F1000Research F1000Research 2046-1402 F1000 Research Limited London, UK 10.12688/f1000research.150600.4 Research Article Articles Prevalence, Characteristics and Potential Risk Factors of Prediabetes in Primary Health Care: A Cross-Sectional Study

[version 4; peer review: 1 approved, 1 approved with reservations, 2 not approved]

 Amelia Rina Conceptualization Data Curation Investigation Methodology Supervision Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0002-0419-9622 a 1 Harahap Juliandi Data Curation Formal Analysis Investigation Methodology Resources Software Supervision Writing – Review & Editing https://orcid.org/0000-0002-1090-2003 1 Wijaya Hendri Conceptualization Funding Acquisition Investigation Resources Supervision Validation Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0002-7309-8227 2 Pase M. Aron Formal Analysis Investigation Methodology Project Administration Software Supervision Validation 3 Suryani Widjaja Sry Data Curation Formal Analysis Funding Acquisition Investigation Validation Visualization Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0001-9738-9339 4 Saktioto Saktioto Supervision Validation Visualization Writing – Original Draft Preparation Writing – Review & Editing https://orcid.org/0000-0001-9200-8998 5 Department of Community Medicine, Universitas Sumatera Utara, Medan, North Sumatra, 20155, Indonesia Department of Pediatrics, Universitas Sumatera Utara, Medan, North Sumatra, 20155, Indonesia Department of Internal Medicine, Universitas Sumatera Utara, Medan, North Sumatra, 20155, Indonesia Department of Biochemistry, Universitas Sumatera Utara, Medan, North Sumatra, 20155, Indonesia Physics Department, Math and Natural Sciences, Universitas Riau, Pekanbaru, Riau, 28293, Indonesia rina2@usu.ac.id

No competing interests were disclosed.

 16 4 2026 2024 13 843 8 4 2026 Copyright: © 2026 Amelia R et al. 2026 This is an open access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Background

Prediabetes, a reversible condition before the onset of diabetes, is a significant concern in healthcare globally. The global prediabetes epidemic has emerged and has considerably impacted health expenditures. Various risk factors play important roles in the progression of prediabetes to diabetes. Intensive lifestyle and pharmacological interventions can significantly reduce the risk of diabetes progression.

 Objective

This study aimed to determine the prevalence, characteristics, and potential risk factors of prediabetes state in primary health care in Medan in August 2023.

 Methods

The sample consisted of 89 participants. This was an analytical cross-sectional study in the community that met the inclusion and exclusion criteria. The determination of prediabetes is based on the results of blood tests, namely, the examination of fasting blood sugar levels (FBGL), 2-hour postprandial oral glucose tolerance test (OGTT), and hemoglobin A1c (HbA1C). Other examinations included lipid profiling (total cholesterol, HDL-C, LDL-C, and triglycerides). Data processing was performed using SPSS via univariate and bivariate analyses (chi-square test).

 Results

Of the 89 participants, the prevalence of prediabetes based on HbA1c, FBGL and 2-hours OGTT levels was 28.1%, 50.6%, and 28.1%, respectively. 82% of the participants were female, and 53.9% were overweight or obese based on body mass index (BMI). The risk factors for prediabetes were age >64 years, female, physical inactivity, and diastolic blood pressure ≥90 mmHg ( p<0.05). Other risk factors such age <45-64 years, consumption of vegetables/fruits, BMI, HDL, LDL, trygliceride, total cholesterol, systolic blood pressure, achantosis nigricans, and waist-hip circumference did not associate significantly ( p>0.05).

 Conclusion

This study found that the prevalence of prediabetes was 67.4% in Medan. Age >64 years, female, physical inactivity, and diastolic blood pressure ≥90 mmHg were the most important risk factors for prediabetes.

 macrovascular complication HbA1c diabetes type 2 Prediabetes Risk Factors Directorate of Research, Technology and Community Service (DRTPM) Directorate General of Higher Education, Research and Technology (Ditjen Diktiristek) Ministry of Education, Culture, Research and Technology Indonesia, and TALENTA Universitas Sumatera Utara, Medan .Indonesia This research was funded by the Directorate of Research, Technology and Community Service (DRTPM) Directorate General of Higher Education, Research and Technology (Ditjen Diktiristek) Ministry of Education, Culture, Research and Technology Indonesia, and TALENTA Universitas Sumatera Utara, Medan .Indonesia The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Amendments from Version 3

Dear Editor, In the following revised version, I have made several corrections based on the reviewer's feedback: 1. Added references to the research methods (references 29 and 30). 2. Improvements to Table 4. Hopefully, it will be well received. Best regards, Rina Amelia

 Introduction

Diabetes is a group of metabolic diseases and a serious, long-term (chronic) condition which characterized by hyperglycemia resulting from defects in insulin secretion, insulin action, or both. 1 – 3 Based on the 10th edition of the International Diabetic Federation (IDF) Atlas, the prevalence of diabetes is estimated to be 537 million adults aged 20–79 years worldwide (10.5%). This includes both type 1 and type 2 diabetes as well as diagnosed and undiagnosed diabetes. The adult population with diabetes aged between 20-79 years is estimated to be 19,465,100 in Indonesia. Instead, the prevalence of diabetes among the ages–20-79 years is 10.6% (of the total adult population aged 20-79 years is 179,720,500). In other words, one in nine people in Indonesia had diabetes. 2 T2DM is one of the most important causes of morbidity and mortality worldwide. 4

Prediabetes is a condition that results in high BGL and often leads to T2DM. 5 People with prediabetes have high BGL (are below the amount needed to be diagnosed with diabetes, but they are at a higher risk of getting diabetes. 6 According to the World Health Organization (WHO), prediabetes is an intermediate level of high blood sugar. They use two specific tests to define it: impaired FBGL, which means BGL of 6.1 to 6.9 mmol/L (110–125 mg/dL) before a meal, or impaired OGTT, which mean OGTT of 7.8 to 11.0 mmol/L (140–199 mg/dL) two hours after eating 75 g of oral glucose. 7

According to the American Diabetes Association (ADA), the criterion for identifying impaired FBGL between 5.6 and 6.9 mmol/L (100-125 mg/dL), impaired OGTT between 7.8 and 11.0 mmol/L (140-199 mg/dL), and HbA1c levels between 5.7% and 6.4% (39-47 mmol/mol). 8 , 9 Both definitions rely on FBGL measurements, 2-hour plasma glucose concentrations during an OGTT, and HbA1c concentrations. 10 In Indonesia, the diagnostic criteria for prediabetes align with those established by the ADA. 11 The significance of Impaired FBGL and impaired 2-hour postprandial OGTT is threefold: they signal an elevated chance of developing T2DM in the future, indicative of an existing heightened risk of cardiovascular disease (CVD), and identifying therapies that can prevent the onset of T2DM. 2

Individuals with Impaired FBGL and impaired OGTT are at a high risk of developing T2DM, with up to 50% within five years. 8 , 9 Untreated T2DM for a prolonged time can lead to complications such as retinopathy, neuropathy, CVD, or stroke. 12 – 14 These chronic implications contribute to diabetes distress and health expenditures. 15 , 16 Diabetes distress is a hidden emotional burden in DM. 17 Healthcare expenditure for people with diabetes is expected to reach 1,054 billion USD by 2045. 18 The cost of managing individuals with T2DM and complications is two times higher than that for individuals without complications. 19 , 20

Risk factors for prediabetes include BMI, waist circumference, ethnicity, family history, and sex. 2 , 9 , 21 Other risk factors include hypertension, low levels of HDL cholesterol, smoking, and low levels of education and income. 22 According to the RISKESDAS (National Basic Health Research) Indonesia, the increase in diabetes data is in line with the rise in obesity rates, a risk factor for diabetes, from 14.8% to 21.8% in 2013-2018. In addition, it is also in line with the increase in BMI from 11.5% to 13.6% and central obesity from 26.6% to 31%. 23 , 24 Intensive lifestyle and pharmacological intervention can significantly reduce the risk of progression to diabetes in patients with impaired FBGL or impaired OGTT. 25 , 26

Medan was chosen as the study location due to its notably high burden of diabetes, which reflects a growing concern in the region. According to the Medan City Health Office (2012), diabetes was the second most common non-communicable disease after hypertension, with 10,347 diabetes patients recorded across 39 primary health care centers. This number has continued to rise, highlighting the urgent need for early detection and prevention strategies. Furthermore, data from the Indonesian Ministry of Health (2018) reported a national diabetes prevalence of 8.5%, or approximately 20.4 million individuals, while Riskesdas (2023) data revealed that North Sumatera’s prevalence was 8.47%, indicating a sustained public health issue. Globally, the International Diabetes Federation estimated that 537 million adults had diabetes in 2021, with Indonesia ranking 7th among countries with the highest number of cases. 27 Given this context, investigating prediabetes in Medan is crucial not only due to the rising local trend but also to support resource allocation by local health authorities and to inform targeted interventions at the primary care level. This study aimed to investigate the prevalence, characteristics, and potential risk factors for prediabetes in primary health care in Medan, Indonesia.

 Methods Study design and selection criteria

A cross-sectional study of a community that fulfilled the eligibility criteria was conducted in Medan, Indonesia. The participants were people who were at least 18 years old. Participants who had been diagnosed with diabetes or were pregnant were excluded criteria. A day before the study, all participants were reminded to fast for 8 hours and were only allowed to drink plain water before we assessed their FBGL. The minimum number of participants was determined using the Slovin formula. This formula allows calculation of the minimum sample size based on an acceptable margin of error. 27

The Slovin formula was used to calculate the sample size in this study due to the unavailability of specific data on prediabetes prevalence in Medan from national sources such as Riskesdas or P2PTM at the city or district level. Since no prior estimates or reliable variability data on prediabetes in Medan could be found, the Slovin formula was deemed appropriate for determining an adequate sample size under such uncertainty. The minimum sample size can be determined using this formula, 28 with an estimated population size (N) of 890 adults attending primary health care centers in Medan within the study period and a margin of error (e) of 10% (0.1).

 Data collection

Data were collected in August 2023 in Medan, Indonesia.

Data were collected in August 2023. The recruitment of participants in this study was conducted at Padang Bulan Primary Health Care. The selection was not assisted by the local statistical agency (BPS); however, the decision was informed by health service data obtained from the Medan City Health Office, which indicated that Padang Bulan Primary Health Care serves a significant portion of the population at risk for metabolic diseases. This health center is among the primary care facilities with a consistently high number of DM cases, making it a relevant and strategic site for investigating prediabetes risk factors. Participant recruitment followed inclusion criteria and was conducted among adult patients who visited the health centre during the study period. Participation in this study is voluntary; participants are not obligated to participate. Earlier, the researcher provided an explanation regarding the ongoing research and their active participation in it. Subsequently, patients who gave their consent would sign the informed consent form.

