RN, NP, M.Sc (Public Health), a lecturer at School of Public Health, Walailak University
BACKGROUND: Type 2 diabetes, especially in the elderly, continues to plague the world. Thailand – a developing country – is not immune to these ravage effects and their distressing upsurge in health and economic societal burdens. Self-care management is an essential strategy to prevent complications and reduce type 2 diabetes complications.
AIM: This study aimed to examine the treatment outcome and factors predicting diabetes self-care behaviors among elderly in Thailand.
METHODS: A cross-sectional correlative predictive design using multiple linear regression models to evaluate data in elderly type 2 diabetics in Thailand (August through December 2017) to assess perceived behavioral control on diabetic self-care management. One hundred thirty-four participant’s data were collected via questionnaire along with individual health records becoming the foundation of this study.
RESULTS: Most patients controlled glycemic outcome (77.9 %) through self-care behaviors at moderate rates (majority – 55.9%). Subjective norms and perceived control strongly correlated with behavioral intention and self-care behaviors. Perceived behavioral control was the most important factor predicting intentions (β 4.025, p < .01) and self-care management behavior (β15.258, p < .001). Patients responding to items regarding self-care behavior for diet, exercise and medication adherence showed favorable outcomes.
CONCLUSION: More than half of the patients had moderate levels in self-care management and the majority had good glycemic outcomes. From the analysis, we find that perceived behavioral control is critical to predicting behavioral intention and diabetic self-care behavior among the elderly.
АКТУАЛЬНОСТЬ. В мире продолжает расти заболеваемость диабетом 2 типа, особенно у пожилых людей. Таиланд, будучи развивающейся страной, не застрахован от этого разрушительного заболевания, а также его тревожащего роста в сфере общественного здоровья и экономического бремени для общества. Самостоятельный контроль за показателями гликемии представляет собой важную стратегию для предотвращения и уменьшения осложнений диабета 2 типа.
ЦЕЛЬ. Данное исследование направлено на изучение результатов лечения, а также факторов, прогнозирующих самоконтроль у пациентов с сахарным диабетом среди пожилых людей в Таиланде.
МЕТОДЫ. Был проведен перекрестный коррелятивный прогностический анализ с использованием моделей множественной линейной регрессии для оценки данных у пожилых пациентов с сахарным диабетом 2 типа в Таиланде (с августа по декабрь 2017 года) для оценки самостоятельного контроля за уровнем гликемии. Данные 134 участников исследования были получены путем заполнения опросника, а также анализа медицинских карт.
РЕЗУЛЬТАТЫ. Большинство пациентов самостоятельно контролировали уровень гликемии (77,9%), причем у большинства показатели были умеренными (55,9%). Субъективные нормы и воспринимаемый контроль тесно связаны с намерениями пациентов. Наиболее важными факторами, определяющими намерения, были воспринимаемый поведенческий контроль (β4,025, р<0,01) и поведение по самоконтролю (β15,258, р<0,001). У пациентов, которые отвечали на вопросы, касающиеся поведения по самоконтролю за диетой, физическими упражнениями и режимом приема лекарств, наблюдались благоприятные результаты.
ЗАКЛЮЧЕНИЕ. Более чем у половины пациентов отмечались умеренные уровни самоконтроля, и у большинства были хорошие гликемические показатели. В результате анализа было выявлено, что воспринимаемый поведенческий контроль имеет решающее значение для прогнозирования поведенческих намерений и поведения по самоконтролю за диабетом среди пожилых людей.
The global diabetes prevalence has nearly doubled since 1980, rising from 4.7% to 8.5% in the adult population [
Previous studies have shown around one third of Asian patients had adequately controlled T2DM, with glycemic hemoglobin (A1C) levels < 7%, as recommended by the American Diabetes Association (ADA). In Thailand (2016) the prevalence of diabetes in adults was reported at 8.3% [
Diabetes prevalence in Nakhon Si Thammarat province also had continuity increased that showed T2DM patients from 2015 (4,409 cases), 2016 (4,797 cases) and 2017 (5,015 cases) per 100,000 population respectively [
The Theory of Planned Behaviors (TPB: see Fig.1) was developed by Ajzen, I. (1991) [
Fig. 1. The Conceptual Frame Work and The Results in Predicting Factors in Regression Model
This study is aimed to examine diabetic self-care management (DSCM) among elderly with T2DM, to analyze the correlated factors and to clarify DSCM by behavioral intention (BI) and perceived behavioral control (PBC).
