ORIGINAL ARTICLES Epidemiology Biostatistics and Public Health - 2020, Volume 17, Number 1

Association Between Triglyceride-Glucose

Index (TyG Index) and Type 2 Diabetes
Mellitus: A Systematic Review
Iche Andriyani Liberty (1), Naufal Karyna Putri (2)

(1) Department of Public Health-Community Medicine, Faculty of Medicine, Sriwijaya University, Palembang, Indonesia
(2) Study Program of Medical Profesion, Faculty of Medicine, Sriwijaya University, Palembang, Indonesia

CORRESPONDING AUTHOR: Iche Andriyani Liberty, Department of Public Health and Community Medicine, Epidemiology and Biostatistics sub-
department, Sriwijaya University, Palembang Indonesia
Email address: iche.aliberty@gmail.com; Phone: 6281215461615; Fax: 62711-373438

DOI: 10-2427/13250
Accepted on xxx

ABSTRACT

Background: Insulin Resistance has a major role in the pathogenesis of diabetes mellitus where there is a decrease
in sensitivity in peripheral tissues. The emergence of insulin resistance, 1-2 decades precedes before the diagnosis of
type 2 diabetes mellitus is established. This theory is supported by the usefulness of insulin resistance as a marker of
future diabetes or the prevention of type 2 diabetes by insulin-sensitizing agents. Recently, there is evidence that has
been suggested to measure insulin resistance as surrogate marker by calculated the triglyceride and glucose which is
so called as “Fasting triglyceride-glucose index” or can be shorten as “TyG index”. This study reported a systematic
review of association between TyG index and type 2 diabetes mellitus from various literatures to make a conclusion
as a basis for further research.
Methods: A literature search in EBSCOhost, ProQuest, MEDLINE, and NCBI database was performed to retrieve and
review studies reporting the association between TyG index and type 2 diabetes.
Results: All studies showed that higher TyG index were associated with higher type 2 diabetes (HR 4.36, 10.38,
9.54 for each).
Conclusion: This systematic review provides further evidence about higher TyG index is related to higher risk of
development type 2 diabetes. It represents that TyG index can predicting the risk of incident T2DM.

Key words: TyG index, Triglyceride Glucose Index, T2DM, Type 2 Diabetes Mellitus

INTRODUCTION in peripheral tissues. The emergence of IR 1-2 decades

precedes before the diagnosis of type 2 diabetes mellitus
Diabetes is the leading cause of death and morbidity is established [6,7]. Therefore, IR is useful as a marker of
worldwide [1-3]. The number of deaths caused by diabetes diabetes mellitus type 2 that is likely to occur.
mellitus in 1990 to 2014 almost doubled [4]. According
to the International Diabetes Federation, in 2015 75% of There are several direct and indirect steps can be
diabetics came from low and middle income countries [5]. used to measuring insulin such as the euglycemic clamp
Insulin Resistance (IR) has a major role in the pathogenesis test which is a standard measurement and other markers
of diabetes mellitus where there is a decrease in sensitivity which include the HOMAIR, Matsuda Index and QUICKI

Association Between Triglyceride-Glucose Index (TyG Index) and Type 2 Diabetes Mellitus: A Systematic Review e13250-1
 Epidemiology Biostatistics and Public Health - 2020, Volume 17, Number 1 ORIGINAL ARTICLES