The sampling method used was non-probability purposive sampling, based on the inclusion criteria of adult individuals visiting Padang Bulan Primary Health Care during the study period. Randomization was not feasible due to limited access to the full population registry and logistical constraints in the primary care setting. Although randomization was not applied, efforts were made to minimize sampling bias by selecting a study site with a high and stable prevalence of diabetes cases and by recruiting participants consecutively to avoid selection based on researcher discretion. Standardized data collection procedures and objective measurement tools were also used to reduce interviewer bias and ensure data consistency.

Participants then provided their background information, physical activity, consumption of vegetables or fruits, and history of high blood glucose (during pregnancy or medical checkups), which were collected through direct interviews conducted by the research team. As the data collection did not involve the use of structured questionnaires or psychometric tools, validity and reliability testing were not required. The information gathered was factual and straightforward, minimizing the risk of misinterpretation by participants. Additionally, since the data were obtained through direct interaction between researchers and participants, the potential for bias or inconsistency was further reduced. No external validator was involved, as the nature of the instrument did not necessitate expert assessment beyond the research team.

We also obtained information about achantosis nigricans to identify additional risk factors for prediabetes. The identification of acanthosis nigricans in this study was conducted by an internal medicine specialist (internist) who was directly involved in the data collection process. The internist performed physical examinations to assess the presence of acanthosis nigricans based on standard clinical diagnostic criteria, including typical skin changes such as hyperpigmentation and velvety thickening commonly found in areas like the neck or axillae. The subjects completed the study form by themselves, after which their height, weight, hip and waist circumferences, systolic and diastolic blood pressures (SBP and DBP), lipid profile, FBGL, HbA1c, and 2-hour postprandial OGTT levels were measured.

 Ethical statement

The research design was approved by the Ethics Committee of the Faculty of Medicine, Universitas Sumatera Utara. The approval number is 896/KEPK/USU/2023 (Approval date: 21 August 2023). Patient participation is voluntary; patients have no compulsion to participate in this research. Previously, the researcher explained the research protocol that would be carried out. If the patient agreed, they signed informed consent.

 Data measurement

Body height, body weight, waist circumference, and hip circumference were measured by trained research assistants. While weighing, we asked participants to remove their footwear and wear only loose clothing. Body Mass Index (BMI) was calculated using WHO criteria as the classification standard. 29

Waist circumference and hip circumference were measured using a non-stretchable tape. Patients were defined as centrally obese if they had a waist circumference of >90 cm in men and >80 cm in women. Waist circumference alone is more practical and is included in the NCEP ATP III Guidelines, we chose WHR due to its potential relevance in predicting metabolic risk in Asian populations. Measurements were performed by trained assistants using standardized protocols to minimize error. 30

The blood pressure was measured using a digital blood pressure monitor (Omron™). FBGL, HbA1c, and 2-hour postprandial OGTT levels were measured using venous blood. The process of collecting blood was conducted in two distinct phases. The initial phase was conducted following an 8-hour fasting period by the patient; the examination included measuring the patient’s FBGL, HbA1C, and lipid profile. Subsequently, the patient was administered 75 grams of glucose (sugar dissolved in water) for the OGTT assessment, and the 2 hours post-prandial was monitored. The lipid profile was checked using the enzymatic colorimetric method (Thermo Scientific™ Indiko™ Plus Clinical Chemistry Analyzer). 28 The hexokinase method (NIPRO Premier S Blood Glucose Monitoring System GM01IAA) was used to find the FBGL and 2-hour post-meal OGTT levels. 31 High-performance liquid chromatography (HPLC) was used to determine HbA1c levels (BIORAD D-10 Hemoglobin Testing System). 32

Statistical analysis

We conducted univariate analysis to determine the prevalence and demographic characteristics. Bivariate analysis was used to analyze the risk factors for prediabetes in Medan, Indonesia, using the Chi Square Test ( p<0.05). The multivariate analysis uses Poisson regression with a stepwise method, which involves entering qualified variables with a p-value of less than 0.25 into the next analysis to obtain the ratio prevalence value. It is statistically significant if p value < 0.05. However, in this study, we used the Chi-square Test for initial bivariate analysis to identify categorical associations and then employed Poisson regression with robust error variance for multivariate analysis. Logistic regression was considered, but given the moderate prevalence of the outcome and the sample size, prevalence ratios from modified Poisson regression were preferred over odds ratios to avoid overestimation of risk. Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS Inc.).

 Results Characteristics of patients

Table 1 shows that the majority of the 89 participants were housewives (69.7%) and women (82%). 40.4 Of the participants, 40.4% had a high school education, 31.5% were age range–45-54 years old, 56.2% were never engaged in physical activity, and 51.7% consumed vegetables/fruits.

Frequency distribution based on demographic characteristics.
Characteristics Frequency (n = 89) Percentage (%)
Gender
 Men 16 18
 Women 73 82
Age (years)
 <45 26 29.2
 45-54 28 31.5
 55-64 23 25.8
 >64 12 13.5
Education
 Elementary School 26 29.2
 Junior High School 13 14.6
 Senior High School 36 40.4
 Diploma 2 2.2
 Bachelor 11 12.4
 Doctoral degree 1 1.1
Employment
 Housewife 62 69.7
 Student 1 1.1
 Entrepreneur 16 18
 Civil servants 6 6.7
 Private employee 4 4.5
Physical activity
 Never 39 43.8
 More than 30 min 50 56.2
Consuming vegetables/fruits
 Everyday 46 51.7
 Not everyday 43 48.3

Note. n; sample size.

A total of 25 people had prediabetes based on HbA1c measurement, 45 had prediabetes based on FBGL measurement, and 25 had prediabetes based on 2-hour postprandial OGTT levels measurement. The prevalence of prediabetes based on HbA1c, FBG, and 2-hour postprandial OGTT was 28.1%, 50.6%, and 28.1%, respectively ( Table 2).

Prevalence of prediabetes based on HbA1c, FBGL and 2-h postprandial OGTT examination.
Measurement Frequency (n= 89) Percentage (%)
HbA1c Levels
 Normal 50 56.2
 Prediabetes 25 28.1
 Diabetes 14 15.7
Fasting Blood Glucose Levels
 Normal 30 33.7
 Prediabetes 45 50.6
 Diabetes 14 15.7
2 Hours Postprandial OGTT Levels
 Normal 49 55.1
 Prediabetes 25 28.1
 Diabetes 15 16.9

Note. n; sample size, OGTT; oral glucose tolerance test.

Table 3 shows that the median value of BMI and waist-hip circumference ratio were 26.14 kg/m 2 and 2, respectively. Based on the blood pressure examination, the median value of systolic and diastolic blood pressures were 142 and 86 mmHg, respectively. Based on lipid profile laboratory results, the median value of serum total cholesterol, HDL, LDL, and triglycerides levels were 206 mg/dL, 53 mg/dL, 122 mg/dl, and 126 mg/dl, respectively. Based on the glucose level test, the median value of FBGL, 2-hours postprandial OGTT, and HbA1c were 106 mg/dl, 136 mg/dl, and 5.6%, respectively.

The result of physical examination and laboratory.
Characteristics Mean Median 95% Confidence Interval (CI) Minimum-Maximum
BMI 26.61 kg/m 2 26.14 kg/m 2 25.53-27.68 13.23-41.87 (kg/m 2)
Waist-hip ratio 1.74 2.00 1.65-1.83 1-2
Systolic Blood Pressure 141.03 mmHg 142 mmHg 136.42-145.65 100-204 (mmHg)
Diastolic Blood Pressure 87.12 mmHg 86 mmHg 84.60-89.64 66-131 (mmHg)
Total Cholesterol 209.66 mg/dL 206 mg/dL 198.76-220.57 127-471 (mg/dL)
HDL 54.48 mg/dL 53 mg/dL 51.70-57.27 27-89 (mg/dL)
LDL 124.87 mg/dL 122 mg/dL 116.53-133.20 30-238 (mg/dL)
Triglycerides 142.87 mg/dL 126 mg/dL 123.87-161.86 51-738 (mg/dL)
Fasting blood glucose 116.64 mg/dL 106 mg/dL 108.24-125.04 77-311 (mg/dL)
2-hours postprandial OGTT 157.70 mg/dL 136 mg/dL 143.44-172.03 75-521 (mg/dL)
HbA1c 6.08% 5.60% 5.72-6.45 4.5-13%

Note. VEGF, vascular endothelial growth factor; IGF-1; Insulin-like growth factor-1; BMI, body mass index; HDL, high-density lipoprotein; LDL, low-density lipoprotein; OGTT, oral glucose tolerance test.

As shown in Table 4, the risk factors for prediabetes were significantly correlated if the p-value was <0.05. According to chi-square analysis, age >64 years, gender, daily exercise, and triglyceride levels were significantly associated with prediabetes events ( p<0.05). There was no significant relationship between age <45-64 years, consumption of vegetables/fruits, BMI, HDL, LDL, total cholesterol, systolic and diastolic blood pressure, achantosis nigricans, and waist-hip circumference ratio in prediabetes patients.

