Prevalence and predictors of suboptimal glycemic control among patients with type 2 diabetes mellitus in northern Thailand: A hospital-based cross-sectional control study.

BACKGROUND: Suboptimal glycemic control among patients with type 2 diabetes mellitus (DM) is a significant public health problem, particularly among people living with poor education and economic statuses, including those with a unique dietary culture. This study aimed to estimate the prevalence and identify the factors associated with suboptimal glycemic control among patients with type 2 DM during the coronavirus disease-2019 (COVID-19) pandemic.
METHODS: A hospital-based cross-sectional study was used to elicit information from DM patients attending six hospitals located in Chiang Rai Province, northern Thailand, between February and May 2021. A validated questionnaire and 5 mL blood specimens were used as the research tools. Glycated hemoglobin (HbA1c) greater than 7.0% among DM patients at least two years after diagnosis was defined as suboptimal glycemic control. Chi-square tests and logistic regression were used to identify the associations between variables at the significance level α = 0.05.
RESULTS: A total of 967 patients were recruited for this study; 54.8% 530 had suboptimal glycemic control, 58.8% were female, 66.5% were aged 50-69 years, and 78.5% were married (78.5%). Six variables were found to be associated with suboptimal glycemic control in multivariable logistic regression. Participants aged <49, 50-59, and 60-69 years had 3.32 times (95% CI = 1.99-5.53), 2.61 times (95% CI = 1.67-4.08), and 1.93 times (95% CI = 1.26-2.95) greater odds of having suboptimal glycemic control, respectively, than those aged ≥70 years. Married individuals had 1.64 times (95% CI = 1.11-2.41) greater odds of having suboptimal glycemic control than those ever married. Participants who consumed sticky rice had 1.61 times (95% CI = 1.19-2.61) greater odds of having suboptimal glycemic control than those who did not consume sticky rice in daily life. Participants who had been diagnosed with DM for 11-20 years and ≥21 years had 1.98 times (95% CI = 1.37-2.86) and 2.46 times (1.50-4.04) greater odds of having suboptimal glycemic control, respectively, than those who had been diagnosed ≤ 10 years. Participants who had experienced forgetting to take their medication had 2.10 times (95% CI = 1.43-3.09) greater odds of having suboptimal glycemic control than those who did not, and those who had their medical expenses covered by the national scheme had 2.67 times (95% CI = 1.00-7.08) greater odds of suboptimal glycemic control than those who self-paid.
CONCLUSION: Effective health interventions to control blood glucose among DM patients during ongoing treatment are urgently required. The interventions should focus on patients aged less than 69 years, marital status, forgetting to take their medication, and a longer time since diagnosis, including reducing their sticky rice consumption. The effects of copayments should also be considered.

Introduction

Diabetes mellitus (DM) with suboptimal glycemic control is one of the major health problems among people around the world, especially those who are living in low- and middle-income countries, including Thailand [1]. The World Health Organization (WHO) estimated that 422 million patients globally are affected by DM [2] and 1.5 million deaths were directly caused by DM, and an additional 3 million deaths were secondarily caused by DM in 2019, primarily from suboptimal glycemic control [1]. In addition to the lives lost from the disease, it also adversely impacts economics, especially due to the resources required for the treatment and care of DM patients for many years [3, 4]. Moreover, the quality of life of those who have this condition and their family members are also reduced due to the need for constant care during everyday life [5, 6]. A significant proportion of DM patients are properly diagnosed and well cared for, but the effectiveness of maintaining or controlling their blood glucose is very poor, particularly in Thailand [7].

Thailand is defined as an upper-middle-income country with approximately 67 million people [8]. The WHO reported that the prevalence of DM among the Thai population aged 30 years and over was 9.6% (9.1% for men and 10.1% for women [9]. Thailand has a policy and strategic plan for reducing diabetes problems, but it has no programs to address overweight and obesity, including physical inactivity problems [9]. The majority of Thai people work in the agricultural sector, including people in northern Thailand [10]. People living in northern Thailand have some special cultures and lifestyles, including unique cooking practices. Sticky rice is the main dish in daily life, and consuming oily noodles for lunch is also common [11]. Since 2018, the northern region has been classified as having the greatest proportion of people aged ≤60 years in the country, namely an aging society [12], while its economic level has been ranked at a lower level than the central and southern regions [13].

In 2020, the Ministry of Public Health, Thailand estimated that 5 million people have developed DM among the Thai population or one in every eleven among people aged 15 years and over [14]. People living in the northern region of Thailand, where the local people have special cultures and lifestyles, including the consumption of sticky rice and popular oily dishes in their daily lives [15], have been reported to have one of the highest DM prevalences in Thailand [16]. Moreover, 200 deaths every day have been attributed to DM [14]. Aekplakorn et al. [7] reported that the prevalence of DM among the Thai population was 10.8% among individuals aged 20 years and over, and the overall suboptimal glycemic control was 29.5% (23.9% among men and 35.7% among women). DM patients with poor economic and educational statuses have a higher risk of having suboptimal glycemic control. In addition, a large amount of money has been allocated for DM case management in Thailand [17], particularly to those who have a problem of suboptimal glycemic control while living with a unique culture. A few scientific studies are available to understand the prevalence and factors contributing to suboptimal glycemic control among DM patients in northern Thailand, especially during the coronavirus disease-2019 (COVID-19) crisis, when the schedules of health care services were modified to fit the situation in hospitals.

The aims of this study were to estimate the prevalence of suboptimal glycemic control and identify factors associated with suboptimal glycemic control among Thai people living in northern Thailand.

Materials and methods

Study design and study setting

A cross-sectional study was used to collect data from participants who were DM patients attending six hospitals: Mae Lao Hospital, Mae Chan Hospital, Wiang Chiang Rung Hospital, Phan Hospital, Wiang Chai Health Promoting Hospital, and Mae Lao Health Promoting Hospital, which were selected by a random method from among 18 hospitals in Chiang Rai Province, Thailand.

Study population and eligible population

DM patients who attended DM clinics in the six hospitals were the study population. Those who attended a clinic between February and May 2021 and had been diagnosed at least two years prior with DM and treated met the inclusion criteria. However, those who had severe illness, were admitted to the inpatient department, pregnant, or could not provide essential information were excluded from the study.

Sample size

The sample size was calculated based on the standard formula for a cross-sectional design [18]; n = [Z2α/2P (1-P)]/d2, where the Z = value from the standard normal distribution corresponded to the desired confidence level (Z = 1.96 for 95% CI), P = the expected true proportion (P = 0.30 [7], and d = precision (d = 0.03); after adding 5% to account for any error in the study, 941 participants were needed for the analysis.

Research instruments

A questionnaire developed by the researcher was used for data collection. It consisted of six parts. In part one, seven items were used to collect physical examination and laboratory data, such as weight, height, blood pressure, triglyceride level, low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C). In part two, twenty questions were used to collect general information, such as sex, age, religion, tribe, and education. In part three, eight questions were used to collect data regarding health behaviors such as smoking, alcohol use, having tea, having coffee, etc. In part four, five questions about the stress test (ST-5) [19] were used to detect the level of stress. In part five, ten questions were used to detect knowledge about DM prevention and control. In the last part, ten questions were used to detect attitudes toward DM prevention and control (S1 Questionnaire).

A five mL blood specimen was drawn to detect HbA1c levels and other lipid profiles, such as LDL-C, HDL-C, and triglycerides.