[8]. However, in reality at primary health care insulin were discussed and resolved by consensus.
measurement is difficult to be done because it is limited The term “hazard ratio” (HR) refers to the probability
by the availability of tools, the costs and the stability of when an individual who is being observed in a clinical trial
insulin which is still doubtful [9]. Therefore, a marker that at time t, experiences an event at that time. It represents the
is simple, accurate, affordable and available in primary instantaneous rate of occurence for an individual who has
health care is still needed to physician estimated the insulin survived until the time “t’’. As for interpret the measurement
resistance that can be developed into type 2 diabetes association, a value of hazard ratio 1 means lack of
melitus (T2DM) [10]. association, a value of hazard ratio bigger than 1 is
Recently, there is evidence that has been suggested interpreted as increased risk, and a value of hazard ratio
to measure insulin resistance as surrogate marker by below 1 is interpreted as a smaller risk [16].
calculated the triglyceride and glucose which is so called In total, we screened 83 literatures from both sources.
as “Fasting triglyceride-glucose index” or can be shorten We obtained full text for 83 titles, of which we retained
as “TyG index” [10]. In the other hand, another evidence 9 for data extraction. After retrieving the full manuscript,
showed that Triglyceride-glucose index (TyG index) has literatures were excluded due to the measurement
association with HOMA-IR, the insulin supression test and association and the included variabel. All of studies that
hyperinsulinemic-euglycemic clamp [11-14]. were included as criteria used cohort as method.
The study about association between Triglyceride- The searching and data extracting was following
glucose index and type 2 diabetes mellitus has been the PRISMA flow diagram (Figure 1). Three studies were
reported on amount of literatures but the systematically conducted in China, Korea and Europe (Table 1). The
reviewed has not been reported. In this study, we conducted median follow-up ranged from 48.5 months to 10 years.
a systematic review about association between Triglyceride-
glucose index and T2DM from various literatures to make a
conclusion as a basis for further research. RESULTS

TyG index cut-off point

METHODS
All of three studies assessed the risk of developing
We used comprehensive and current database to the T2DM through the four quartiles of the TyG index, hazard
literature of association between TyG index and T2DM. ratio (HR) and the 95% confidence interval (CI) using
This systematic review followed recommendations from the cox proportional hazard model. Those studies used
the Preferred Reporting Items for Systematic Review and first quartile as reference group. Lee et al and Zhang
Meta-analysis (PRISMA). Articles included in this review et al predicted risk of T2DM through follow-up using
were identified throught literature searches of EBSCOhost, receiver operating character (ROC) curve to. Lee et al
ProQuest, MEDLINE, and NCBI for the years 2008 to conducted a Kaplan-Meier survival analysis for assesed
2018. Literature search was performed from July 4 to 5 the progression to diabetes after 4 years according to the
2019. The search terms used were (Triglyceride-glucose[All TyG index quartiles.
Fields] AND (“abstracting and indexing”[MeSH Terms] Zhang, et al explained the risk of incident diabetes
OR (“abstracting”[All Fields] AND “indexing”[All Fields]) increased at quartile 4 of TyG index which is 8.82 or
OR “abstracting and indexing”[All Fields] OR “index”[All more. For men, the area under ROC curve (AUC) for TyG
Fields])) AND (“diabetes mellitus, type 2”[MeSH Terms] index was 0.602 (0.583–0.622). Women performed
OR “type 2 diabetes mellitus”[All Fields]). The literatures larger AUC for TyG index than men 0.733 (0.717–
inclusion that used in this review must be conducted 0.748). It reported the best TyG index cut-off value for
on human, in english and on adult subjects. Therefore, diagnosis T2DM for men was 8.64 with sensitivity 0.522
studies published before 2008, editorials, grey literature, and specificity 0.642 and for women was 8.76 with
interventin studies and poor-quality studies were excluded sensitivity 0.650 and specificity 0.702 [17].
in this review. The evidence database for the literature was Research by Lee et.al, explained that during follow
assembled using established systematic review methods. up period the proportion of subjects with incidence of
All relevant studies were assessed for risk of bias diabetes increased in the TyG index quartile with quartile
using the Newcastle Ottawa Scale (NOS) in order to be four reported had a significantly high risk of progression
included in the review. Studies with NOS score above 7 to diabetes. Quartile 4 had value of TyG index 8.97 or
were considered as high- quality; a score of 6 to 7 was more. The cut-off point had optimal value at 8.8 and the
considered as moderate; and a score less than 6 was AUC value was 0.751 with 95% CI 0.704–0.799 [18].
considered as poor-quality [15]. The study of Gonzales et al reported that the risk of
Retrieved articles were reviewed independently by diabetes increased progressively at value of TyG index
two investigators (IAL and NKP) in order to gain potentially was 8.31 or more. This study demonstrated subjects that
relevant articles. All disagreements on inclusion/exclusion had fourth quartile of TyG index (TyG index value between

e13250-2 Association Between Triglyceride-Glucose Index (TyG Index) and Type 2 Diabetes Mellitus: A Systematic Review
ORIGINAL ARTICLES Epidemiology Biostatistics and Public Health - 2020, Volume 17, Number 1