A cross-sectional descriptive-relative correlational design was conducted to examine a diabetic self-care management practice of 212 T2DM patients. The study sample includes patients previously diagnosed with T2DM who attended and registered at the Pak Phanang Health Center, Nakhon Si Thammarat, Thailand for routine care and follow-up.
Tools: All interviews were conduct at a health center and Thai language was used in a hard copy that divided into 3 parts including: 1) characteristics and personal health records, 2) applied TPB constructs in attitude toward behavior (AB), subjective norm (SN), perceived behavioral control (PBC) and behavioral intention (BI) and 3) diabetic self-care management behaviors (DSCM).
Conceptual Frame Work: A conceptual frame work of this study can be presented and applied with TPB construct by Ajzen, I. (1991) as in Fig. 1(3.6).
Statistical plan: Data were analyzed by using descriptive statistics, Pearson’s Product Moment Correlation, and multiple regression analysis.
The inclusion criteria were elderly diagnosed with T2DM attendance at a regional DM-clinic treatment for at least 1 year, with laboratory data confirming their state, despite being treated with anti-diabetic medication for at least 6 months. Participants were excluded if they were under the age of 55 or had incomplete data.
This present study was performed at Pak Panang Community Health Center in Pak Phanang subdistrict,Pak Phanang district of Nakhon Si Thammarat province. The health personal record was received under the Director Decision and ethical consideration.
This research project was conducted between December 2017 and September 2018.Data was collected after January 15, 2017 after the research project was approved by the Ethics Committee.Participants were selected to form specific criteria and enrolled in the study in February 2017. The data were collected until May 2017 then analyzed in July 2017, the final results were reported in September 2017.
The main method of conducting this research, participants will be interviewed about their diabetic self-management behavior, which was developed by the researcher according to The Theory of plan behavior. The results of the treatment and the blood test results were not conducted in this research, but were allowed to apply the treatment results from the main unit which has been regularly performed.
All participants were asked and interviewed by questionnaire with 3 parts including:
The focus group discussion in DSCM was collected and analyzed.
All patients were informed that their participation was voluntary and that they could withdraw from the research at any time (Human Ethics Research Committee, Nakhon Si Thammarat Provincial Health Office – EC: NSTPH 016/2017 by the date Jan 15, 2017).
All 450 elderly patients with type 2 diabetes have been attended at The west and The East community health centers in Pakpanung subdistrict, each center have a similar amount of T2DMaround 220 cases and alsosimilarin the cultural and lifestyle. The Pak Phanang community health center was selected by cluster area sampling and 212 patients were attended, then 134 samples were selected purposively by eligible criteria.
All data were analyzed using programmed calculations (SPSS version 11.5) where percentages, means, and standard deviations were used to describe all continuous variables. A Pearson correlation (r) was conducted to explain the association between variables. Pearson correlation coefficient (R) and Correlation of Determinants (R2) were analyzed to identify factors associated with diabetic self-care. Using the structural equation modeling technique, measurement and structural regression models were developed for both diabetic self-behavior intention and self-care behavior to predict glycemic control (Enter - MRA). All statistical analysis, p < 0.05 was considered significant for both models.
Population of 212 T2DM patients, who attended at the Pak Phanang Health Center, Nakhon Si Thammarat, Thailand. Participants of 134 elderly were purposive sampling at age over 55 years old, good oriented, be able to communicate, be able to read the Thai language, had the latest A1C result within 6 months.
Descriptive analysis in demographics
Medical data among the 134 (out of 152 cases - 88.16%) older subjects with T2DM showed: about 73.1% were female with average age of 64.8 (Standard Deviation - SD 9.41) years. Most were married 64.9% and have a life partner (53.0%) to care for and the other (minor) was divorced and had children and caregivers (32.1%). Most of them were non-working or housewives (49.7%), business owners (26.9%) and gardeners (20.9%) – respectively. 32.8% of the cases had incomes between 5,001-10,000 baht and 26.1% had incomes less than 5,000 baht – on a monthly basis, as shown in Table 1.