Validated questionnaire

The validity and reliability of the questionnaire were detected by different methods. Item-objective congruence was used to detect the validity of the questionnaire. Using this method, three external experts assessed the congruence between the questions and the context of the study, including the objectives of the study. The experts provided the score for each item: “-1” means that the question is not related to the content and objectives of the study; “0” means that the question is related to the content and objectives of the study but requires improvement before use; and “+1” means that the question is relevant to the content and the objectives of the study and does not require any improvement. The scores from the experts were pooled and divided by three before interpretation. If the questions had an average score less than 0.5, the questions were deleted from the questionnaire. If the questions were scored between 0.5-0.7, they were improved before being added to the questionnaire. The items with a score greater than 0.7 were included in the final questionnaire.

The questionnaire was piloted with 20 people who had similar characteristics to the study population and attended the DM clinic at Mae Sai Hospital. In this step, the feasibility, proper words or sentences used, and ordering of the questions, including the reliability, were analyzed. Finally, Cronbach’s alpha was found to be 0.72 for the final questionnaire.

Measures

Body weight index (BMI) is a standard used by the WHO that is classified into three categories: underweight (≤18.49), normal weight (18.50-24.99), and overweight (≥25.00) [20]. Stress was classified into three categories [21]: low (≤4 scores), moderate (5-7 scores), and high (≥8 scores). The triglycerides were classified into two major groups according to guidelines from the World Health Organization [22]: normal (<150 mg/dL) and high (≥150 mg/dL). According to the World Health Organization (WHO), low-density lipoprotein cholesterol (LDL-C) levels were classified into two main groups: normal (<100 mg/dL) and high (≥100 mg/dL) [22]. High-density lipoprotein cholesterol (HDL-C) was classified into two groups: low (<40 mg/dL) and normal (≥40 mg/dL) [22]. A glycated hemoglobin (HbA1c) level ≥ 7 was defined as suboptimal glycemic control [23].

Data gathering procedures

The hospital directors and the chiefs of DM clinics were contacted and we explained the study objectives, including the procedures of data collection, after obtaining ethical approval for conducting this research from the Chiang Rai Provincial Public Health Office. On the day of data collection, all DM patients who attended the clinic were invited to join the study on a voluntary basis. Those who were willing to participate in the study were given an explanation of the study objectives and the procedure of data and blood specimen collection. After signing the written consent form, the participants were asked to fill out the questionnaire. For those who could not use Thai, the participants were requested to use fingerprints to sign the consent form. Researchers completed the questionnaire for those who could not use Thai after obtaining the information from the participants. A 5 mL blood sample was collected by medical technicians who had validated licensing. The participating DM patients were asked to fast (nothing per oral (NPO)) for 12 hrs before blood collection. All specimens were stored properly and transferred to the Mae Fah Laung Medical Laboratory Center for analysis on the same day.

Statistical analysis

The data were coded in an Excel sheet followed by fact-checking before transfer into SPSS with the SPS program (version 24, Chicago, IL). Descriptive statistics were used to describe the general characteristics of the participants. Percentages are used to describe the categorical data, while other continuous data are described by the mean and standard deviation (SD) for a normal distribution, and the median and interquartile range (IQR) for the skewed data. Chi-square and Fisher exact tests were used to detect the associations. Logistic regression was used to detect the associations between the independent variables and dependent variable at a significance of α = 0.05. Selection variable into the model, the “Enter” mode was used. The Cox-Snell R2and Nagelkerke R2and the Hosmer-Lemshow were used to determine the fit of the model in all steps. These variables found to be significant in the univariable logistic model were considered to be put into the multivariable model. Before fitting the final multivariable logistic model, some variables were controlled as the confounder factors before interpretation.

Ethics approval and consent to participate

Consent to participate, all study instruments and procedures were approved by the Ethics Committee for Human Research, Chiang Rai Provincial Public Health Office, Chiang Rai, Thailand (CRPHO No.6/2465). All participants received an oral and written explanation and provided their consent before a voluntary agreement was witnessed and documented by signature or fingerprint.

Results

General characteristics of the participants

A total of 967 cases (530 suboptimal glycemic control (54.8%) and 437 controlled blood glucose (45.2%)) were recruited from 6 hospitals: 294 cases (30.4%) from Phan Hospital, 301 cases (31.1%) from Mae Chan Hospital, 208 cases (21.5%) from Mae Lao Hospital, 84 cases (8.7%) from Wiang Chiang Rung Hospital, 40 cases (4.1%) from Mae Lao Health Promoting Hospital, and 40 cases (4.1%) from Wiang Chai Health Promoting Hospital.

More than half were women (58.8%), 66.5% were aged 50-69 years (mean = 58.7, min = 18, max = 97, and SD = 11.3), and 99.1% held Thai identification cards (IDs). The majority were married (78.5%), Buddhist (99.5%), graduated from primary school (67.1%), worked as farmers (36.4%), and had an annual family income less than 50,000 baht (68.7%), with a median of 40,000 baht and an IQR of 30,000 baht (Table 1).

Table 1

General characteristics of participants.

Characteristics	Total		Suboptimal glycemic control				χ2	p-value
			Yes		NO
	n	%	n	%	n	%
Total	967	100.0	530	54.8	437	45.2	N/A	A/A
Sex
Male	398	41.2	200	50.3	198	49.7	5.67	0.017*
Female	569	58.8	330	58.0	239	42.0
Age (years)
≤49	187	19.3	66	35.3	121	64.7	27.49	<0.001*
50-59	274	28.3	163	59.5	111	40.5
60-69	369	38.2	195	52.8	174	47.2
≥70	137	14.2	51	37.2	86	62.8
Tribe
Hill tribe	28	2.9	18	64.3	10	35.7	1.04	0.307
Thai people	939	97.1	512	54.5	427	45.5
Thai ID card
No	9	0.9	7	77.8	2	22.2	1.93	0.164
Yes	958	99.1	523	54.6	435	45.4
Marital status
Single	70	7.2	43	61.4	27	38.6	8.03	0.018*
Married	759	78.5	426	56.1	333	43.9
Ever married	138	14.3	61	44.2	77	55.8
Religion
Buddhist	962	99.5	528	54.9	434	45.1	0.44	0.505a
Christian or Muslim	5	0.5	2	40.0	3	60.0
Education
No education	122	12.6	58	47.5	64	52.5	4.11	0.128
Primary school	649	67.1	356	54.9	293	45.1
Secondary school and higher	196	20.3	116	59.2	80	40.8
Occupation
Unemployed	283	29.3	147	51.9	136	48.1	3.41	0.331
Famer	352	36.4	188	53.4	164	46.6
Employed	224	23.2	130	58.0	94	42.0
Trade and government officer	108	11.2	65	60.2	43	39.8
Annual income (baht)
≤ 50,000	664	68.7	366	55.1	298	44.9	1.67	0.433
50,001-100,000	189	19.5	97	51.3	92	48.7
≥100,001	114	11.8	67	58.8	47	41.2
Family debt
No	566	58.5	334	59.0	232	41.0	9.72	0.002*
Yes	401	41.5	196	48.9	205	51.1
Family members (people)
≤ 4	828	85.6	444	53.6	384	46.4	3.26	0.071
≥ 5	139	14.4	86	61.9	53	38.1
Living with
Alone	61	6.3	28	45.9	33	54.1	2.383	0.497
Spouse	589	60.9	330	56.0	259	44.0
Child	216	22.3	118	54.6	98	45.4
Relatives	101	10.4	54	53.5	47	46.5

*Significance level α = 0.05

a Fisher’s exact test

Four factors were significantly different between the uncontrolled and controlled blood glucose groups: sex (p-value = 0.017), age (p-value<0.001), marital status (p-value = 0.018), and family debt (p-value = 0.002) (Table 1).