Analysis risk factors of prediabetes.
Risk Factors Prediabetes Non-Prediabetes Unadjusted Prevalence Ratio (PR) 95% CI PR p-value
Frequency (n) Percentage (%) Frequency (n) Percentage (%)
Age (years) 0.005
 <45 9 34.6 17 65.4 Ref.
 45-54 10 35.7 18 64.3 1.032 0.498 – 2.140 0.933
 55-64 9 39.1 14 69.9 1.130 0.540 – 2.365 0.745
 >64 11 91.7 1 8.3 2.648 1.515 – 4.628 0.001
Gender
 Male 13 81.3 3 18.7 Ref.
 Female 26 35.6 47 64.4 0.438 0.297 – 0.646 0.001
Daily Exercise
 More than 30 min 17 34.0 33 66.0 Ref.
 Never 22 56.4 17 43.6 1.659 1.032 – 2.667 0.034
Consuming vegetables/fruits
 Everyday 20 43.5 26 56.5 Ref.
 Not everyday 19 44.2 24 55.8 1.016 0.635 – 1.627 0.946
Body Mass Index
 Normoweight 7 36.8 12 63.2 Ref.
 Obesity 32 45.7 38 54.3 1.241 0.653 – 2.357 0.489
High Density Lipoprotein (mg/dL)
 ≥60 12 40.0 18 60.0 Ref.
 <60 27 45.8 32 54.2 1.144 0.681 – 1.922 0.604
Low Density Lipoprotein (mg/dL)
 <100 11 47.8 12 52.2 Ref.
 ≥100 28 42.4 38 57.6 0.887 0.532 – 1.479 0.653
Trygliceride (mg/dL)
 <150 18 32.1 38 67.9 Ref.
 ≥150 21 63.6 12 36.4 1.980 1.250 – 3.135 0.004
Total Cholesterol (mg/dL)
 <200 14 37.8 23 62.2 Ref.
 ≥200 25 48.1 27 51.9 1.271 0.770 – 2.096 0.337
2-h Postprandial OGTT (mg/dL) 0.0001
 <140 14 28.6 35 71.4 Ref.
 140-199 10 40.0 15 60.0 1.400 0.726 – 2.700 0.315
 ≥200 15 100 0 0 3.500 2.242 – 5.463 0.0001
Fasting Blood Glucose (mg/dL) 0.0001
 <100 8 26.7 22 73.3 Ref.
 100-125 18 39.1 28 60.9 1.467 0.730 – 2.950 0.282
 ≥126 13 100 0 0 3.750 2.065 – 6.811 0.0001
Systolic Blood Pressure (mmHg)
 <140 16 38.1 26 61.9 Ref.
 ≥140 23 48.9 24 51.1 1.285 0.792 – 2.084 0.303
Diastolic Blood Pressure (mmHg)
 <90 22 37.9 36 62.1 Ref.
 ≥90 17 54.8 14 41.2 1.446 0.914 – 2.287 0.126
Achantosis Nigricans
 No 37 42.5 50 57.5 Ref.
 Yes 2 100 0 0 2.351 0.567 – 9.756 0.105
Waist-hip Ratio
 Normal 12 52.2 11 47.8 Ref.
 Obesity 27 40.9 39 59.1 0.784 0.482 – 1.276 0.384

Note. N, sample size; p, probability less than 0.05; OGTT, oral glucose tolerance test.

Table 5 explains that based on multivariate data analysis using the Poisson regression test with the stepwise method, it was found that the factors that were purely risk factors for prediabetes were age >64 years, female, never doing daily exercise, and diastolic blood pressure ≥90 mmHg ( p<0.05).

Multivarite.
Risk Factors Adjusted Prevalence Ratio (PR) 95% CI PR p-value
Age (years)
 >64 3.130 1.787 – 5.481 0.0001
Gender
 Female 0.404 0.244 – 0.669 0.0001
Daily Exercise
 Never 1.949 1.227 – 3.096 0.005
Low Density Lipoprotein (mg/dL)
 ≥100 0.777 0.510 – 1.183 0.240
2-h Postprandial OGTT (mg/dL)
 140-199 1.527 0.857 – 2.722 0.151
 ≥200 1.460 0.701 – 3.042 0.312
Fasting Blood Glucose (mg/dL)
 ≥126 2.661 1.172 – 6.041 0.019
Systolic Blood Pressure (mmHg)
 ≥140 0.642 0.379 – 1.088 0.100
Diastolic Blood Pressure (mmHg)
 ≥90 1.805 1.125 – 2.897 0.014
Waist-hip Ratio
 Obesity 0.582 0.328 – 1.031 0.064
Discussion

Our study demonstrated that the prevalence of individuals with prediabetes in Medan, Indonesia, using HbA1c, FBGL, and 2-hours postprandial OGTT were 28.1%, 50.6%, and 28.1%, respectively. The prevalence of prediabetes based on FBGL examinations in 33 provinces in Indonesia was 10%. 33 In this study, the prevalence of prediabetes using FBGL was one-fifth of that in the previous study. Another study found that the prevalence of prediabetes in Pontianak with an FBGL > 100 mg/dL was 76.5%. A significant increase in the prevalence of prediabetes has also been reported in the US, Europe, North America, the Caribbean, Africa, West Iran, and Malaysia. 34 , 35 The global prevalence of prediabetes using FBGL and 2-hours postprandial OGTT is estimated to increase to 6.5% and 10%, respectively, in 2045. 8

The American Diabetes Association (ADA) defines impaired fasting glucose (IFG) as having a fasting plasma glucose (FPG) level of 100–125 mg/dL, impaired glucose tolerance (IGT) as having a 2-hour postprandial glucose of 140–199 mg/dL, or elevated HbA1c (5.7%–6.4%) as having “prediabetes,” or intermediate hyperglycemia, and advises this population to make preventative efforts. 9

In this study, we found that most patients with prediabetes were 55-64 years old (n=69, 39.1%), but the prevalence was higher among age >64 years (n=11, 91.7%). In this study, we also found that age >64 years was significantly associated with the incidence of prediabetes ( p=0.001, 95% CI; 1.515-4.628) based on chi square test, and was a risk factor of prediabetes ( p=0.0001, 95% CI; 1.787-5.481) based on poisson regression test. Respondents aged >64 years were 3.13 times more at risk of experiencing prediabetes. Similar to the study by Andriani et al., the majority of prediabetic patients were >50 years old. A previous study also found a significant relationship between age and incidence of prediabetes ( p=0.029). 22 Numerous additional factors that may impact the etiology of prediabetes are also associated with advanced age. Peripheral insulin resistance is increasing in tandem with these processes. with a poor diet, little exercise, or obesity. Hyperglycemia results if this process occurs in people who are at risk of developing prediabetes. The degree of environmental exposure and lifestyle choices have a significant impact on the rate and timing of development. 22

In this study, we found that 82% of the respondents were female and the most patients with prediabetes were female (n=26, 35.6%). Female was significantly associated with the incidence of prediabetes ( p=0.001, 95% CI; 0.297-4.0.646) based on Chi square test, and was a risk factor of prediabetes ( p=0.0001, 95% CI; 0.244-0.669) based on poisson regression test. Female was 0.4 times more at risk of experiencing prediabetes. This study is consistent with research conducted in Pontianak, showing that female are more prevalent among people with prediabetes. 22 Women of reproductive age are less susceptible to cardiovascular disease because of the protective effects of estrogen. Estrogen commonly reduces triglyceride and LDL-C circulating levels, while increasing HDL-C levels. However, some studies have mentioned the development of cardiovascular disease in women with lower blood glucose levels than in men. 36 – 39

The prevalence of physical inactivity in the subjects diagnosed with prediabetes in this study was 56.4%. It was significantly associated with the incidence of prediabetes ( p=0.034, 95% CI; 1.032-4.2.667) based on chi square test, and was a risk factor of prediabetes ( p=0.005, 95% CI; 1.227-3.096) based on poisson regression test. Peoples that never do daily exercise were 1.94 times more at risk of experiencing prediabetes. This is inline with previous research which states that a sedentary lifestyle influences the development of prediabetes and diabetes. Exercise helps to avoid obesity and increases insulin sensitivity. Compared to people who exercise, those who do not exercise may be more susceptible to developing prediabetes and diabetes, 40 and physical activity is known to be protective against the onset of type 2 diabetes. 41 Program-intensive lifestyle interventions from the DPP were to achieve and maintain a minimum weight loss of 7% and a physical activity of 150 min per week identical in intensity to brisk walking. The goal of physical activity was to approximate at least 700 kcal/week of physical activity. 42

The prevalence of prediabetes among groups who do not consume vegetables/fruits every day is higher (44.2%), compared to those who consume vegetables/fruits every day (43.5%). But, we did not find an association between consuming vegetables/fruits every day and prediabetes. Consuming fruits and vegetables has been linked to the prevention of a number of chronic conditions, such as diabetes and prediabetes. These benefits have been attributed to the high nutrient content and low energy of fruits and vegetables. Fruit consumption up to 200 g/day was associated with a lower relative risk of type 2 diabetes; intakes beyond this threshold were associated with an increased risk of diabetes, possibly due to the increased intake of fructose from fruit, which has been associated with a reduction in insulin sensitivity. On the other hand, results from future research have been mixed. In one study, fruit and vegetable intake was compared to prediabetes risk in 150 prediabetes cases and 150 controls. The results showed an inverse relationship. There could be a connection between these discrepancies and the nutritional evaluation technique that was employed. 43 The 12-week intervention consisted of four nutrition visits and instructions on a high-carbohydrate diet (60% to 70% daily calories), high-fiber diet, and low-fat diet (<7% calories from saturated fat). The results showed 5% weight loss. 44 In a study of participants at a high risk of diabetes, dietary fiber intake lowered postprandial blood glucose and insulin resistance. The recommended dietary fibre intake recommendation is 3.0 g or 1,000 kcal of total energy per day to prevent T2DM. 16

The pravelence of obesity based on BMI and waist-hip circumference ratio in the subjects diagnosed with prediabetes in this study was 45.7% and 40.9%, respectively. The prevalence of obesity based on BMI was higher than normoweight subjects, but the prevalence of obesity based on waist-hip circumference ratio was lower than normoweight subject.

While BMI is widely used for general adiposity assessment, we recognize its limitations in distinguishing between fat and lean mass. BMI poses bigger bias than waist circumference and waist-hip ratio despite its practicability. In this study, we discuss BMI cautiously and acknowledge that central obesity measures such as waist circumference or WHR may better reflect cardiometabolic risk, especially in Asian populations who are more prone to visceral fat accumulation. 29 , 30 However, in this study, we did not find an association between BMI, waist-hip circumference ratio and prediabetes. This study is consistent with research conducted in Pontianak, showing that overweight or obesity are more prevalent among people with prediabetes. 22 BMI is a simple anthropometric measure commonly used to measure general adiposity. 45 Asian populations have more visceral fat than Caucasian populations do. This results in metabolic disorders, lipotoxicity, and insulin resistance. In addition, limited insulin secretory capacity and genetic predisposition play important roles in the development of insulin resistance. Several studies have reported that there is no relationship between BMI and the obesity paradox, and BMI acts as a simple indicator for evaluating the risk of blood glucose and lipid metabolism. 44 Maintaining a normal weight BMI is essential in the education of patients with prediabetes and is a concern for physicians. 45

Our study found that low HDL (<60 mg/dl), high LDL (≥100 mg/dl), high trygliceride (≥150 mg/dl), and high total cholesterol (≥200 mg/dl) were more frequent among prediabetes patients. Among all of lipid profiles, high trygliceride (≥150 mg/dl) was significantly associated with the incidence of prediabetes ( p=0.004, 95% CI; 1.250-3.135) based on chi square test. This is consistent with a study by Li et al. from 2024, which also revealed that, after controlling for confounding variables, standard lipid measures showed trygliceride to be an independent risk factor for prediabetes, while HDL and LDL seemed to be possibly protective. There is evidence that a common dyslipidemia feature in prediabetic patients is hypertriglyceridemia. Increased free fatty acids from elevated trygliceride levels stimulate changes in pancreatic α cell insulin signaling and excessive glucagon release, which ultimately culminates in insulin resistance. On the other hand, insulin resistance increases trygliceride levels by blocking trygliceride lipolysis, which raises free fatty acids in the liver and lowers HDL by lowering the expression of apolipoprotein A-I, which is required for HDL synthesis. The “vicious circle” of diabetes development is aided by the causal link that exists between trygliceride and insulin resistance. Out of all the conventional lipid measures, trygliceride was found to be the most significant factor linked to prediabetes in the current investigation. 46

Triglyceride levels are known to be the most variable lipid parameter and are highly sensitive to fasting duration. While our participants fasted for approximately 8–10 hours, which is slightly below the recommended 12–14 hours for triglyceride stability, interpretation was made with caution. As supported by Nordestgaard et al. (2018)​, non-fasting lipid profiles can still be valid in large-scale screenings, especially when analytical consistency and interpretation guidelines are applied. 44 However, from the results of the multivariate analysis, none of the lipid profiles were risk factors for prediabetes in this study. 46

In this study, we found that high blood pressure was more frequent among prediabetes patients. High diastolic blood pressure (≥90 mmHg) was a risk factor of prediabetes ( p=0.014, 95% CI; 1.125-2.897) based on poisson regression test. Peoples with high diastolic blood pressure (≥90 mmHg) were 1.8 times more at risk of experiencing prediabetes.