Table 1. Personal medical record in cases: percentage, mean and standard deviation
Characteristic | MinMean SD | Max | Case (n=134) | % |
Sex | ||||
Male | 36 | 26.9 | ||
Female | 98 | 73.1 | ||
Age ( years) | Min 26.0 | Max 91 | Mean 64.9 | SD 9.4 |
≤60 | 36 | 26.9 | ||
61–70 | 61 | 45.5 | ||
≥71 | 37 | 27.6 | ||
BMI | Min 17.3 | Max 53.0 | Mean 26.4 | SD 4.9 |
<18.5 | 1 | 0.7 | ||
18.5-24.9 | 31 | 23.1 | ||
25.0-29.9 | 25 | 18.7 | ||
>30.0 | 77 | 57.5 | ||
Duration Of DM(years) | Min1 | Max 30 | Mean 12.42 | SD 6.4 |
<5 | 20 | 14.9 | ||
5.1-10 | 43 | 32.1 | ||
11-15 | 34 | 25.4 | ||
>15 | 37 | 27.6 | ||
Comorbidity | ||||
None | 18 | 13.4 | ||
HT, CVD, or DLP | 116 | 86.6 | ||
Eye Check up | ||||
Non Diabetic retinopathy | 113 | 84.3 | ||
Diabetic retinopathy | 18 | 13.4 | ||
No check up | 3 | 2.2 | ||
Diabetic Foot | ||||
Normal | 116 | 86.6 | ||
Abnormal | 18 | 13.4 | ||
Follow up | ||||
Every1 month | 87 | 64.9 | ||
Every2 month | 31 | 23.1 | ||
Every3 month/miss | 16 | 11.9 | ||
Treatment Outcome | ||||
FBS≤125, BP≤139/89, A1C<7 | 45 | 33.6 | ||
FBS=126-154, BP=140/90-159/99, A1C <7 | 58 | 43.3 | ||
FBS=155-182, BP=160/100-179/109, A1C 7-8 | 18 | 13.4 | ||
FBS≥180, BP≥180/110-159/99, A1C >8 | 13 | 9.7 |
The results of treatment outcome with glucose controlled (A1C <7, about 77.9 %) and uncontrolled (A1C ≥7, about 21.1%). Diabetic Self–Care Management (DSCM) were divided into 3 classes (levels) – with mean score of medium level at 55.9% and minorities showing in both high level (DSCM > 41.94) at 20.89% and low (DSCM < 29.88) level at 23.14%.
Assessing the level of perceived control for this sample population,this research focused on the trend to care for T2DM in the long term by a term in grounded theory. This findings showed that most participants had higher proportion in exercise and anti-diabetic adherence than eating healthy food. In opposite way the results was shown the stronger intention in their healthy food than exercise and medicinal adherence. Therefore, the descriptive results about perceived behavioral control (PBC) and behavioral intention (BI) mean score in all items were determined to be a high level, as shown in table 2.