Health behaviors, knowledge and attitudes toward DM prevention and control

Almost one-fifth smoked (16.2%), 15.9% used alcohol, and more than half did not exercise in their daily life. A large proportion cooked their own food (88.4%), 31.9% regularly ate sticky rice, 15.2% drank tea, and 31.7% drank coffee. One-fourth (25.6%) had stress between moderate-to-high levels, 47.8% had knowledge regarding DM prevention and control at low-to-moderate levels, and 86% had attitudes toward DM prevention and control at poor-to-moderate levels (Table 2).

Table 2

Substance use and health behaviors among the participants.

Health behaviors	Total		Suboptimal glycemic control				χ2	p-value
			Yes		NO
	n	%	n	%	n	%
Smoking
No	810	83.8	423	52.2	387	47.8	13.47	<0.001*
Yes	157	16.2	107	68.2	50	31.8
Alcohol use
No	813	84.1	441	54.2	372	45.8	0.65	0.417
Yes	154	15.9	89	57.8	65	42.2
Exercise
No	499	51.6	198	39.7	301	60.3	95.42	<0.001*
Sometime	265	27.4	190	71.7	75	28.3
Regular	203	21.0	142	70.0	61	30.0
Preparing food
Themselves	855	88.4	457	53.5	398	46.5	5.49	0.019*
Buying	112	11.6	73	65.2	39	34.8
Type of rice eaten daily
Non-sticky rice	659	68.1	332	50.4	327	49.6	16.38	<0.001*
Sticky rice	308	31.9	198	64.3	110	35.7
Frequency of having sticky rice per day
One	16	5.2	12	75.0	4	25.0	2.75	0.252
Two	22	7.1	17	77.3	5	22.7
Three	270	87.7	169	62.6	101	37.4
Drinking tea
No	820	84.8	425	51.8	395	48.2	19.33	<0.001*
Yes	147	15.2	105	71.4	42	28.6
Drinking coffee
No	660	68.3	323	48.9	337	51.1	28.91	<0.001*
Yes	307	31.7	207	67.4	100	32.6
Stress (ST-5)
Low	719	74.4	390	54.2	329	45.8	0.36	0.832
Moderate	159	16.4	90	56.6	69	43.4
High	89	9.2	50	56.2	39	43.8
Knowledge regarding DM prevention and control
Low	78	8.1	37	47.4	41	52.6	4.50	0.105
Moderate	384	39.7	201	52.3	183	47.7
High	505	52.2	292	57.8	213	42.2
Attitudes toward DM prevention and control
Poor	270	27.9	122	45.2	148	54.8	14.00	0.001*
Moderate	562	58.1	329	58.5	233	41.5
Positive	135	14.0	79	58.5	56	41.5

*Significance level α = 0.05

Seven variables were found to be different health behaviors between the uncontrolled and controlled blood glucose groups: smoking (p-value<0.001, preparing food (p-value = 0.019), type of rice eaten daily (p-value<0.001), drinking tea (p-value<0.001), drinking coffee (p-value<0.001), and attitude toward DM prevention and control (p-value = 0.001) (Table 2).

Experiences related to DM treatment and care and biomarkers

Among the participants, 26.7% were diagnosed with DM more than 10 years prior, 2.1% were self-paying for all medical expenses, 19.0% had experienced forgetting to take a medication, and 13.5% had experienced side effects while using a medication related to DM treatment. Sixty people had experienced wounds on their feet, 50.3% had diabetic nephropathy, and 53.5% had HT. A large proportion were overweight (65.3%), 44.8% had high LDL cholesterol levels, and 49.7% had high triglyceride levels (Table 3).

Table 3

DM experiences and biomarkers of the participants.

Health indicators	Total		Suboptimal glycemic control				χ2	p-value
			Yes		No
	n	%	n	%	n	%
Length of DM diagnosed (year)
≤ 10	709	73.3	362	51.1	347	48.9	15.16	0.001*
11-20	172	17.8	113	65.7	59	34.3
> 20	86	8.9	55	64.0	31	36.0
Medical expenses
Covered by the national universal scheme	947	97.9	519	54.8	428	45.2	0.00	0.986
Self-paid	20	2.1	11	55.0	9	45.0
Experience of forgetting to take a medication
No	783	81.0	398	50.8	385	49.2	26.29	<0.001*
Yes	184	19.0	132	71.7	52	28.3
Having side effects from DM medications
No	836	86.5	434	51.9	402	48.1	20.87	<0.001*
Yes	131	13.5	96	73.3	35	26.7
History of wounds on foot
No	907	93.8	482	53.1	425	46.9	16.39	<0.001*
Yes	60	6.2	48	80.0	12	20.0
Diabetic nephropathy
No	375	38.8	191	50.9	184	49.1	14.57	0.001*
Yes	486	50.3	263	54.1	223	45.9
Do not know	106	11.0	76	71.7	30	28.3
Having HT
No	395	40.8	225	57.0	170	43.0	14.29	0.001*
Yes	517	53.5	263	50.9	254	49.1
Do not know	55	5.7	42	76.4	13	23.6
BMI (kg/m2)
Underweight (≤ 18.50)	53	5.5	27	50.9	26	49.1	14.03	0.001*
Normal weight (18.51-22.99)	283	29.3	130	45.9	153	54.1
Overweight (≥ 23.00)	631	65.3	373	59.1	258	40.9
HDL cholesterol (mg/dL)
Low (< 40)	262	27.1	144	55.0	118	45.0	0.00	0.953
Normal (≥ 40)	705	72.9	386	54.8	319	45.2
LDL cholesterol (mg/dL)
Normal (<100)	534	55.2	272	50.9	262	49.1	7.22	0.007*
High (≥100)	433	44.8	258	59.6	175	40.4
Triglycerides (mg/dL)
Normal (< 150)	486	50.3	253	52.1	233	47.9	2.98	0.084
High (≥ 150)	481	49.7	277	57.6	204	42.4

*Significance level α = 0.05

Nine (9) variables were found to be significantly different between the suboptimal glycemic control and the controlled blood glucose groups: the duration since the DM diagnosis (p-value = 0.001), experience of forgetting to take a medication (p-value<0.001), having side effects from DM medications (p-value<0.001), having wounds on the foot (p-value<0.001), diabetic nephropathy (p-value = 0.001), having HT (p-value = 0.001), BMI (p-value = 0.001), and LDL cholesterol (p-value = 0.007) (Table 3).

Prevalence and factors associated with suboptimal glycemic control

The prevalence of suboptimal glycemic control was 54.8% (50.3% in men and 58.0% in women). The highest prevalence of suboptimal glycemic control was among those aged 50-59 (59.5%) and 60-69 years (58.0%).

In the model to identify socio-demographics that associated with suboptimal glycemic control among DM patients by univariable logistic regressions, five (5) variables were found to be associated with suboptimal glycemic control: sex, age, marital status, education, and family debt (Table 4).

Table 4

Identifying socio-demographics that associated with suboptimal glycemic control among DM patients by univariable and multivariable logistic regressions.