This is consistent with a study by Huang et al. from 2020, which also stated that, one of the most prevalent chronic illnesses, hypertension, is a significant modifiable risk factor for metabolic and cardiovascular disorders, including prediabetes and diabetes. In populations with a reduced risk of cardiovascular disease, hypertension was linked to a higher risk of mortality. When prediabetes was added, this risk increased even further. 47 In this study, we also found that the prevalence of prediabetes was higher among groups who have achantosis nigricans (100%). But, it was not statistically significant ( p=0.105), possibly due to the small number of cases. This contradict the claim that achantosis nigricans was found to be associated with insulin resistance. Achantosis nigricans is a dermatosis that usually affects the neck and intertriginous surfaces. Acanthosis nigricans is a clinical marker of insulin resistance and has been found to correlate with prediabetes and T2DM risk. It is characterized by velvety, papillomatous, brownish-black, hyperkeratotic plaques. 48 Additionally, this study also contradicts a 2020 study by Alvarez that found people with normoglycemia, prediabetes, and type 2 diabetes had significant (>85%) achantosis nigricans specificity and positive predictive value. Achantosis nigricans’s positive predictive value for insulin resistance were high across nearly all categories of carbohydrate tolerance. This implies that in those who are euglycemic or have prediabetes, achantosis nigricans is a reliable and early clinical indicator of insulin resistance. 49

In this study, of 13 risk factors, only 4 risk factors have significant correlation with prediabetes in Medan ( p<0.05), namely age >64 years, female, never doing daily exercise, and diastolic blood pressure ≥90 mmHg. Preventing type 2 diabetes mellitus provides significant public health benefits, including lower rates of complications. 44 Implementing lifestyle counselling in clinical practice is feasible and cost-effective. 35 Holistic and integrated coordination is needed to assess the disease, including early detection in high-risk factor populations, targeted treatment, and intensive lifestyle modification.

This study has several limitations. The sample size was relatively small and drawn from a single health center, which may affect generalizability. The use of a non-randomized sampling method could introduce selection bias. Additionally, the cross-sectional design limits the ability to infer causality. However, we minimized bias by using standardized measurement tools, trained examiners, and validated laboratory methods. Future studies with larger, multi-site samples and longitudinal follow-up are recommended to confirm and expand these findings.

 Conclusion

This study found that the prevalence of prediabetes based on HbA1c, FBGL and 2-hours OGTT levels was 28.1%, 50.6%, and 28.1%, respectively in Medan. Age >64 years, female, physical inactivity, and diastolic blood pressure ≥90 mmHg were the most important risk factors for prediabetes. Early detection is necessary to assess high-risk factors, targeted treatment, and intensive lifestyle modifications.

Ethical statement

The research design was approved by the Ethics Committee of the Faculty of Medicine, Universitas Sumatera Utara. The approval number is 896/KEPK/USU/2023 (Approval date: 21 August 2023). Patient participation is voluntary; patients are not compelled to participate in this research. Before being included in the research, patients are given an informed consent sheet containing information about research procedures, blood examinations, the discomfort they will experience when taking blood, and other matters related to the research. If the patient understands and is willing to participate, they must sign the informed consent sheet.

 Data availability Underlying data

Figshare: Prediabetes data, https://doi.org/10.6084/m9.figshare.25612098.v1. 50

This project contains the following underlying data:

Data file: Prediabetes Data

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0)