Table 2. Perceived Behavioral Control and Behavioral Intention in cases; percentage, min, max, mean and standard deviation. (n=134)
Variables | Cases | % | ||
Perceived Behavioral Control | ||||
Plan a healthy diet | Min1 | Max 3 | Mean 2.28 | SD 0.75 |
Disagree | 24 | 17.9 | ||
Not sure | 49 | 36.6 | ||
Agree | 61 | 45.5 | ||
Plan to join exercise | Min2 | Max 3 | Mean 2.97 | SD 0.17 |
Not sure | 4 | 3.0 | ||
Agree | 130 | 97.0 | ||
Plan to perform medical adherence | Min2 | Max 3 | Mean 2.90 | SD 0.31 |
Not sure | 14 | 10.4 | ||
Agree | 120 | 89.6 | ||
Behavioral Intention | ||||
Intend to eat healthy diet | Min2 | Max 3 | Mean 2.78 | SD 0.41 |
Not sure | 29 | 21.6 | ||
Agree | 105 | 78.4 | ||
Intend to join exercise | Min1 | Max 3 | Mean 2.48 | SD 0.59 |
Disagree | 6 | 4.5 | ||
Not sure | 58 | 43.3 | ||
Agree | 70 | 52.2 | ||
Intend to perform medicaladherence | Min1 | Max 3 | Mean 2.48 | SD 0.59 |
Disagree | 6 | 4.5 | ||
Not sure | 58 | 43.3 | ||
Agree | 70 | 52.2 |
The Correlation Between Variables
The Pearson correlation coefficient (r) indicated that all TPB constructs were significantly correlated with intentions. Pearson Correlation (r) analysis of paired variables showed strong correlation: Attitude-Perceive Control (r .327, p < .01) Self-care Management-Perceive Control (r .317, p < .01), Subjective Norm-Perceive Control (r .275, p < .01). Behavioral Intention - Perceive Control (r .198, p < .05), respectively. Perceived Control was significantly correlated with other variables, prominently (Table 3).
Table 3. Pearson Correlation(r) n =134
Notes: **, * Correlation is significant at the 0.01 level, and 0.05 level (2-tailed) respectively. |
Results of the preliminary test of the regression equation
The 6 assumptions were tested according to the preliminary agreement to MRA with the following results
Therefore, the assumptions were correctly tested and can explain the dependent variables (BI and DSCM) and their significance and strength in influence on self-care management.
This study analyzed and provided both models through multiple regression analysis, with multicollinearity not influencing the elementary agreement.
Multiple correlation coefficient (R) - Coefficient of determination (R٢)
In deriving the predictive equation through regression methods, calculations of the coefficient of correlation for behavioral Intention (R .301, F 4.325, P < .01, Table 4) were determined and the coefficient of determination (R٢ -.091), which is the variance of the explanatory variables in describing the variables, considering the weight and power of the predictor in the multiple regression equations.
A predictive equation for DSCMwas also derived by PBC and BI, with the results of (R .332, F 8.099, P < .001, Table 4) and the coefficient of determination (R٢ .110). PBC was found to be a stronger predictor for SCM than BI - as shown in table 4. However, the regression models were weak predictors for this research, overall.
Table 4. The Coefficients of Multiple Correlation (R) for Behavioral Intention (BI), and Diabetic self-care management (DSCM)
The coefficient of determination (R2 -.091, P < .01) in the multiple regression equation predicts BI. The results show that AB, SN, and PBC variables shape Behavioral Intentions about 9.10%; BI, PBC and the coefficient of determination (R2 .110, P < 001). The 2 variables together describe DSCM about 11%.
Results demonstrated that AB, SN, and PBC can explain the variables forBI (Model 1) as AB (β -0.91, 4.20%), SN (β .490, 25.3%) and PBC (β .229, 14.2%) – respectively. The variable SN with standardized coefficients (β.253, P<.01) is a significant predictor for BI (Table 5). It is important to note that there was at least one variable that can be described in terms of the variance of the variables. Also, the same results were found for the Coefficient of determination (R٢) for DSCM. PBC (Model 2) had explained outcomes about 29.7% (β.297, p<.001) BI about 10.0 %. Therefore, PBC is a significant predictor for explaining DSCM (Table 5).