Factors	Suboptimal glycemic control		Univariable analysis			Multivariable analysis
	Yes (%)	No (%)	OR	95% CI	p-value	AOR	95% CI	p-value
Total	530 (54.8)	437 (45.2)	N/A	N/A	N/A	N/A	N/A	N/A
Sex
Male	200 (50.3)	198 (49.7)	1.00
Female	330 (58.0)	239 (42.0)	1.39	1.07-1.79	0.012*
Age (years)
≤49	66 (35.3)	121 (64.7)	3.09	1.95-4.88	<0.001*	3.32	1.99-5.53	<0.001*
50-59	163 (59.5)	111 (40.5)	2.47	1.62-3.77	<0.001*	2.61	1.67-4.08	<0.001*
60-69	195 (52.8)	174 (47.2)	1.89	1.26-2.82	0.002*	1.93	1.26-2.95	0.011*
≥70	51 (37.2)	86 (62.8)	1.00			1.00
Marital status
Single	43 (61.4)	27 (38.6)	2.07	1.15-3.72	0.015*	1.78	0.95-3.31	0.070
Married	426 (56.1)	333 (43.9)	1.67	1.16-2.41	0.006*	1.64	1.11-2.41	0.011*
Ever married	61 (44.2)	77 (55.8)	1.00			1.00
Tribe
Hill tribe	18 (64.3)	10 (35.7)	1.50	0.68-3.28	0.310
Thai people	512 (54.5)	427 (45.5)	1.00
Thai ID card
No	7 (77.8)	2 (22.2)	2.91	0.60-14.08	0.184
Yes	523 (54.6)	435 (45.4)	1.00
Religion
Buddhist	528 (54.9)	434 (45.1)	1.82	0.30-10.97	0.511
Christian or Muslim	2 (40.0)	3 (60.0)	1.00
Education
No education	58 (47.5)	64 (52.5)	1.00
Primary school	356 (54.9)	293 (45.1)	1.33	0.90-1.96	0.147
Secondary school and higher	116 (59.2)	80 (40.8)	1.63	1.03-2.57	0.035*
Occupation
Unemployed	147 (51.9)	136 (48.1)	1.00
Agriculturist	188 (53.4)	164 (46.6)	1.10	0.80-1.51	0.538
Employed	130 (58.0)	94 (42.0)	1.31	0.92-1.87	0.127
Trade and government officer	65 (60.2)	43 (39.8)	1.49	0.95-2.34	0.081
Annual income (baht)
≤ 50,000	366 (55.1)	298 (44.9)	1.00
50,001-100,000	97 (51.3)	92 (48.7)	0.88	0.64-1.22	0.470
≥100,001	67 (58.8)	47 (41.2)	1.21	0.81-1.82	0.338
Family debt
No	334 (59.0)	232 (41.0)	1.53	1.18-1.98	0.001*
Yes	196 (48.9)	205 (51.1)	1.00
Family members (people)
≤ 4	444 (53.6)	384 (46.4)	1.00
≥ 5	86 (61.9)	53 (38.1)	1.40	0.97-2.02	0.071
Living with
Alone	28 (45.9)	33 (54.1)	0.69	0.36-1.30	0.257
Spouse	330(56.0)	259 (44.0)	1.11	0.73-1.70	0.610
Child	118 (54.6)	98 (45.4)	1.04	0.65-1.68	0.846
Relatives	54 (53.5)	47 (46.5)	1.00

*Significance level α = 0.05

Only two (2) variables were found to be associated with suboptimal glycemic control in multivariable logistic regression: age, and marital status. Participants aged <40, 40-49, 50-59, and 60-69 years had 3.32 times (95% CI = 1.99-5.53), 2.61 times (95% CI = 1.67-4.08), and 1.93 times (95% CI = 1.26-2.95) greater odds of having suboptimal glycemic control, respectively, than those aged ≥70 years. Married participants had 1.64 times (95% CI = 1.11-2.41) greater odds of having suboptimal glycemic control than those ever married (Table 4).

In the model to identify behavioral and psychological determinants associated with suboptimal glycemic control among DM patients by univariable logistic regression, six (6) variables were found to be associated with suboptimal glycemic control among the DM: smoking, exercise, preparing food in daily life, type of rice consumed daily, drinking tea, drinking coffee, and attitudes toward DM prevention and control (Table 5).

Table 5

Identifying behavioral and psychological determinants associated with suboptimal glycemic control among DM patients by univariable and multivariable logistic regressions.

Factors	Suboptimal glycemic control		Univariable analysis			Multivariable analysis
	Yes (%)	No (%)	OR	95% CI	p-value	AOR	95% CI	p-value
Total	530 (54.8)	437 (45.2)	N/A	N/A	N/A	N/A	N/A	N/A
Smoking
No	423 (52.2)	387 (47.8)	1.00
Yes	107 (68.2)	50 (31.8)	1.95	1.36-2.81	<0.001*
Alcohol use
No	441 (54.2)	372 (45.8)	1.00
Yes	89 (57.8)	65 (42.2)	1.11	0.79-1.58	0.526
Exercise
No	301 (60.3)	198 (39.7)	1.00
Sometime	75 (28.3)	190 (71.7)	3.85	2.79-5.31	<0.001*
Regular	61 (30.0)	142 (70.0)	3.53	2.49-5.02	<0.001*
Preparing food in daily life
Themselves	457 (53.5)	398 (46.5)	1.00
Buying	73 (65.2)	39 (34.8)	1.56	1.03-2.34	0.033*
Type of rice eaten daily
Non-sticky rice	332 (50.4)	327 (49.6)	1.00			1.00
Sticky rice	198 (64.3)	110 (35.7)	1.77	1.34-2.34	<0.001*	1.61	1.19-2.61	0.002*
Drinking tea
No	425 (51.8)	395 (48.2)	1.00
Yes	105 (71.4)	42 (28.6)	2.32	1.58-3.40	<0.001*
Drinking coffee
No	323 (48.9)	337 (51.1)	1.00
Yes	207 (67.4)	100 (32.6)	2.16	1.62-2.86	<0.001*
Stress (ST-5)
Low	390 (54.2)	329 (45.8)	1.00
Moderate	90 (56.6)	69 (43.4)	1.07	0.75-1.51	0.691
High	50 (56.2)	39 (43.8)	1.13	0.72-1.76	0.584
Knowledge regarding DM prevention and control
Low	37 (47.4)	41 (52.6)	1.00
Moderate	201 (52.3)	183 (47.7)	1.21	0.74-1.98	0.430
High	292 (57.8)	213 (42.2)	1.51	0.94-2.45	0.087
Attitudes toward DM prevention and control
Poor	122 (45.2)	148 (54.8)	1.00
Moderate	329 (58.5)	233 (41.5)	1.70	1.26-2.27	<0.001*
Positive	79 (58.5)	56 (41.5)	1.76	1.16-2.68	0.008*

*Significance level α = 0.05

Only one variable was found to be associated with suboptimal glycemic control in multivariable logistic regression. Participants who consumed sticky rice in daily life had 1.61 times (95% CI = 1.19-2.61) greater odds of having suboptimal glycemic control than those who did not (Table 5).

In the model to identify medication and biochemical markers that associated with suboptimal glycemic control among DM patients by univariable logistic regressions, seven (7) variables were found to be associated with suboptimal glycemic control among the DM patients attending hospitals in northern Thailand: duration since the DM diagnosis, forgetting to take a medication, having side effects from taking medications, a history of having wounds on the feet, having diabetic nephropathy, and having elevated HT and LDL cholesterol (Table 6).

Table 6

Identifying medication and biochemical markers associated with suboptimal glycemic control among DM patients by univariable and multivariable logistic regressions.