 References Kharroubi AT Darwish HM : Diabetes mellitus: The epidemic of the century. World J. Diabetes. 2015;6(6):850–867. 26131326 10.4239/wjd.v6.i6.850 PMC4478580 Magliano DJ Boyko EJ : IDF Diabetes Atlas 10th edition scientific committee. IDF Diabetes Atlas. 10th ed. Brussels: International Diabetes Federation;2021 Amelia R Harahap J Zulham : Educational Model and Prevention on Prediabetes: A Systematic Review. Curr. Diabetes Rev. 2024;20:e101023221945. 37818560 10.2174/0115733998275518231006074504 PMC10909827 Lee CMY Colagiuri S Woodward M : Comparing different definitions of prediabetes with subsequent risk of diabetes: an individual participant data meta-analysis involving 76 513 individuals and 8208 cases of incident diabetes. BMJ Open Diabetes Res. Care. 2019;7(1):e000794. 31908797 10.1136/bmjdrc-2019-000794. eCollection 2019 PMC6936411 Vicks WS Lo JC Guo L : Prevalence of prediabetes and diabetes vary by ethnicity among U.S. Asian adults at healthy weight, overweight, and obesity ranges: an electronic health record study. BMC Public Health. 2022;22(1):1954. 36273116 10.1186/s12889-022-14362-8 PMC9587616 Khan RMM Chua ZJY Tan JC : From Pre-Diabetes to Diabetes: Diagnosis, Treatments and Translational Research. Medicina (Kaunas). 2019;55(9):546. 31470636 10.3390/medicina55090546 PMC6780236 Bansal N : Prediabetes diagnosis and treatment: A review. World J. Diabetes. 2015;6(2):296–303. 25789110 10.4239/wjd.v6.i2.296 PMC4360422 Rooney MR Fang M Ogurtsova K : Global Prevalence of Prediabetes. Diabetes Care. 2023;46(7):1388–1394. 37196350 10.2337/dc22-2376 PMC10442190 American Diabetes Association: 2. Classification and Diagnosis of Diabetes: Standards of Medical Care in Diabetes—2021. Diabetes Care. 2021;44(Supplement1):S15–S33. 33298413 10.2337/dc21-S002 Beulens J Rutters F Rydén L : Risk and management of pre-diabetes. Eur. J. Prev. Cardiol. 2019;26(2):47–54. 10.1177/2047487319880041 Soelestijo S Sanusi H Sasiarini L : Pedoman Pengelolaan dan Pencegahan Diabetes Melitus Tipe 2 Dewasa di Indonesia. PB PERKENI [Internet]. 7th ed. Indonesia: Perkumpulan Endokrinologi Indonesia;2021. Amelia R Sari Wahyuni A Yunanda Y : Atherosclerotic Cardiovascular Disease in Diabetes Patients. Curr. Diabetes Rev. 2023;19(8):e060223213457. 10.2174/1573399819666230206124638 Amelia R Wijaya H Rusdiana R : Risk of Cardiovascular Complication Among Type 2 Diabetes Mellitus Patients in Medan, Indonesia. A Cross-sectional Study. Med. Arch. 2022 Oct;76(5):324–328. 36545452 10.5455/medarh.2022.76.324-328 PMC9760233 Amelia R Harahap J Lelo A : Risk analysis for cardiovascular complication based on the atherogenic index of plasma of type 2 diabetes mellitus patients in Medan, Indonesia. Family Med. Prim. Care Rev. 2020;22(3):197–201. 10.5114/fmpcr.2020.98242 Hostalek U : Global epidemiology of prediabetes - present and future perspectives. Clin. Diabetes Endocrinol. 2019;5:5. 31086677 10.1186/s40842-019-0080-0 PMC6507173 Feldman EL Callaghan BC Pop-Busui R : Diabetic neuropathy. Nat. Rev. Dis. Prim. 2019;5:42. 10.1038/s41572-019-0092-1 Amelia R Sahbudin DKNSB Yamamoto Z : Stress level and self-concept among type 2 diabetes mellitus patients in Indonesia. Family Med. Prim. Care Rev. 2020;22(2):111–115. 10.5114/fmpcr.2020.95313 Sun H Saeedi P Karuranga S : IDF Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res. Clin. Pract. 2022;183:109119. 34879977 10.1016/j.diabres.2021.109119 PMC11057359 Hidayat B Ramadani RV Rudijanto A : Direct Medical Cost of Type 2 Diabetes Mellitus and Its Associated Complications in Indonesia. Value Health Reg. Issues. 2022;28:82–89. 34839111 10.1016/j.vhri.2021.04.006 Amelia R : The Model of Self Care Behaviour and the Relationship with Quality Of Life, Metabolic Control and Lipid Control of Type 2 Diabetes Mellitus Patients in Binjai City, Indonesia. Open Access Maced. J. Med. Sci. 2018 Sep 21;6(9):1762–1767. 30338004 10.3889/oamjms.2018.363 PMC6182544 Amelia R Harahap J Wijaya H : The role of physical activity on glucose transporter-4, fasting blood glucose level and glycate hemoglobin in type 2 diabetes mellitus patients in Medan, Indonesia. Family Med. Prim. Care Rev. 2021;23(3):274–278. 10.5114/fmpcr.2021.108188 Budiastutik I Kartasurya MI Subagio HW : High Prevalence of Prediabetes and Associated Risk Factors in Urban Areas of Pontianak, Indonesia: A Cross-Sectional Study. J. Obes. 2022;2022:4851044–4851049. 36536959 10.1155/2022/4851044 PMC9759381 Kementerian Kesehatan Republik Indonesia: InfoDATin 2020 Diabetes Melitus. Pusdatin. 2020 Amelia R Burhan B Lelo A : The Relationship between Anthropometry and Ankle-Brachial Index with Blood Glucose Level in Patients with Type 2 Diabetes Mellitus at Community Health Center in Medan, Indonesia. Family Med. Prim. Care Rev. 2018;20(4):307–312. 10.5114/fmpcr.2018.79339 Amelia R Lelo A Lindarto D : Analysis of factors affecting the self-care behaviors of diabetes mellitus type 2 patients in Binjai, North Sumatera-Indonesia. Asian J. Microbiol., Biotechnol. Environ. Sci. 2018;20(2):361–367. Rusdiana SM Amelia R : The Effect of Diabetes Self-Management Education on Hba1c Level and Fasting Blood Sugar in Type 2 Diabetes Mellitus Patients in Primary Health Care in Binjai City of North Sumatera, Indonesia. Open Access Maced. J. Med. Sci. 2018 Apr 12;6(4):715–718. 29731946 10.3889/oamjms.2018.169 PMC5927509 Anoyke MA : Sample Size Determination in Survey Research. J. Sci. Res. Rep. 2020;26(5):90–97. 10.9734/jsrr/2020/v26i530263 Kathak RR Sumon AH Molla NH : The association between elevated lipid profile and liver enzymes: a study on Bangladeshi adults. Sci. Rep. 2022;12(1):1711. 35110625 10.1038/s41598-022-05766-y PMC8810783 Tan KC : Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004. National Cholesterol Education Program (US). : Expert Panel on Detection, Treatment of High Blood Cholesterol in Adults. Third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). The Program;2002:285:2486–2497. 10.1001/jama.285.19.2486 Sonagra A Zubair M Motiani A : Hexokinase method. StatPearls;2024. Zechmeister B Erden T Kreutzig B : Analytical interference of 33 different hemoglobin variants on HbA1c measurements comparing high-performance liquid chromatography with whole blood enzymatic assay: a multi-center study. Clin. Chim. Acta. 2022;531:145–151. 35378091 10.1016/j.cca.2022.03.028 Soewondo P Pramono LA : Prevalence, Characteristics, and Predictors Of Pre-diabetes In Indonesia. Med. J. Indones. 2011;20(4):283–294. 10.13181/mji.v20i4.465 Center for Disease Control and Prevention: Prevalence of Prediabetes Among Adults. The United States of America (The US): CDC;2022 [updated 2022; cited 2023 Sept 06]. Prevalence of Prediabetes Among Adults|Diabetes|CDC. Moradpour F Rezaei S Piroozi B : Prevalence of prediabetes, diabetes, diabetes awareness, treatment, and its socioeconomic inequality in west of Iran. Sci. Rep. 2022;12(1):17892. 36284227 10.1038/s41598-022-22779-9 PMC9596718 Akhtar S Nasir JA Ali A : Prevalence of type-2 diabetes and prediabetes in Malaysia: A systematic review and meta-analysis. PLoS One. 2022;17(1):e0263139. 35085366 10.1371/journal.pone.0263139 PMC8794132 Bernabe-Ortiz A Perel P Miranda JJ : Diagnostic accuracy of the Finnish Diabetes Risk Score (FINDRISC) for undiagnosed T2DM in Peruvian population. Prim. Care Diabetes. 2018;12(6):517–525. 30131300 10.1016/j.pcd.2018.07.015 PMC6249987 Siddiqui S Zainal H Harun SN : Gender differences in the modifiable risk factors associated with the presence of prediabetes: A systematic review. Diabetes Metab. Syndr. Clin. Res. Rev. 2020;14(5):1243–1252. 32688241 10.1016/j.dsx.2020.06.069 Rao G Powell-Wiley TM Ancheta I : Identification of obesity and cardiovascular risk in ethnically and racially diverse populations: a scientific statement from the american heart association. Circulation. 2015;132(5):457–472. 26149446 10.1161/CIR.0000000000000223 Andriani RN Fitriadi Y Danawati CW : Analysis of prediabetes risk factors at primary health care centers. RevPrim. Care Prac. Educ. 2022;5(2):52–56. 10.22146/rpcpe.74552 Galaviz KI Narayan KMV Lobelo F : Lifestyle and the Prevention of Type 2 Diabetes: A Status Report. Am. J. Lifestyle Med. 2018;12(1):4–20. 30202378 10.1177/1559827615619159 PMC6125024 American Diabetes Association: Standards of Medical Care in Diabetes—2020 Abridged for Primary Care Providers. Clin. Diabetes. 2020;38(1):10–38. 31975748 10.2337/cd20-as01 PMC6969656 Barouti AA Tynelius P Lager A : Fruit and vegetable intake and risk of prediabetes and type 2 diabetes: results from a 20-year long prospective cohort study in Swedish men and women. Eur. J. Nutr. 2022;61(6):3175–3187. 35435501 10.1007/s00394-022-02871-6 PMC9363331 Diabetes Canada Clinical Practice Guidelines Expert Committee Lipscombe L Booth G : Pharmacologic Glycemic Management of Type 2 Diabetes in Adults. Can. J. Diabetes 2018;42(Suppl 1):S88–S103. 10.1016/j.jcjd.2017.10.034 Yoshida Y Jia W Yuanhao Z : Rising Prediabetes, Undiagnosed Diabetes, and Risk Factors in Young Women. Am. J. Prev. Med. 2023;64(3):423–427. 36437142 10.1016/j.amepre.2022.10.001 PMC9974837 Li M Zhang W Zhang M : Nonlinear relationship between untraditional lipid parameters and the risk of prediabetes: a large retrospective study based on Chinese adults. Cardiovasc. Diabetol. 2024;23(1):12. 38184606 10.1186/s12933-023-02103-z PMC10771669 Huang YQ Liu L Huang JY : Prediabetes and risk for all-cause and cardiovascular mortality based on hypertension status. Ann. Transl. Med. 2020;8(23):1580. 33437779 10.21037/atm-20-1155 PMC7791204 Singh SK Agrawal NK Vishwakarma AK : Association of Acanthosis Nigricans and Acrochordon with Insulin Resistance: A Cross-Sectional Hospital-Based Study from North India. Indian J. Dermatol. 2020;65(2):112–117. 32180596 10.4103/ijd. IJD_646_18 PMC7059470 Álvarez-Villalobos NA Rodríguez-Gutiérrez R González-Saldivar G : Acanthosis nigricans in middle-age adults: A highly prevalent and specific clinical sign of insulin resistance. Int. J. Clin. Pract. 2020;74(3): e13453. 31769902 10.1111/ijcp.13453 Amelia R : Prediabetes Data. 10.6084/m9.figshare.25612098.v1 10.5256/f1000research.198495.r476311 Reviewer response for version 4 Murray Leech Jacques 1 Referee https://orcid.org/0000-0001-6990-5826 Department of Clinical and Biomedical Sciences, Faculty of Health and Life Sciences, University of Exeter, Exeter, UK

Competing interests: No competing interests were disclosed.

 23 4 2026 Copyright: © 2026 Murray Leech J 2026 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation reject

This cross-sectional study investigates the prevalence of prediabetes and its associated risk factors among adults attending a primary health care centre in Medan, Indonesia. By utilizing multiple diagnostic markers (HbA1c, fasting blood glucose, and OGTT), the authors aim to provide local evidence to support early detection strategies in a region with a high diabetes burden.

However, the study’s impact is significantly limited by a very small and non-representative sample, alongside several important methodological, analytical, and reporting inconsistencies that undermine the validity and interpretation of the findings.

In its current form, the manuscript contains several substantial methodological and interpretative issues that limit confidence in the findings. I therefore recommend not approved at this stage, pending major revision.

Major points

The most critical concern is the mismatch between the study’s limited design and its overly ambitious conclusions. The authors state that the findings are “crucial… to support resource allocation by local health authorities,” yet the methodology, based on a small sample (n = 89) from a single primary care centre using non-probability sampling, is not capable of supporting such claims. While the authors report prevalence estimates based on three diagnostic criteria (HbA1c 28.1%, fasting blood glucose 50.6%, and OGTT 28.1%), these results reflect a specific high-risk clinical population rather than the general population of Medan. Extrapolation beyond this setting is therefore not justified and introduces substantial selection bias.

The study population is highly skewed (82% female), and recruitment was conducted at a single centre with a high burden of metabolic disease. This strongly suggests selection bias rather than true population characteristics. This imbalance has direct implications for interpretation. For example, the identification of “female” as a risk factor is not reliable in a dataset with such limited gender variability and may simply reflect recruitment patterns rather than a biological association.

The manuscript also suffers from significant reporting inconsistencies and technical errors that cast doubt on the data's integrity. For example, the data tables contain physiologically improbable values, such as a median waist-hip ratio of "2.00". These unintentional errors suggest a lack of rigorous data verification and compromise the reliability of the statistical outputs. There are also inconsistencies between the in-text results and the provided additional data, and many of the in-text measurements not being provided in the source data, which hinders the ability to validate the authors' findings and suggests a failure in data management.

The reporting of female as a risk factor. There is a critical inconsistency in the interpretation of the regression results presented in Table 5. The adjusted prevalence ratio for female sex is reported as 0.404 (95% CI: 0.244–0.669), which indicates a negative (protective) association with prediabetes. However, the manuscript describes female sex as a risk factor. This represents a fundamental misinterpretation of the model output and should be corrected throughout the manuscript. Taken together, the combination of substantial sampling bias and concerns regarding data quality undermines confidence in the reported associations. Additionally, the inclusion of multiple correlated variables in a model with only 89 participants raises a high potential for overfitting. Given these constraints an alternative methods such as penalized regression may be preferred.

Other points

Please clarify whether participants were taking medications that could influence metabolic measurements; if so, this should be reported and adjusted for where possible.

The manuscript would benefit from a more cautious interpretation of these findings, using less causal language, and a thorough correction of mathematical errors throughout the text.

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

No

Are all the source data underlying the results available to ensure full reproducibility?

Partly

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Partly

Reviewer Expertise:

Diabetes, Genetics, Epidemiology

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.

 10.5256/f1000research.198495.r475723 Reviewer response for version 4 Dany Frans 1 Referee https://orcid.org/0000-0003-1412-5894 National Research and Innovation Agency (BRIN), Cibinong, Indonesia

Competing interests: No competing interests were disclosed.

 17 4 2026 Copyright: © 2026 Dany F 2026 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

Dear authors, 

Thank you for sending your revision. However, some data inconsistencies remain.