Table 5. Predicting Behavioral Intention (BI), and Diabetic self-care management (DSCM)
Variable | F (t) | B7 | Standard E of β8 | P-value | R5 | Adjusted R26 | 95.0% CI9 for B |
Model 1 | 4.325 | 4.025 | 1.34815 | .006 | .301a | .070 | .762 - 7.287 |
AB1 | (-.413) | -.091 | -.042 | .680 | (-.528 -.345) | ||
SN2 | (2.559) | .490 | .253 | .012 | (.111 -.868) | ||
PBC3 | (1.592) | .229 | .142 | .114 | (-.056 -.513) | ||
Model 2 | 8.099 | 15.258 | 6.04422 | .000 | .332a | .096 | 4.736 - 25.780 |
PBC | (3.531) | 2.174 | .297 | .001 | (.956 - 3.392) | ||
BI4 | (1.195) | .457 | .100 | .234 | (-.300 - 1.214) | ||
a. Predictors: (Constant), PBC, SN, AB, BI | b. Dependent Variable: BI (Model 1) | ||||||
a. Predictors: (Constant), BI, PBC | b. Dependent Variable: DSCM (Model 2) | ||||||
1AB Attitude toward behavior, 2SN Subjective Norm, 3PBC Perceived Behavioral Control, 4BI Behavioral Intention ,5R Multiple correlation coefficient (R) - correlation 6R2 Coefficient of determination (R2), 7 - 8beta coefficient (β), 9statistically significant when 95% Confidence Interval does not include 0 |
Results revealed that attitude (AB), subjective norm (SN), and perceived behavioral control (PBC) involved predicted behavioral intentions in diabetic patients (Model 1) - focusing on emphasizing dietary control, exercise and medicinal adherence. The regression equation is as follows:
Model 1 (BI): 4.025-AB 0.091+SN 0.490+PBC 0.229
Model 2 demonstrated that Perceived Behavioral Control is a significantly strong predictor for predicting DSCM behaviors and is expressed as follows:
Model 2 (DSCM): 15.258+ PBC 2.174+ BI 0.457
The Framework and the findings in Diabetic Self Care Management (DSCM) from this research is better understood by Fig. 1.
Most of (76.9%) of participants in this present study had controlled the glycemic outcome (A1C < 7) and few of them (22.1%) had uncontrolled.
The proportion in PBC means score in «plan for healthy diet» showed: disagree-1, not sure -2, and agree-3, which differs from the exercise and drug adherence that was similar proportion; not sure -1, and agree-9.
The proportion in BI means score in «intend for healthy diet» showed: not sure -2, and agree-8, which differs from the exercise and drug adherence that was similar proportion; disagree-1, not sure -4, and agree-5.
Subjective Norm (SN) was highly significantly correlated with all variables. All predictors: AB, SN, PBC had the influence to predict intention to perform diabetic self-management behaviors and PBC was a highest significant predictor in both regression model.
This result is a direct benefit for elderly patients with diabetes in encouraging awareness and compliance with structured behavioral patterns in order to be able to plan behaviors based on perceived problems, obstacles, and benefits in controlling type 2 diabetes.
In addition, this research has resulted in the development of care systems for diabetics and health service providers in evaluating and improving better T2DM self-management outcomes.
The reatment outcome among elderly with T2DM
The entire elderly population living with diabetes was stable in distribution for at least 5 years and sustained for more than 15 years and is similar with previous studies [
This study compared treatment outcome among controlled and uncontrolled patients at a 4:1 ratio, respectively, consequently with the percentages of controlled slightly improved among men (45.9%) but not among women (36.4%) [
Discussion of the primary research results
Results demonstrated that more than half the patients can manage their T2DM health without complications, but the remainder had high risk in renal insufficiency and/or cerebral-vascular disease – serious complications and the main cause of diabetic death. This finding is similar with others where the participants (36%) suffer from diabetes complications especially neuropathy [
The correlation between TPB variables
Predicted variables by attitude, key reference group(s) encourages planned behavioral control, appropriately. As a result, the behavior of the patients was found to be relatively high. This means that patients with similar demographic features have a similar life context, having attitudes and behaviors in like direction. In fact, there is little variance in the sample group as is noted.
Behavioral control correlates with other variables quite clearly. Consistent with the other TPB constructs [
Factors related with diabetic self-care management
This research explored the explanatory variables focused on diabetic self-care management;
Dietary pattern, Physical Activities, and Drug Adherence: T2DM being a chronic disorder requires multiple therapeutic approaches including;
Dietary pattern; the patient intention to manage behavior according to doctor’s advice, reduced blood sugar levels and complications [
Physical Activities; results indicate that TPB-based interventions including planning strategies may encourage physical activity among older people with diabetes. The previous study shows nearly 87% of the variance in exercise behavior and 72% of the variance in healthy eating behavior were explainable by TPB constructs [13,14]. The frequency of regular exercise benefits patients, reducing long term monitored blood glucose. The types of exercise should be easy and appropriate for elderly - such as arm swing exercises, small and short step exercises and mild to moderate intensity aerobic exercises [16,19].