Factors	Suboptimal glycemic control		Univariable analysis			Multivariable analysis
	Yes (%)	No (%)	OR	95% CI	p-value	AOR	95% CI	p-value
Total	530 (54.8)	437 (45.2)	N/A	N/A	N/A	N/A	N/A	N/A
Length of having DM (year)
≤ 10	362 (51.1)	347 (48.9)	1.00			1.00
11-20	113 (65.7)	59 (34.3)	1.72	1.22-2.43	0.002*	1.98	1.37-2.86	<0.001*
> 20	55 (64.0)	31 (36.0)	1.68	1.05-2.67	0.028*	2.46	1.50-4.04	<0.001*
Experience of forgetting of taking a medicine
No	398 (50.8)	385 (49.2)	1.00			1.00
Yes	132 (71.7)	52 (28.3)	2.45	1.73-3.48	<0.001*	2.10	1.43-3.09	<0.001*
Having side effect of taking medicine
No	434 (51.9)	402 (48.1)	1.00
Yes	96 (73.3)	35 (26.7)	2.43	1.61-3.65	<0.001*
Medical expenses
Covered by the national universal scheme	519 (54.8)	428 (45.2)	0.99	0.40-2.41	0.986	2.67	1.00-7.08	0.049*
Self-paid	11 (55.0)	9 (45.0)	1.00			1.00
History of having wound on foots
No	482 (53.1)	425 (46.9)	1.00
Yes	48 (80.0)	12 (20.0)	3.17	1.69-5.94	<0.001*
Diabetic nephropathy
No	191 (50.9)	184 (49.1)	1.00
Yes	263 (54.1)	223 (45.9)	1.14	0.87-1.50	0.324
Do not know	76 (71.7)	30 (28.3)	2.33	1.46-3.71	<0.001*
Having HT
No	225 (57.0)	170 (43.0)	1.00
Yes	263 (50.9)	254 (49.1)	0.76	0.59-1.00	0.050*
Do not know	42 (76.4)	13 (23.6)	2.41	1.25-4.64	0.008*
BMI (kg/m2)
Underweight (≤ 18.50)	27 (50.9)	26 (49.1)	1.00
Normal weight (18.51-22.99)	130 (45.9)	153 (54.1)	0.81	0.45-1.47	0.503
Overweight (≥ 23.00)	373 (59.1)	258 (40.9)	1.39	0.79-2.44	0.248
HDL cholesterol (mg/dL)
Low (< 40)	144 (55.0)	118 (45.0)	1.00	0.75-1.34	0.953
Normal (≥ 40)	386 (54.8)	319 (45.2)	1.00
LDL cholesterol (mg/dL)
Normal (<100)	272 (50.9)	262 (49.1)	1.00
High (≥100)	258 (59.6)	175 (40.4)	1.46	1.13-1.89	0.003*
Triglycerides (mg/dL)
Normal (<150)	253 (52.1)	233 (47.9)	1.00
High (≥ 150)	277 (57.6)	204 (42.4)	1.23	0.95-1.58	0.110

*Significance level α = 0.05

Only three (3) variables were found to be associated with suboptimal glycemic control in multivariable logistic regression: time since being diagnosed with DM, forgetting to take a medication, and medical expenses. Participants who had been diagnosed with DM 11-20 years ago and more than 20 years ago had 1.98 times (95% CI = 1.37-2.86) and 2.46 times (1.50-4.04) greater odds of having suboptimal glycemic control, respectively, than those who had been diagnosed ≤ 10 years prior. Participants who had experienced forgetting to take medication had 2.10 times (95% CI = 1.43-3.09) greater odds of having suboptimal glycemic control than those who did not, and those who were covered for their medical expenses by the national scheme had 2.67 times (95% CI = 1.00-7.08) greater odds of more suboptimal glycemic control than those who self-paid (Table 6).

Discussion

Majority of the DM patients in Chiang Rai Province were female, aged 50-69 years, and had a poor economic and education status. Approximately, 6.6% smoked and used alcohol. A large proportion had poor knowledge and attitudes toward DM prevention and control. One-fourth had been diagnosed with DM more than 10 years ago, one-fifth had experienced forgetting to take their medication, and almost all were covered by the universal medical scheme. Being older, married, forgetting to take medication, consuming sticky rice, having a long-term diagnosis of DM, and having a fully supported medical cost were found to be associated with suboptimal glycemic control among these DM patients.

In this study, the prevalence of suboptimal glycemic control among DM patients attending hospitals was very high (54.8%), which is similar to studies conducted in Iran (58.3%) [24], India (65.4%) [25], and Oman (54.0%) [26]. However, the prevalence of suboptimal glycemic control was found to be higher than that in a study conducted by Aekplakorn et al. [7] in Thailand, which was reported to be 30.0%. The differences might be impacted by the population’s culture, especially the daily consumption of sticky rice [11] and living in a lower socioeconomic status [10] in northern Thailand compared to nationwide representation from a study conducted by Aekplakorn et al. [7]. The difference could also be impacted by the rescheduling of DM care and management during the COVID-19 pandemic [27, 28].

In our study, age was found to be associated with suboptimal glycemic control. Those who were younger were more likely to have suboptimal glycemic control than those who were older. This coincided with a study conducted in Pakistan, which showed that younger individuals were at a greater risk of having suboptimal glycemic control than older age groups [29]. However, a study conducted in Ghana [30] and Eastern Sudan [31] did not detect an association between age and suboptimal glycemic control. A study conducted in Saudi Arabia [32] reported that older age was significantly associated with suboptimal glycemic control. A study conducted in central Thailand reported that DM patients aged 60 years or older had a significantly greater chance of having suboptimal glycemic control than those aged less than 60 years [33]. The possible reason for suboptimal glycemic control among the people living in northern Thailand could be their busy farming life, and during the COVID-19 pandemic, the schedule for meeting with a doctor was extended to 3 or 6 months [34]. Moreover, many patients who have previously had good control of their blood glucose have been directed to receive their medication at health-promoting hospitals where no medical doctors or medical equipment for detecting blood glucose were available.

Married participants had a greater risk of suboptimal glycemic control than those who ever married. However, a study conducted in Ethiopia did not show a difference in suboptimal glycemic control between married and ever-married groups [35]. A study in Eastern Sudan reported that being unmarried had a greater risk of suboptimal glycemic control than being married [31]. A study in South Thailand reported that marital status was not associated with suboptimal glycemic control [36]. The people living in northern Thailand will have a busy daily life while being married due to working at their farm to support their family members. Therefore, married DM patients could have trouble following medical advice to control blood glucose.

Moreover, those who regularly consumed sticky rice were at a greater risk of having suboptimal glycemic control. This is the first report about an association between the type of rice consumed and suboptimal glycemic control. These results are supported by a study of the glycemic indexes of different types of rice in Thailand, which reported that sticky rice had the highest rapid available glucose (RAG) [37]. Sanmuangken et al. [38] reported that DM patients who consumed sticky rice were more likely to have suboptimal glycemic control than those who did not. Sticky rice consumption is common among people living in northern Thailand [11], and this is expected to be one of the most difficult issues to change in order to improve the blood glucose control among DM patients.

Haghighatpanah et al. [39] reported that those who had been diagnosed with DM more than 10 years ago had a greater chance of having poor glycemic control than those who had been diagnosed with DM more recently. This was confirmed by studies in Ethiopia [40], Palestine [41], Malaysia [42], and Myanmar [43], which reported that DM patients who had been diagnosed more than 10 years ago had a greater risk of suboptimal glycemic control than those who had been diagnosed for less than 10 years. Moreover, a study in Thailand also reported that DM patients with 10 years or more after diagnosis had a greater chance of having suboptimal glycemic control than those who had been diagnosed less than 10 years prior [33].

In this study, it was found that those who did not adhere to their prescribed medication had a greater chance of having suboptimal glycemic control than those who did. A study in Ethiopia reported that DM patients with poor medication adherence were significantly associated with suboptimal glycemic control [44]. Moreover, Basu [45] noticed that medication adherence was a major contributing factor to suboptimal glycemic control. Gebrie et al. [35] reported that poor medication adherence was a significant contributing factor to suboptimal glycemic control. A study in Thailand reported that medication adherence was one of the significant factors for suboptimal glycemic control [46].