1. The number of prediabetes cases is reported as 25 based on HbA1c alone, 45 based on fasting glucose alone, and 25 based on 2hPP alone. However, Table 4 lists 39 cases for the multivariate analysis. To prevent confusion between Table 2 and Table 4, consider adding an additional row in Table 2 indicating the final number of prediabetes cases (39) included in the multivariate analysis. It would also be helpful to explain in the main manuscript how these 39 individuals were selected for inclusion in the multivariate analysis  

2. It seems that you haven't done much to revise your supplementary Excel data (Data-prediabetes-2026.xlsx) in terms of sample size/numbers, terminology/typo consistency and participant confidentiality. The followings are notable, but not limited to:

a. In the Education column, some entries still use Indonesian terms such as SD and SMP. Additionally, multiple labels are used for the same category—for example, “Elementary,” “Elementry,” and “SD” all refer to elementary school. Please standardize the terminology and use a single, consistent label throughout 

b. In Table 2, the number of Housewife is 62, but there is 63 in the supplementary Excel

c. In Table 2, there are categories for Entrepreneur with 16 and Private employee with 4, but there are only 19 for the category self-employed

d. In Table 2, there are 46 individuals who consumed vegetables/fruits every day. However, there are 86 who consumed vegetables/fruits every day in the Excel data

e. There is a misclassification in the Excel column for fruit and vegetable consumption, where “Mother” has been incorrectly included

f. Please remove the Name column from the Excel data. You don't need to share this confidential information with the public. You also don't have to share the FINDRISC and Coding sheet, so you must remove both from the supplementary Excel

g. The number of prediabetes based on fasting blood glucose in the Excel is still 46, while there are 45 in the Table 2 

h. The number of prediabetes highlighted red in the Conclusion column is 28, not matched with those reported either in Table 2 or Table 4. There are also some blank cells without any category in this Conclusion column. 

i. In the supplementary data access link in the journal system, there are still two previous unrevised data (Prediabetes data.xlsx) besides the latest one. I think you no longer need these two, and you should remove both to avoid confusion. You just need to upload the revised and completed one (Data prediabetes-2026.xlsx) provided that you have already revised its contents

j. Several English spelling errors are present, such as “hipertensi” instead of “hypertension” and “neuropaty” instead of “neuropathy”, etc. Please review these carefully, as the supplementary materials will be accessible to an international audience.

As noted in my earlier feedback, it’s important to thoroughly verify data consistency across the manuscript and the Excel supplementary files, not just for the issues mentioned above. Involve your co-authors in this process, as having multiple reviewers will improve accuracy

Is the work clearly and accurately presented and does it cite the current literature?

No

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Partly

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

No

Are sufficient details of methods and analysis provided to allow replication by others?

No

Reviewer Expertise:

Diabetes and non-communicable diseases, stem cells, bioinformatics (notably molecular modeling and metabolomics).

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

 10.5256/f1000research.185117.r409725 Reviewer response for version 3 Dany Frans 1 Referee https://orcid.org/0000-0003-1412-5894 National Research and Innovation Agency (BRIN), Cibinong, Indonesia

Competing interests: No competing interests were disclosed.

 17 9 2025 Copyright: © 2025 Dany F 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

Some revisions have been made in the revised manuscript, yet there are still many unanswered/ questions that warrant further response. Please note that if all of these questions are not addressed, this manuscript cannot be approved for indexing.

The following are still unanswered:

Did any participant take medications (antihypertensive agents, antidyslipidemia drugs etc.) or herbal medicines prior to data collection? If yes, this should be described them in detail as this can affect your measurements.

How many times did researchers measure participant blood pressure and how did they decide the final number to be included in their data if there was any deviation between measurements?

Did the authors process and use the blood samples for glucose and lipid profiling in the same day? If not, how did you store them? Please specify.

Please explain on how did the authors validate the clinical chemistry (glucose and lipid profile) measurements. Did they use control beads/calibrators before and after sample measurements? Were any clinical pathologists involved in the validation?

Which cutoff criteria did investigators choose for BMI calculation? WHO in general or Asia-Pacific criteria since both have different threshold to define overweight and obesity.

Triglyceride is the most fluctuating lipid variable among common lipid parameters and usually fasting period for its reliable measurement lasts for minimum 12 hours, preferably up to 14 hours. Did the authors measure the lipid profile more than once? If the measurement was only once, then they should refer to other reliable literature or references instead of existing consensus/guidelines to define hypertriglyceridemia. (hint: the authors may refer to a review by Keirns BH et al with the title “Fasting, non-fasting and postprandial triglycerides for screening cardiometabolic risk”)

Please mention if there was any participant drop out and their proportion from the total samples in the Method

Please use the different cutoff of HDL level for women and men since both have different physiological/hormonal conditions. You might refer to NCEP ATP III Guidelines or national consensus (PERKENI)

Please also add the proportion of prediabetes as a whole (combining FBGl, OGTT and HbA1c) in Table 2 since there is a possibility of an individual have more than one abnormal blood glucose profile (e.g., have abnormal FBG1, OGTT and abnormal HbA1c value all in one participant)

Please add more discussions of HDL level difference between men and women and their influence on risk of prediabetes and acanthosis nigricans (if you also choose to evaluate this as outcome variables)

In conclusion part in Abstract, it is said that the prevalence of prediabetes was 67.4% in Medan. How did the researchers get this number? Is this the proportion of prediabetes as a whole (combining FBGl, OGTT and HbA1c)? Since what it is said in the Conclusion part before Ethical Statement, this number does not appear.

Please check again the supplementary data. Both Excel files (45686667_Prediabetes data.xlsx and 47638837_Prediabetes data.xlsx) still have the same contents (same predictor and outcome variables). There is no data of acanthosis nigricans and other measurements (blood pressure, anthropometric parameters) yet. Please complete the additional information

The study instruments (questionnaires), both in Bahasa Indonesia and English version, have not been attached. Please attach them as Supplementary Files

Is the work clearly and accurately presented and does it cite the current literature?

No

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Partly

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

No

Are sufficient details of methods and analysis provided to allow replication by others?

No

Reviewer Expertise:

Diabetes and non-communicable diseases, stem cells, bioinformatics (notably molecular modeling and metabolomics).

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

References : Fasting, non-fasting and postprandial triglycerides for screening cardiometabolic risk. Journal of Nutritional Science .2021;10: 10.1017/jns.2021.73 10.1017/jns.2021.73 Amelia Rina Community Medicine/Faculty of, Universitas Sumatera Utara, Medan, Universitas Sumatera Utara, Indonesia

Competing interests: No competing interests

 28 9 2025

Dear Editor,

First of all, I would like to sincerely thank the reviewers for their interest in my manuscript and for taking the time to provide valuable feedback. The review process has given me many important insights and made me realize that my research is still not methodologically perfect, highlighting several aspects that need improvement for future studies.

In this response letter, I will try to address the reviewers’ comments and questions in detail. I hope that the explanations provided here clarify the remaining concerns and will be acceptable. Should there still be any shortcomings, I welcome further input and corrections.The following are still unanswered:

 

Q: Did any participant take medications (antihypertensive agents, antidyslipidemia drugs etc.) or herbal medicines prior to data collection? If yes, this should be described them in detail as this can affect your measurements.   Answer: The primary objective of this study was to determine the prevalence of prediabetes and its associated risk factors. The study sample consisted of adults (minimum age 18 years) who had never been diagnosed with diabetes, and who were not taking any medications, including glucose-lowering or lipid-lowering drugs. However, during the study, some individuals were found to have undiagnosed diabetes based on laboratory test results.

Q: How many times did researchers measure participant blood pressure and how did they decide the final number to be included in their data if there was any deviation between measurements?  Answer:  Blood pressure was measured twice for each participant. Prior to measurement, each participant was required to rest in a supine position for 5 minutes to ensure a calm and stable condition. The first measurement was then taken, followed by a second measurement 2 minutes later. The blood pressure values reported in the tables represent the mean of these two measurements.

Q: Did the authors process and use the blood samples for glucose and lipid profiling in the same day? If not, how did you store them? Please specify .  Answer: All participants were asked to fast for 8–10 hours before blood collection. Blood glucose and lipid profiles were measured simultaneously. Blood was separated for HbA1C, fasting plasma glucose (FBG), and lipid profile analysis. For lipid profile, samples were collected in red-top tubes with CAT Serum Clot Activator; HbA1C samples were placed in EDTA tubes; fasting glucose and 2-hour postprandial glucose samples were also placed in EDTA tubes. Each tube was appropriately labeled with the patient’s name and grouped before being transported to the laboratory

Q: Please explain on how did the authors validate the clinical chemistry (glucose and lipid profile) measurements. Did they use control beads/calibrators before and after sample measurements? Were any clinical pathologists involved in the validation?  Anwer: All laboratory analyses (lipid profile, HbA1C, fasting and 2-hour glucose) were performed in a certified and standardized laboratory by professional analysts. According to the laboratory staff, calibration procedures were routinely conducted to ensure accuracy and reliability of the results. Calibration is performed to adjust the instruments to measure analyte concentrations based on chemical reactions, ensuring consistency and validity of results. The lipid profile was processed using an Auto Analyzer (Indiko Thermo Scientific™). Total cholesterol (TC) was measured using the cholesterol oxidase method; HDL-C and LDL-C were analyzed using the enzymatic colorimetric method; and triglycerides (TG) were measured using the GPO–Trinder method. The laboratory procedures were supervised by a clinical pathology specialist.

Q: Which cutoff criteria did investigators choose for BMI calculation? WHO in general or Asia-Pacific criteria since both have different threshold to define overweight and obesity.  Answer: Nutritional status was assessed using the Body Mass Index (BMI), calculated as weight in kilograms divided by height in meters squared (kg/m²). The World Health Organization (WHO) classification was used as the reference standard for overweight and obesity, as it is widely adopted and allows for international comparison -> Determination of the patient's nutritional status used the Body Mass Index (BMI), which is defined as body weight in kilograms divided by the square of body height in meters (kg/m2), which is then adjusted to the World Health Organization classification [15]. Reference: WHO. Body mass index - BMI, available: https://www.euro.who.int/en/health-topics/disease-prevention/nutrition/a-healthy-lifestyle/body-mass-index-bmi