Drug adherence; in contrast to other studies, the results obtained reported high medication adherence over previous studies showing the rate of non-adherence to the treatment prescription being high as compared to other studies that reported moderate adherence levels to medication [17,20]. Similar findings reported that suboptimal medication adherence leads to negative consequences, such as suboptimal metabolic control, increased risk of diabetes complications and hospitalizations, and additional healthcare expenditures [
This result consequently clarified that all 3 variables predisposed patient intention to accomplish their self-care behaviors. To apply the model to changing lifestyle requires focus on perceived behavioral control, and subjective norms - such as planning to prepare healthy food, planning to exercise or perform slight physical activities (even activities in gardens or housing keeping help), monitoring and empowerment regularly from family members and doctors or other paramedical practitioners. Being concerned with self-management support may improve self-care activities and A1C in patients with comorbid diabetes [
Social support (Subjective norms)
The referral groups such as medical providers, nurses, and caregivers (subjective norms) were the main predictors of behavioral intention in model 1. The results showed that life partners or children make patients trust and adhere to all those behaviors, resulting in good behavioral outcomes. Results in medical treatment planning, such as annual year biochemistry and eye checkup, were strong, only 2.2 % were missed. Most patients had healthy eating patterns, physical activity and drug adherence consistent with doctor-patient relationships. Concern by family members support strong predicting with behavioral control for patients to perform self-care behaviors (these variables explained 30% of the common variance). It is important for the health care provider to assess sources of social support and integrate the results of this assessment to ensure the empowerment of the patient during diabetes education [18,20].
The study data presented obstacles such as poor ability of elderly patients and lack of education by caregivers to conduct accurate glucose monitoring. Unrealistic perceptions of cost of test strips and needles, lack of basic knowledge due to the absence of diabetes self-management educational programs [15,20]. Not only absence of education but also lack in skills to perform, lack of home health care provider support, fear of testing and associated pain and preference for traditional medicine overwhelmed the sample group [
Perceived Behavioral Control and Behavioral Intention
Increased ability to adjust medication dosages, dietary intake and physical activity depended on Perceived Control and Intention. The patients who have strong intention to perform self-care behaviors should have a good plan and appropriate time to change their life style. The significance of both predictors serves greatly to control blood sugar. The perception of behavior control has a distinct effect followed by conclusions, in accordance to the reference group [
Diet behavior was most important for determining lifestyle modification. Therefore, most patients should have learned by peer group discussion through role models who have well-controlled and uncontrolled storylines about the health benefits or multiple complications.
The limitation in tested outcome that the A1C can diagnose pre-diabetes and diabetes, but is not recommended for screening for diabetes in Thai people due to high costs And standardized laboratories Certified by NGSP and standards are still low when referring to the Diabetes Control and Complications Trial reference assay (DCCT) [
The results showed patients had a moderate level in self-care management, consistent with their treatment outcome. It was noted that behavioral intention and diabetic self-management behaviors are being managed by perceived behavioral control for predicting model and health-related diabetic self-care management. The perceived control is a helpful guide in managing obstacles in long-term health benefits among elderly T2DM. Further research to identify the barriers with self-efficacy, possibly through building patient empowerment skills and a deeper perceived quality of life among populations with diabetes, should be undertaken.
Source of funding. Pak Phanang Municipality Health Fund, NHSO.
Conflict of interests. Authors declare no explicit and potential conflicts of interests associated with the publication of this article.
Acknowledgements. Thanks to the Provincial Public Health Doctor at Nakhon Si Thamarat Province, and the public health teams at Pak Phanang Community Health Centre who provided assistance and support in collecting medical records and tracking volunteers. Thank you very much to Pak Phanang Municipality Health Fund, NHSO for budget into the research project, and thank you to all participants, for your kindness and cooperation. “This research was partially supported by the new strategic research (P2P) project, Walailak University, Thailand.”
The authors declare that there are no conflicts of interest present.