Those who were supported with all medical fees paid by the national scheme had a greater chance of having suboptimal glycemic control than those who were self-paying. This might be because those who did not have to pay medical fees were not concerned about taking their medication or strictly following the medical advice about their blood glucose management. Those who had to pay medical fees followed the medical advice. A systematic review [47] reported that the cost of diabetes treatment in low- and middle-income countries was $5 to $40 for a visit as an outpatient, and annual inpatient costs ranged from approximately $10 to over $1,000. The medical cost for DM treatment is very high for people in northern Thailand; however, most people have been supported under the universal coverage scheme.

Some limitations were detected throughout the study that might impact the analysis and interpretation of the findings. First, this study investigated some culture factors related to uncontrolled blood glucose, such as sticky rice consumption, which is a common dietary behavior among Thai people, but a few hill tribe people (2.9%) participated in the study, which might have little impact on the interpretation. Second, some questions ask about participants’ past experiences, such as forgetting to take medications, having wounds on their feet, and experiencing side effects from taking medicine. Such questions might introduce recall bias, especially among those who were diagnosed a long time ago. Third, participants were not asked about the use of traditional herbs to heal their disease, which is practiced by some people and might interfere with HbA1c levels. Fourth, information on the prescription of regular DM drugs, which directly impact suboptimal glycemic control, was not collected in this study. Fifth, lifestyle modification should also be detected, especially those who were diagnosed several years previously, because lifestyle could also impact suboptimal glycemic control. Last, very few people in northern Thailand carry Hb variants [48], which might impact the interpretation of HbA1C. Thus, any further research should be aware of these points.

Conclusions

A large proportion of DM patients in Chiang Rai people are facing a problem of uncontrolled blood glucose in their daily disease management. Several factors are associated with suboptimal glycemic control, including personal characteristics (age and marital status), sticky rice consumption, duration since DM diagnosis, forgetting to take medications and being given free access to care. Effective public health programs are needed to improve the quality of DM management systems among both patients and health care service systems. Among the elderly, having been diagnosed a long time ago and eating sticky rice regularly require the implementation of tools or innovations to improve the blood glucose control of these patients. Copayment for medical fees should be carefully considered among DM patients. Some associated factors, such as eating sticky rice, need to be studied in more detail.

(DOCX)

Click here for additional data file.

(XLSX)

Click here for additional data file.

21 Sep 2021

PONE-D-21-24649Prevalence and predictors of suboptimal glycemic control among patients with type 2 diabetes mellitus in northern Thailand: a hospital-based cross-sectional control studyPLOS ONE

Dear Dr. Apidechkul,

Thank you for submitting your manuscript to PLOS ONE. After careful consideration, we feel that it has merit but does not fully meet PLOS ONE’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

The reviewers gave positive reviews; however, there are multiple issues especially in the Methods, Results and Discussion needed to be clarified.

==============================

Please submit your revised manuscript by Nov 05 2021 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at plosone@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pone/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Sompop Bencharit, DDS, MS, PhD, FACP

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“The author is also grateful to all research assistants from the Center of Excellence for the Hill tribe Health Research for their help in data collection.”

We note that you have provided additional information within the Acknowledgements Section that is not currently declared in your Funding Statement. Please note that funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“This research was supported by the Center of Excellence for the Hill tribe Health Research, Mae Fah Lung University, Thailand (Grant Number 1-2021). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

Additional Editor Comments:

The reviewers gave positive reviews; however, there are multiple issues especially in the Methods, Results and Discussion needed to be clarified.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: 1. As sample size calculation; why do you use e as desired precision? Because the reference journal use d as absolute error or precision.

2. Please explain for P value in the sample size calculation, I could not find this figure in reference journal.

3. For validation questionnaire, why you do use 20 Thai people to validate this questionnaire? The Cronbanch’ s alpha showed 0.72 this is acceptable value, but it’s too low. What is the explanation for this low level?

4. For data gathering procedure, how do you believe in HbA1C level? Because HbA1C can’t use among hemoglobinopathy patients and hemoglobinopathy patients are common in the northern part of Thailand.

5. If you use HbA1C level to category the group of patients, how do you sure for one time of HbA1C test?

6. Ethics approval and consent to participate, how do you make a consent among minority group? Because the data also showed for non-Thai patients.

7. Do you exclude for additional medication or herbal that might interfere with HbA1C level? If not, this is important point as in questionnaire to identified cofounding factors to suboptimal glycemic?

8. As table 2, eating sticky rice is one factor to become suboptimal glycemic patients. Could you identify more than frequency of having sticky rice, such as amount of sticky rice.

9. As table 2, exercise is a factor to stimulate insulin releasing into the blood circulation. Could you add this information to identify as factor.

10. As table 4, How do you clarify for answer to experience of forgetting of taking medicine? Because this information is recall bias and not really true answer.

11. As table 4, I didn’t agree to use medical expense as one factor to be suboptimal glycemic, this should be a medicine type that patients received.

12. As table 4, diabetic nephropathy as do not know answer look to have statistically significant. What does it means?

Reviewer #2: This article is very interesting, and it is a valuable health service research article for diabetes outcome in Thailand. It is worth to consider publication of the article.

Although it is a cross-sectional study, it was well designed considering the power, and adjusting covariates.

Abstract

Conclusion highlighted the old age, but the multiple logistic regression results showed the higher number of adjusted odds ratios for age<49. Author may check carefully and revise the conclusion. Overall conclusion should be revised carefully in the abstract.

Background and literature review were well written.

Method is clearly written.

Important comments are

1. The level of HBA1C 7 is used as cut-off point. Authors need to explain with international guideline and Thai guideline.

2. When researchers target a treatment outcome, epidemiological study becomes prognostic in clinical research. Thinking as prognostic study authors must consider "treatment" that the patient is receiving at the time of recruitment for the study. 1. Types of the diabetes treatment such as oral hypoglycemic drugs, or insulin or life-style modification should be adjusted in the final model. 2. If the majority are on the oral hypoglycemic drugs, the type of agents and their effect are worth to consider as covariates. At least 1 should be performed to control the possible bias caused by medication. If those are not possible, authors should mention it in the limitation.

4. Adherence is of the highlight. Social support is important for adherence and diabetes lifestyle modification. I recommend reviewing and cite https://pubmed.ncbi.nlm.nih.gov/34299754/ it for discussion.

5. Author should mention the analysis especially how the final model was constructed. How each variable was decided to be included in the model. Table 4 is not informing how the variables are selected in term of concept, or in term of p-value to be in the multi-variable model.

6. ***Authors may divide the multi-variable models to be separate tables such as Table 5 andTable 6.

***Table 5 can have more than one models for the outcome of poor glycemic controls. Eg Model of social demographic, model of behavioral and psychological determinants, models of medication and biochemical markers

***Table 6 is to inform non-adherence and influencing factors.

These ***revisions are important to get published.

7. It is noticed that knowledge, attitude and stress-management were carefully measured. Author may present a model with carefully measured variables to report how those impact on the glycemic control or separate model for adherence. Revision may refer to above comment.

8. Reporting analysis results are qualified with 95% confidence interval values.

Results:

Characteristic of the sample and associations are well presented in tables and written adequately. Some comments in the analysis have already covered results.

Discussion:

1.The first sentence seems to reflect the sample characteristic. It is too dogmatic. I recommend writing simply just to describe the characters of the sample.

2.It is already well known that non-adherence will cause poor-glycemic control. What factors caused non-adherence in this study setting will be the interest for the reader. It would be more interesting to learn the factors associated to non-adherence of diabetes treatment. We recommend an additional model of logistic regression analysis and thorough discussion. Again, social support should be discussed.