Q: Triglyceride is the most fluctuating lipid variable among common lipid parameters and usually fasting period for its reliable measurement lasts for minimum 12 hours, preferably up to 14 hours. Did the authors measure the lipid profile more than once? If the measurement was only once, then they should refer to other reliable literature or references instead of existing consensus/guidelines to define hypertriglyceridemia. (hint: the authors may refer to a review by Keirns BH et al with the title “Fasting, non-fasting and postprandial triglycerides for screening cardiometabolic risk”)  Answer:  Several studies have compared fasting and non-fasting triglyceride levels, with variable results. Some reported small differences (approximately 3–5 points), while others found no significant difference. In this study, fasting samples were chosen because multiple parameters were measured (FBG, HbA1C, TC, HDL-C, LDL-C, and TG). Reference: Keirns, B. H., Sciarrillo, C. M., Koemel, N. A., & Emerson, S. R. (2021). Fasting, non-fasting and postprandial triglycerides for screening cardiometabolic risk.  Journal of Nutritional Science,  10. https://doi.org/10.1017/JNS.2021.73 .  The difference between fasting and non-fasting triglycerides was 4 mg/dL in the overall samples, and factors that determine the magnitude of differences were hypertension status, antihyperlipidemic agent use and LDL cholesterol levels. These findings may help us to use fasting and non-fasting triglycerides properly to assess the risk of CVDs. Reference: Yang, S., Liu, M., & Wu, T. (2015). Magnitude of the Difference between Fasting and Non-fasting Triglycerides, and Its Dependent Factors.  Journal of Community Medicine & Health Education,  2015(5), 1–8. https://doi.org/10.4172/2161-0711.1000375).   Non-fasting blood lipid tests could find more individuals with HTG as well as those with marked HTG among Chinese outpatients with HBP. It indicates that non-fasting blood lipid tests are worth being recommended in patients with HBP. Reference: Nabijonov, S. A. (2023). Comparison between Fasting and Non-Fasting Cut-Off Values of Triglyceride in Diagnosing High Triglyceride in Chinese Hypertensive Outpatients.  Stomatology,  12(7), 2539. https://doi.org/10.3390/jcm12072539)

Q: Please mention if there was any participant drop out and their proportion from the total samples in the Method.   Answer: As stated in the methods, the inclusion criteria were applied in a cross-sectional design. There were no participant dropouts (inclusion dan exclusion criteria)

Q: Please use the different cutoff of HDL level for women and men since both have different physiological/hormonal conditions. You might refer to NCEP ATP III Guidelines or national consensus (PERKENI).   Answer: HDL values were not stratified by sex in this study. Pregnant women were excluded due to potential hormonal influences on glucose levels. The 2021 PERKENI guidelines for dyslipidemia management were used as the reference for HDL cut-off values, since the primary aim was not to assess cardiovascular complications but rather to classify metabolic risk factors appropriately.

Q: Please also add the proportion of prediabetes as a whole (combining FBGl, OGTT and HbA1c) in Table 2 since there is a possibility of an individual have more than one abnormal blood glucose profile (e.g., have abnormal FBG1, OGTT and abnormal HbA1c value all in one participant).   Answer:Prediabetes was defined based on three markers: fasting blood glucose, OGTT, and HbA1C. Because individuals may meet only one or two of these criteria, an additional table has been included to show the diagnostic classification of prediabetes using all three parameters.

          No.      Parameter                                                      Frequency            %

         1.         Normal                                                                13                 14.6

         2         Prediabetes (FBGL and OGTT, and HbA1C)        9                  10.1

         3         Prediabetes (FBGL or OGTT or HbA1C

                    or both FBGL and OGTT, FBGL

                    and HbA1c, and OGTT and HbA1C)                      52                 58.4

        4.        Diabetes                                                                 15                 16.9

10.  Q: Please add more discussions of HDL level difference between men and women and their influence on risk of prediabetes and acanthosis nigricans (if you also choose to evaluate this as outcome variables).   Answer: As suggested by the reviewer, additional discussion has now been included in the manuscript regarding differences in HDL levels between men and women, as well as their possible influence on prediabetes and acanthosis nigricans risk.

11. Q: In conclusion part in Abstract, it is said that the prevalence of prediabetes was 67.4% in Medan. How did the researchers get this number? Is this the proportion of prediabetes as a whole (combining FBGl, OGTT and HbA1c)? Since what it is said in the Conclusion part before Ethical Statement, this number does not appear.

Answer: We would like to clarify the discrepancy in the prevalence reported in the Conclusion section of the Abstract. The correct results of our study indicate that the prevalence of prediabetes was 28.1% based on HbA1c, 50.6% based on FBGL, and 28.1% based on 2-hours OGTT. In the first version of the manuscript that we submitted, these values were reported correctly and consistently across the Abstract and main text. However, during the revision process, an unintentional error occurred, and the Abstract conclusion was mistakenly edited to “the prevalence of prediabetes was 67.4% in Medan.” This figure does not represent a calculated prevalence from our study but was erroneously inserted in the revision. We sincerely apologize for this oversight and would like to confirm that the correct prevalence values are those stated in the Results and Conclusion sections of the main text prior to the Ethical Statement.

12. Q: Please check again the supplementary data. Both Excel files (45686667_Prediabetes data.xlsx and 47638837_Prediabetes data.xlsx) still have the same contents (same predictor and outcome variables). There is no data of acanthosis nigricans and other measurements (blood pressure, anthropometric parameters) yet. Please complete the additional information.   Answer: The supplementary data files have now been revised and updated. The corrected version includes additional variables such as acanthosis nigricans, blood pressure, and anthropometric parameters.

13.  Q: The study instruments (questionnaires), both in Bahasa Indonesia and English version, have not been attached. Please attach them as Supplementary Files

Answer: The standardized Finnish Diabetes Risk Score (FINDRISC) questionnaire was used to assess prediabetes risk factors. This validated tool estimates the 10-year risk of developing type 2 diabetes without requiring  laboratory tests, making it suitable for initial screening. It consists of 8 items covering age, BMI, waist circumference, physical activity, fruit/vegetable consumption, history of hypertension treatment, history of hyperglycemia, and   family history of diabetes. In this study, these questions were used to determine the risk of prediabetes, and laboratory tests were also performed to diagnose prediabetes.

I hope these clarifications and revisions satisfactorily address the reviewers’ concerns. Thank you very much for your time, thoughtful comments, and constructive feedback, which have significantly improved the quality of this manuscript.

Best regards,

Rina Amelia

 Dany Frans

Competing interests: No competing interests were disclosed.

 31 3 2026

Dear authors,

Thank you for sending the complete supplementary data. However, there are some inconsistencies regarding the number of cases of prediabetes between your Excel data and those presented in the Tables:

1. The number of prediabetes based on Fasting Blood Glucose level (FBGL) is 45, and diabetes is 14 in Table 2 (Prevalence of prediabetes based on HbA1c, FBGL, and 2-h postprandial OGTT examination), but in the Supplementary Excel, the number of prediabetes is 46, and diabetes is 13

2. The number of prediabetes in the conclusion column (highlighted in red) in the Supplementary Excel is 28, but the number of prediabetes in the Table 4 (Analysis risk factors of prediabetes) is 39 (pay attention to the columnwise total frequency (n) of prediabetes) and 50 in the non-prediabetes column for the variable gender, daily exercise, etc. Which figures are correct and regarded as the final ones?

3. The total number or frequency (n) of prediabetes for the age variable in Table 4 is 99 (among these, 69 in the age group 55-64 years old), inconsistent with other risk factor variables (gender, daily exercise, etc.).

Please check again the entire data consistency between the data presented in the manuscript (Tables, Results, Discussion) and those in the supplementary Excel. You should perform re-analysis for the statistical test after you make sure the final numbers, notably for the prediabetes vs non-prediabetes classification.

 10.5256/f1000research.173232.r370535 Reviewer response for version 2 Dany Frans 1 Referee https://orcid.org/0000-0003-1412-5894 National Research and Innovation Agency (BRIN), Cibinong, Indonesia

Competing interests: No competing interests were disclosed.

 20 3 2025 Copyright: © 2025 Dany F 2025 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve-with-reservations

Overall, similar studies of prediabetes risk factor evaluation are already available and considerable flaws are encountered in the methods. However, it seems that the authors would like raise awareness of acanthosis nigricans as potential prediabetes determinants, which deserves further consideration.

Title

1. By nature, cross-sectional approach is not appropriate for risk factor identification in community-based studies. Cohort design is more suitable so the title should be changed by adding 'potential' word before risk factors.

2. You should also put your location of study in the title, in this case in Medan.

3. In which settings was your study conducted? In community-based such as Posyandu or Puskesmas (primary health care), referral clinics or hospital-based? Please be more specific and include it in the title as well.

Introduction

4. Please describe why did you choose Medan as your study location? Is there any unusual spike of prediabetes cases in that area compared to other regions in North Sumatera or other strong justifications?

5. Please highlight the importance of bringing up this issue in your background. In addition to risk of diabetes and its increasing trend, is there any particular interests to conduct this study? for example, rise of diabetes-related complications in Medan and the need to allocate enough resources to mitigate its increasing prevalence from local health authority (Dinkes) perspectives, development of software to assess risk of diabetes etc.? If yes, please provide the data or references to support your reasonings.

Methods

6. Please explain why did you choose the Slovin formula to calculate sample size while prediabetes prevalence estimation is known from Riskesdas or perhaps P2PTM data? You can track the data at the district (Kabupaten) level. Slovin formula is usually used when we don't have enough estimates of variability or expected proportions of a particular condition (prediabetes in this case).

7. Please describe how you get 89 participants by using the Slovin formula. Please specifiy the population size (N) and margin of error threshold (e). Please provide the data/reference to support your N (population size) and margin of error value.  

8. Please explain more on how you recruited participants. How did you choose particular regions in Medan to collect the data? Did you involve local statistical agency (BPS) to help you pick the sample area?

9. Did you perform randomization (simplified or stratified) and blinding? Sample randomization and blinding during data collection are important to minimize bias in epidemiological studies. If you didn't perform one or both, state your solid rationale and your effort to minimize sampling bias.

10. How did you get the information of acanthosis nigricans? Did you involve dermatologists, internists or clinicians to assess it? If yes, did they receive proper training prior to data collection in the field?

11. Acanthosis nigricans is actually a manifestation of insulin resistance. You might consider this variable as the outcome instead of prediabetes or you may complement your findings to make your study 'different' from the rest. There are already many studies of prediabetes risk factors with bigger sample size and determinants in different settings so you would have to make your study standout. If you are interested in analyzing acanthosis nigricans as the outcome variable then you should add your justifications in the Introduction as well.

12. Since the study instrument was filled in by the participants, how did you ensure their validity?

13. Did any external validator check their answers? External validators are usually experts who don't have any conflict of interests and are not involved in your study, but they will assess the validity of your research and/or data collection.

14. Please attach the study instrument (questionnaires) both in Bahasa Indonesia and English version so readers can get an overview of your data collection process.

15. Did any participant take medications (antihypertensive agents, antidyslipidemia drugs etc.) or herbal medicines prior to data collection? If yes, describe them in detail as this can affect your measurements. Short fasting (one day) is not enough to eliminate drug effects. If your participants went fasting only for 8 hours, please describe your effort to minimize bias of data analysis. Did you exclude them or else?

16. Please describe the instrument brand name for the anthropometric measurement (particularly, weighing scale and stadiometer for height)

17. Did you perform calibration on your anthropometric measurement tools (weighing scale, stadiometer)? If yes, please describe.

18. How many times did you measure participant blood pressure and how did you decide the final number to be included in your data if there was any deviation between measurements?

