2018;21(15):2735-2743. doi:10.1017/S1368980018001830

24.

Armstrong

Bull

Development of the world health organization global physical activity questionnaire (GPAQ). J Public Health. 2006;14(2):66-70. doi:10.1007/s10389-006-0024-x

25.

World Health Organization. Global recommendations on physical activity for health. Published 2010. Accessed August 8, 2024. https://www.who.int/publications/i/item/9789241599979

26.

Chu

AHY

SHX

Koh

Müller-Riemenschneider

. Reliability and validity of the self- and interviewer-administered versions of the Global Physical Activity Questionnaire (GPAQ). PLoS One. 2015;10(9):e0136944. doi:10.1371/journal.pone.0136944

27.

Toobert

Hampson

Glasgow

RE.

The summary of diabetes self-care activities measure: results from 7 studies and a revised scale. Diabetes Care. 2000;23(7):943-950. doi:10.2337/diacare.23.7.943

28.

Chew

Bradley

Boyko

EJ.

Brief questions to identify patients with inadequate health literacy. Fam Med. 2004;36(8):588-594.

29.

Cavanaugh

Osborn

Tentori

Rothman

Ikizler

Wallston

KA.

Performance of a brief survey to assess health literacy in patients receiving hemodialysis. Clin Kidney J. 2015;8(4):462-468. doi:10.1093/ckj/sfv037

30.

McNaughton

Cawthon

Kripalani

Liu

Storrow

Roumie

CL.

Health literacy and mortality: a cohort study of patients hospitalized for acute heart failure. J Am Heart Assoc. 2015;4(5):e001799. doi:10.1161/JAHA.115.001799

31.

Chew

Griffin

Partin

, et al. Validation of screening questions for limited health literacy in a large VA outpatient population. J Gen Intern Med. 2008;23(5):561-566. doi:10.1007/s11606-008-0520-5

32.

Kirk

Stegner

Self-monitoring of blood glucose: practical aspects. J Diabetes Sci Technol. 2010;4(2):435-439. doi:10.1177/193229681000400225

33.

Tan

Magarey

Self-care practices of Malaysian adults with diabetes and sub-optimal glycaemic control. Patient Educ Couns. 2008;72(2):252-267. doi:10.1016/j.pec.2008.03.017

34.

Krass

Schieback

Dhippayom

Adherence to diabetes medication: a systematic review. Diabet Med. 2015;32(6):725-737. doi:10.1111/dme.12651

35.

Campbell

AP.

DASH eating plan: an eating pattern for diabetes management. Diabetes Spectr. 2017;30(2):76-81. doi:10.2337/ds16-0084

36.

Liese

Nichols

Sun

D’Agostino

Haffner

SM.

Adherence to the DASH diet is inversely associated with incidence of type 2 diabetes: the insulin resistance atherosclerosis study. Diabetes Care. 2009;32(8):1434-1436. doi:10.2337/dc09-0228

37.

Barnes

Crandell

Bell

, et al. Change in DASH diet score and cardiovascular risk factors in youth with type 1 and type 2 diabetes mellitus: the SEARCH for diabetes in youth study. Nutr Diabetes. 2013;3(10):e91. doi:10.1038/nutd.2013.32

38.

Tan

YWB

Lau

AshaRani

, et al. Dietary patterns of persons with chronic conditions within a multi-ethnic population: results from the nationwide knowledge, attitudes and practices survey on diabetes in Singapore. Arch Public Health. 2022;80(1):62. doi:10.1186/s13690-022-00817-2

39.

Wong

Velasquez

Powe

Tuot

DS.

Association between health literacy and self-care behaviors among patients with chronic kidney disease. BMC Nephrol. 2018;19(1):196. doi:10.1186/s12882-018-0988-0

40.

Prime Minister’s Office Singapore. PM Lee Hsien Loong at the “Singapore Ageing: Issues and Challenges Ahead.” 2023. Accessed October 16, 2023. https://www.pmo.gov.sg/Newsroom/PM-Lee-Hsien-Loong-at-the-Singapore-Ageing-Issues-and-Challenges-Ahead-Book-Launch