Conclusion:

Authors made a strong conclusion to the consumption of sticky rice as a factor associated with poorer glycemic control. It should be noted that the exposure measurement is crude for sticky rice in this study, without the measurement of dose of exposure such as amount, frequency etc. We are hesitant to agree with a strong claim in the conclusion about sticky rice. It should be a recommendation for a further study finding out the association of sticky rice and diabetes in Thailand. Please revise the conclusion.

Otherwise, it is obvious that the authors had great effort in writing background and literature review and good discussion. English language is good standard. It is a carefully designed study with important impact on the public health. Just the final part of analysis needs reorientation. We should consider publication after a revision. I hope that my comments help to improve authors’ valuable work.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Myo Nyein Aung

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

21 Oct 2021

Response to reviewers’ comments

Dear Editor and reviewers,

Thank you very much for such wonderful comments and suggestions which are greatly advantages to improve the manuscript. We have carefully revised and improved all comments accordingly. Many comments raised are very important. Some missing points in our study, We have put all these into the limitations.

Thank you so much for the great comments.

Regards,

TK

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

: Thank you. We have carefully followed the instruction of PLOS ONE.

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

“The author is also grateful to all research assistants from the Center of Excellence for the Hill tribe Health Research for their help in data collection.”

We note that you have provided additional information within the Acknowledgements Section that is not currently declared in your Funding Statement. Please note that funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

“This research was supported by the Center of Excellence for the Hill tribe Health Research, Mae Fah Lung University, Thailand (Grant Number 1-2021). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

: Thank you, we have deleted the information of funding and other information which might make confusing in acknowledgements.

Additional Editor Comments:

The reviewers gave positive reviews; however, there are multiple issues especially in the Methods, Results and Discussion needed to be clarified.

Reviewer #1: 1. As sample size calculation; why do you use e as desired precision? Because the reference journal use d as absolute error or precision.

: Sorry for making you confusing in using “e” or “d” which is different from the reference, we have replaced “d” in “e” as used in the reference.

2. Please explain for P value in the sample size calculation, I could not find this figure in reference journal.

: The p-value in the reference mentioned is 0.05 in the reference is needed to be converted into the statistic value use in the formular which is Z =1.96. Then it would say that we are using the p-value in terms of tits statistics value (opposite value) in the sample size calculation.

While Z=1.96, then the p-value =0.05 in the standard normal distribution which is the same number but in opposite interpretation. Please see reference following

https://emj.bmj.com/content/emermed/17/6/409.full.pdf?__cf_chl_jschl_tk__=pmd_log0XXIITsRikSRqJsjyRRc99G6A8IQ3oLnpO6PzrEA-1632370178-0-gqNtZGzNAfujcnBszQ3R

3. For validation questionnaire, why you do use 20 Thai people to validate this questionnaire? The Cronbach’ s alpha showed 0.72 this is acceptable value, but it’s too low. What is the explanation for this low level?

: Thank you very much for the comment. Collecting data for 20 samples in the pilot phase (to validate the questionnaire) is one of the practical steps because if a set of questions in a questions either work or not, it will be detected within 20 samples. One more point with a number of questions in the questionnaire, 20 samples is enough to detect.

Basically, in the observational study especially in a population-based research, to set up the Cronbach’ s alpha between 0.7-0.8 is one most possible and advantage. To get over 0.8 is very difficult and also does not reflect the real situation. Setting up at 0.8 and over is a bit concrete, fix, and difficult to get the number. We had tried to increase this number to be 0.8 and over, but the number of the questions remaining in the research tools become very short, required a large sample in the pilot, and do not reflect well on the study samples characteristic. In addition, with more than 10 years research experience in doing observational study of us, set up at 0.7-0.8 (acceptable level in large sample size) is much more flexible to measure the variable related to human behaviors.

We always follow the information following reference which might help both of us understand my point;

https://www.apjhs.com/index.php/apjhs/article/view/559/467

http://cda.psych.uiuc.edu/psychometrika_highly_cited_articles/cronbach_1951.pdf

: We do very hope that you understand us.

4. For data gathering procedure, how do you believe in HbA1C level? Because HbA1C can’t use among hemoglobinopathy patients and hemoglobinopathy patients are common in the northern part of Thailand.

: Thank you for the comment. Using HbA1C is the standard guideline of clinical practice according to the WHO (https://www.who.int/diabetes/publications/report-hba1c_2011.pdf) and Ministry of Public Health Thailand to identify, monitor DM patients (https://www.dmthai.org/attachments/article/443/guideline-diabetes-care-2017.pdf)

:With these references, then we used HbA1C as the marker in the study. We do understand your point, but in clinical practice HbA1C is most used.

: To make more for the point, we have put your concern on the limitation.

: Moreover only 0.88% were found 14 Hb variants among people in northern Thailand, with a few people had a form that might interfere HbA1C (Panyasai S, Fucharoen G, Fucharoen S. Hemaglobin variants in northern Thailand: prevalence, heterogeneity and molecular characteristics. Genet Test Mol Biomarkers. 2016; 20(1): 37-43.

5. If you use HbA1C level to category the group of patients, how do you sure for one time of HbA1C test?

: Basically, those who have been diagnosed for DM, they were monitored at least twice a year by HbA1C and by fasting blood glucose in every month before adjusting a prescription. Because HbA1C is used for assessing the remained glucose level which is a bit better than fasting blood glucose. Sometime, the patients do not strictly follow the prescription, but able to seriously restrict their eating behavior a few days before meeting a doctor to make sure that fasting blood glucose is met the satisfaction of a doctor. Then, the HbA1C is better to use for monitoring.

In addition, in our inclusion criterion page 4, lines 7-9, is clearly defined the inclusion criteria that at least two years of the diagnosis required to participate the study.

6. Ethics approval and consent to participate, how do you make a consent among minority group? Because the data also showed for non-Thai patients.

: Almost of the non-Thai people requited into the study, they could use Thai fluently. Only three cases who could not use Thai fluent, the information regarding the study were completed by the help of their relative who spoke Thai. Even being Thai people, some people could not read Thai, then they were asked in fingerprinting to the consent from after having been clearly explained. Please see the statement on the topic of ethic in page 6, lines 17-22.

7. Do you exclude for additional medication or herbal that might interfere with HbA1C level? If not, this is important point as in questionnaire to identified cofounding factors to suboptimal glycemic?

: Thank you for such great comment. We did not ask questions about using any herbs along the treatment to individual. We have added into the limitations, page 21, lines 19-21.

8. As table 2, eating sticky rice is one factor to become suboptimal glycemic patients. Could you identify more than frequency of having sticky rice, such as amount of sticky rice.

: So sorry, with the preliminary research objectives in identifying all possible risk factors associated with suboptimal glycemic, and the nature of a cross-sectional, we did not focus on the detail of any specific factor or behavior such as sticky rice eating. However, thank you very much for the comment, we have planned to work in the point oi our next project.

9. As table 2, exercise is a factor to stimulate insulin releasing into the blood circulation. Could you add this information to identify as factor.

:Thank you the comment, we have added information of exercise in table

10. As table 4, How do you clarify for answer to experience of forgetting of taking medicine? Because this information is recall bias and not really true answer.

: Thank you for the great notice. Asking questions on experience of forgetting of taking medicine, we used two question 15 and 16, as following;

“Q15. Have you forget taking diabetes medication last week? �Yes �No

Q16. Have you forget taking diabetes medication last month? �Yes �No”

: We agree with you that it could lead to have recall bias, then we added in a limitation, please see page 22, lines 17-19.

11. As table 4, I didn’t agree to use medical expense as one factor to be suboptimal glycemic, this should be a medicine type that patients received.