19. Did you process and use the blood samples for glucose and lipid profiling in the same day? If not, how did you store them? Please specify.

20. Please explain on how you validated the clinical chemistry (glucose and lipid profile) measurements. Did you use control beads/calibrators before and after sample measurements? Were any clinical pathologists involved in the validation?

21. Which cutoff criteria did you choose for BMI calculation? WHO in general or Asia-Pacific criteria since both have different threshold to define overweight and obesity.

22. Triglyceride is the most variable among common lipid parameters and usually fasting period for its reliable measurement lasts for minimum 12 hours, preferrably up to 14 hours. Did you measure the lipid profile more than once? If the measurement was only once, then you should refer to other reliable literature or references instead of exisiting consensus/guidelines to define hypertriglyceridemia.

23. Please mention if there was any participant drop out and their proportion from the total samples.  

24. Please provide a separate table for operational definition for each variable and their respective citation.

Statistical Analysis

25. Chi square test is appropriate when you have only categorical data. Several numerical variables might loss some information if they are aggregated into a composite categorical variable. Since some predictors are of continuous numerical type, you could perform binary logistic regression as an 'early screening' for potential risk factors although bigger sample size is preferred (rule of thumb minimum 10 samples per predictor variable).

26. Poisson regression is usually not for risk factor prediction, but to estimate how many events (of prediabetes in this case or number of participant visits to primary health care facilities) or case rates (could be incidence or prevalence rates) instead of binary outcome prediction (disease vs normal). Logistic regression is more appropriate for this purpose although the sample size in this study was not adequate. You might need to change your main objectives of this study (as well as in the title) from 'risk factor prediction' to 'prevalence rate estimation' if you keep on using Poisson regression.

27. Please use the different cutoff of HDL level for women and men since both have different physiological/hormonal conditions. You might refer to NCEP ATP III Guidelines or national consensus (PERKENI).  

28. As mentioned earlier, triglyceride cutoff might be different if the fasting period is less than 12 hours. Please check other references for this regard.

Results

29. Please also add the proportion of prediabetes as a whole (combining FBGl, OGTT and HbA1c) in Table 2 since there is a possibility of an individual have more than one abnormal blood glucose profile (e.g., have abnormal OGTT and HbA1c value)

30. In Table 3, why did you choose waist-hip-ratio over waist circumference? Waist circumference is preferred to predict abdominal obesity and metabolic syndrome risk as it is more reproducible and practical in clinical settings (already included in the NCEP ATP III Guidelines). Please note that measuring hip circumference can be more difficult than measuring waist circumference only. If the person is not accustomed to measuring hip circumference, this could be the source of errors in waist-hip ratio calculation. Unless you are properly trained and utilize specific scales that have the ability to measure lean body mass, fat mass, non-lean mass and fat-free mass accurately, waist-hip ratio calculation can be more prone to errors than waist circumference measurement.

You could refer to Waist Circumference and Waist-Hip Ratio: Report of a WHO Expert Consultation in this matter.

31. It is advisable not to include both BMI and waist circumference/waist-hip ratio in multivariate analysis as this may result in collinearity issue when performing the regression test. BMI can be omitted in this analysis.

32. Please add the binary logistic regression test results for the bivariate analysis.

Discussion

33. As mentioned earlier, BMI poses bigger bias than waist circumference and waist-hip ratio despite its practicability. Please revise this in your discussion.

34. Please add more discussions of HDL level difference between men and women and their influence on risk of prediabetes and acanthosis nigricans (if you also choose to evaluate this as outcome variables).

35. Please review the influence of fasting period on triglyceride level fluctuation / variability and the implication of measurement stability and cutoff determination.

36. As described earlier, Poisson regression is performed to estimate number of events or case rates instead of risk factors. Please revise this part in your discussion.

37. Please elaborate more on acanthosis nigricans and its risk factors, notably when associating it with predictors in your study.

38. Please describe the limitations of your study, any effort to minimize biases and further suggestions to improve your research.

Conclusion

39. Please adjust accordingly by considering the changes made in earlier sections.

Abstracts

40. Please adjust accordingly by considering the changes made in the main sections of manuscript.

41. In conclusion part in Abstract, it is said that the prevalence of prediabetes was 67.4% in Medan. How did you get this number? Is this the proportion of prediabetes as a whole (combining FBGl, OGTT and HbA1c)? If yes, you should state this in other sections (Methods, Results, Discussion)

Data Availability

42. Please check again the supplementary data. Both Excel files (45686667_Prediabetes data.xlsx and 47638837_Prediabetes data.xlsx) have the same contents (same predictor and outcome variables). There is no data of acanthosis nigricans and other measurements (blood pressure, anthropometric parameters) yet. Please complete the additional information.

43. As notified in the Methods. please attach the study instrument (questionnaires) both in Bahasa Indonesia and English version.

Is the work clearly and accurately presented and does it cite the current literature?

No

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Partly

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

No

Are sufficient details of methods and analysis provided to allow replication by others?

No

Reviewer Expertise:

Diabetes and non-communicable diseases, stem cells, bioinformatics (notably molecular modeling and metabolomics).

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.

References : Update on the NCEP ATP-III emerging cardiometabolic risk factors. BMC Med .2014;12: 10.1186/1741-7015-12-115 115 25154373 10.1186/1741-7015-12-115 Consultation WE. Waist circumference and waist-hip ratio. Report of a WHO Expert Consultation. Geneva: World Health Organization.2008; Baioumi AY. Comparing measures of obesity: waist circumference, waist-hip, and waist-height ratios. InNutrition in the Prevention and Treatment of Abdominal Obesity 2019 Jan 1 (pp. 29-40). Academic Press.2019; Sathiyakumar V, Park J, Golozar A, Lazo M, Quispe R, Guallar E, Blumenthal RS, Jones SR, Martin SS. Fasting versus nonfasting and low-density lipoprotein cholesterol accuracy. Circulation. 2018 Jan 2;137(1):10-9. : Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification. BMC Med Res Methodol .2018;18(1) : 10.1186/s12874-018-0519-5 63 29929477 10.1186/s12874-018-0519-5 : Association between Three Waist Circumference-Related Obesity Metrics and Estimated Glomerular Filtration Rates. J Clin Med .2022;11(10) : 10.3390/jcm11102876 35629005 10.3390/jcm11102876 10.5256/f1000research.173232.r333445 Reviewer response for version 2 Pradeepa Rajendra 1 Referee https://orcid.org/0000-0002-4909-3733 Madras Diabetes Research Foundation, Tamil Nadu, India

Competing interests: No competing interests were disclosed.

 11 11 2024 Copyright: © 2024 Pradeepa R 2024 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation approve

I am fine with the revision, and I approve the current manuscript.

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

No

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

Yes

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

NA

I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

 10.5256/f1000research.165188.r317377 Reviewer response for version 1 Pradeepa Rajendra 1 Referee https://orcid.org/0000-0002-4909-3733 Madras Diabetes Research Foundation, Tamil Nadu, India

Competing interests: No competing interests were disclosed.

 18 9 2024 Copyright: © 2024 Pradeepa R 2024 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation reject

Thank you for asking me to review the article entitled “Prevalence, Characteristics and Risk Factors Analysis of Prediabetes: A Cross-Sectional Study” conducted in Medan, Indonesia. The study aimed to determine the prevalence, characteristics, and risk factors of prediabetes state of Medan in August 2023 among 89 participants. The authors have reported the prevalence of prediabetes based on HbA1c, FBG and 2-hours OGTT levels as 28.1%, 50.6%, and 28.1%, respectively. They have provided the clinical and biochemical characteristics and risk factors assessed in the study population.

 The study has been conducted in one month with a very small sample size of 89 participants. The authors could also assess the prevalence using combinations of criteria such as using both HbA1c and FBG etc. The statistical analysis done is very basic, suggest to include multivariate regression to assess the association of risk factors with prediabetes.

This manuscript needs major modification

Is the work clearly and accurately presented and does it cite the current literature?

Yes

If applicable, is the statistical analysis and its interpretation appropriate?

No

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

No

Are the conclusions drawn adequately supported by the results?

Yes

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

Epidemiology of diabetes and associated complications

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.

 10.5256/f1000research.165188.r313859 Reviewer response for version 1 Abdallah Hanaa Reyad 1 Referee National Research Centre, Cairo, Egypt

Competing interests: No competing interests were disclosed.

 30 8 2024 Copyright: © 2024 Abdallah HR 2024 This is an open access peer review report distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. recommendation reject

This manuscript gives information about the prevalence of pre-diabetes in  Medan state in Indonesia which was diagnosed by three methods; HbA1c, OGTT and FBG. the authors also investigated factors associated with pre-diabetes occurrence in those people. Here are my comments about this manuscript:

1- The title: please remove the word "analysis".

2- Methods:

- I think the sample size is small for a community based study representing a state like Medan.

-  The anthropometric measures needs to be in more details and reference is needed and BMI calculation and classification is shoes insufficiency.

- The statistical analysis: the statistical tests were not used appropriately as follows:

 - The Chi square test is not used to detect associated factors; it is used to compare two groups according to qualitative data. so you should calculate the percentages in each group ( prediabetes group and normal group) separately.

- The correlation was not assessed accurately as you should use the tests of correlation e.g Pearson's correlation coefficient test.

- The risk factors associated with the occurrence of prediabetes should be assessed by Multiple logistic regression analysis not Chi square test.

3- Results:

- The prevalence of prediabetes was mentioned by three percentages according to the method used to diagnose it then latter in the text it was mentioned in only one percentage which differed completely from these findings.

- there is discrepancies in results in the tables than those in the text so please revise your results accurately.

-  HbA1c, FBGL, and 2-hours postprandial OGTT, are not risk factors for prediabetes, they are diagnostic tests  used to detect prediabetes.

4- The conclusion:

the prevalence of prediabetes mentioned was completely different than this in the results.

5-The legends of tables contains some abbreviations which was not present in the tables as IGF1 & VEGF.

Is the work clearly and accurately presented and does it cite the current literature?

Partly

If applicable, is the statistical analysis and its interpretation appropriate?

Partly

Are all the source data underlying the results available to ensure full reproducibility?

Yes

Is the study design appropriate and is the work technically sound?

Yes

Are the conclusions drawn adequately supported by the results?

Partly

Are sufficient details of methods and analysis provided to allow replication by others?

Yes

Reviewer Expertise:

Child growth and development, Diabetes, NAFLD, Obesity, nutrition, stunted growth, Autism, Down Syndrome, Oxidative stress, Gut microbiota.

I confirm that I have read this submission and believe that I have an appropriate level of expertise to state that I do not consider it to be of an acceptable scientific standard, for reasons outlined above.