: Thank you for the great comment. However, as in Thailand, there is not everyone who has been grated to access medical fee without charging. Then, at the step of literature review, we found that an affordable was one of the factors contributing to the suboptimal glycemic. Then, it had been induced in the set of questionnaire and in the model.

: However, we totally agree with you on the type of medicines, unfortunately the information was not collected. Then we put this point into the limitation of the study. Please see page 22, lines

12. As table 4, diabetic nephropathy as do not know answer look to have statistically significant. What does it means?

: In terms of “Do not know” means that the participants (DM patients) did not know their status of diabetic nephropathy. Even the statistic significance found in the univariate analysis, but it’s not in the multivariable model.

Reviewer #2: This article is very interesting, and it is a valuable health service research article for diabetes outcome in Thailand. It is worth to consider publication of the article. Although it is a cross-sectional study, it was well designed considering the power, and adjusting covariates.

Abstract

Conclusion highlighted the old age, but the multiple logistic regression results showed the higher number of adjusted odds ratios for age<49. Author may check carefully and revise the conclusion. Overall conclusion should be revised carefully in the abstract.

: Thank you for the great notice. We have revised the conclusion, see page 2, lines 13-14.

Background and literature review were well written.

: Thank you.

Method is clearly written.

: Thank you.

Important comments are

1. The level of HBA1C 7 is used as cut-off point. Authors need to explain with international guideline and Thai guideline.

: Thank you for the concern, we had looked carefully during proposal development on the cutoff the HbA1C which is found the that standard of the American Diabetes Association, reference no.24, is used widely including the WHO has recommended. Then we used this standard as the cut off.

2. When researchers target a treatment outcome, epidemiological study becomes prognostic in clinical research. Thinking as prognostic study authors must consider "treatment" that the patient is receiving at the time of recruitment for the study.

: Thank you for the comment. We totally agreed with your point and it is one of the mistakes of our study that did not get information of the prescriptions. We have added this in the limitation. Thank you so much!

2.1 Types of the diabetes treatment such as oral hypoglycemic drugs, or insulin or life-style modification should be adjusted in the final model.

: Thank you for the great comment. However, as response previously, we did not collect information about drugs prescribes individually including insulin and we have added these points in to the limitation.

: Since a cross-sectional was used in this study, which means that a snap shot data collection was executed, then collecting information of the behavioral modification was not available. We did not do prospective study which could be able to collect data on the life-style modification.

: Thank you so much, we have kept it into the limitation for further study.

2.2 If the majority are on the oral hypoglycemic drugs, the type of agents and their effect are worth to consider as covariates. At least 1 should be performed to control the possible bias caused by medication. If those are not possible, authors should mention it in the limitation.

: Thank you for the comment. We have added the point in limitation, please see page 22, lines 19-20.

3. Adherence is of the highlight. Social support is important for adherence and diabetes lifestyle modification. I recommend reviewing and cite https://pubmed.ncbi.nlm.nih.gov/34299754/ it for discussion.

: Thank you for the suggestion. We aimed to detect the factors associated with of suboptimal glycemic control by a cross sectional which was not focus on a particularly social support for adherence and diabetes life style modification. However, these point are great to study especially those people living northern Thailand. To make readers having most benefits, we have added into the recommendation for further study. Please see in page 22, lines 20-22.

4. Author should mention the analysis especially how the final model was constructed. How each variable was decided to be included in the model. Table 4 is not informing how the variables are selected in term of concept, or in term of p-value to be in the multi-variable model.

: Thank you, we have added information of executing the analysis in page 6, lines 14-20.

5. ***Authors may divide the multi-variable models to be separate tables such as Table 5 and Table 6.

***Table 5 can have more than one models for the outcome of poor glycemic controls. Eg Model of social demographic, model of behavioral and psychological determinants, models of medication and biochemical markers

***Table 6 is to inform non-adherence and influencing factors.

These ***revisions are important to get published.

: Thank you for the great comment, we have divided into three univariable and multivariable analysis for social demographic to the Table 4, model of behavioral and psychological determinants to the Table 5, and models of medication and biochemical markers to the Table 6.

6. It is noticed that knowledge, attitude, and stress-management were carefully measured. Author may present a model with carefully measured variables to report how those impact on the glycemic control or separate model for adherence. Revision may refer to above comment.

: Thank you again for the comment. As response you in previous comment that after discussion with team and statisticians to have all variables into the same model to predict the outcome would be better in term of the power of the statistics (1-�) by reducing �-error in each step of the model including to compliance the nature of a cross-sectional study using in the study.

7. Reporting analysis results are qualified with 95% confidence interval values.

: Thank you, it is because we collect large samples and the interval a bit narrow which is presented a high precision of the estimation.

Results:

Characteristic of the sample and associations are well presented in tables and written adequately. Some comments in the analysis have already covered results.

: Thank you.

Discussion:

1.The first sentence seems to reflect the sample characteristic. It is too dogmatic. I recommend writing simply just to describe the characters of the sample.

: Thank you for the comment, we have revised it.

2.It is already well known that non-adherence will cause poor-glycemic control. What factors caused non-adherence in this study setting will be the interest for the reader. It would be more interesting to learn the factors associated to non-adherence of diabetes treatment. We recommend an additional model of logistic regression analysis and thorough discussion. Again, social support should be discussed.

: Thank you so much for the comment. Again, we aimed to investigate the factors associated with suboptimal glycemic control among the DM patients in Chiang Rai. We did not focus our objective to factors adherence to the suboptimal glycemic control, however, it is excellent point for further study. We will keep go on to make more understand the factors related to adherence of access DM clinic and follow treatment guideline of DM patients.

Conclusion:

Authors made a strong conclusion to the consumption of sticky rice as a factor associated with poorer glycemic control. It should be noted that the exposure measurement is crude for sticky rice in this study, without the measurement of dose of exposure such as amount, frequency etc. We are hesitant to agree with a strong claim in the conclusion about sticky rice. It should be a recommendation for a further study finding out the association of sticky rice and diabetes in Thailand. Please revise the conclusion.

: Thank you for the great comment. We have revised the point of conclusion, page 22 (lines 31-33) – 23 (lines 1-2)

Otherwise, it is obvious that the authors had great effort in writing background and literature review and good discussion. English language is good standard. It is a carefully designed study with important impact on the public health. Just the final part of analysis needs reorientation. We should consider publication after a revision. I hope that my comments help to improve authors’ valuable work.

: Thank you

TK

Assist Prof. Dr. Tawatchai Apidechkul, MSc (Infectious Epidemiology), Dr.P.H (Epidemiology)

Dean, School of Health Science, MFU

Director, Center of Excellence of Hill Tribe Health Research, WHO-CC

Former Hubert H Humphrey Fellow (2013-2014), Emory University

Global Health Delivery Intensive (Harvard T.H. Chan School of Public Health)

Submitted filename: Response to reviewersDMTK-2.docx

Click here for additional data file.

3 Jan 2022

Prevalence and predictors of suboptimal glycemic control among patients with type 2 diabetes mellitus in northern Thailand: a hospital-based cross-sectional control study

PONE-D-21-24649R1

Dear Dr. Apidechkul,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Sompop Bencharit, DDS, MS, PhD, FACP

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Thank you for submitting the revision and response to reviewer's comments.

Reviewers' comments:

6 Jan 2022

PONE-D-21-24649R1

Prevalence and predictors of suboptimal glycemic control among patients with type 2 diabetes mellitus in northern Thailand: a hospital-based cross-sectional control study

Dear Dr. Apidechkul:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Dr. Sompop Bencharit

Academic Editor

PLOS ONE