FIGURE 1. PRISMA 2009 Flow Diagram of search and data extraction

8.67 and 10.15) with adjusted age and sex more likely increased across TyG index quartiles 2 to 4 with the
to develop T2DM. This study reported the risk of type 2 significantly higher risk of diabetes attached to quartile
diabetes development increased above TyG 8.43 for men 4. The HR adjusted for age and sex of quartile 4 versus
and 8.19 for women [19]. quartile 1 was 10.38 (95% CI, 3.68-29.28) [18].
Gonzalez et al., found 332 cases of incident type
2 diabetes involved with 10 years median of follow
Hazard ratios of incident diabetes up. This study reported the value of HR that adjusted
to age and sex for type 2 diabetes mellitus for fourth
Zhang et al. Reported that an increased risk quartile reach out 9.54 with 95% CL: 6.13-14.05
of incident T2DM occurred in quartiles 2, 3 and 4 and P for trend <0.001. the HRs for quartiles 2 and
compared to quartile 1 on the TyG index adjusted for 3 1.44 (95% CI, 0.84-2.46), 4.21 (95% CI, 2.64-
gender, age and family history of diabetes.The HR for 6.72) respectively [19].
each quartiles were 1.19 (95% CI 0.43-3.30), 3.50
(1.50-8.16), 4.36 (1.89-10.05) with Ptrend < 0.001
In summary, this study concluded that the risk of T2DM Study quality
incident is increasing followed with increasing of TyG
index among Chinese population, so that the index might The critical review and bias risk analyses were
be an important indicator to identify people with high risk conducted by using the Newcastle Ottawa Scale (Table
of T2DM [17]. 2). All of the included studies were identified as good
Lee et al., found that there were 101 cases of incident quality as they reached a score of more than 7. All studies
diabetes after follow up. The proportions of subjects with were included to this review considered as highquality
the incidence of diabetes during follow-up time span based on each scores gained are 9.

Association Between Triglyceride-Glucose Index (TyG Index) and Type 2 Diabetes Mellitus: A Systematic Review e13250-3
 Epidemiology Biostatistics and Public Health - 2020, Volume 17, Number 1 ORIGINAL ARTICLES

TABLE 1. Characteristics of studies included in systematic review of associations between TyG index with Type 2 Diabetes Mellitus

Publication TyG index Measure of

Author Population Design Outcome assesment
year cut-off point associations
ROC curve
5,706 (3,195 analysis:
women and AUC for The prevalence of high TyG index (quartile
2,511 men); women 0.733 4) was measured on 14.95% of women and
normoweight; (0.717-0.748) 10.00% of men
rural Chinese AUC for
Zhang et al15 2017 people; family Cohort men: 0.602 HR Risk of incident T2DM increased on quartile 2,
history of (0.583-0.622) 3 and 4 versus quartile 1 of TyG index (1.19
diabetes; waist [95% CI 0.43-3.30], 3.50 [1.50-8.16],
circumference; Cut-off points: 4.36 [1.89-10.05]
age ≥18 8.64 for men
years old 8.76 for
women
ROC curve
2,900 (882 analysis: AUC
women 0.751 (95%
and 2,078 Risk of incident T2DM significantly increased
CI 0.704-
Lee et al16 2016 men); Korean Cohort HR on quartile 4 versus quartile 1 of TyG index
0.799)
population; 10.38 (95% CI 3.68-29.28)
age ≥20 Cut-off points:
years old 8.8
ROC curve
analysis: AUC
0.75 (95% CI During 322 incident cases of diabetes 5.2%
4,820 (1,889 0.70-0.81) for women and 7.9% for men.
women and
Gonzalez et 2016 2,931 men); Cohort Cut-off points: HR
al17 Risk of incident T2DM significantly increased
European Quartile 3 with quartile 4 vs 1 of TyG index 9.54 (95%
population; versus quartile CI 6.13-14.05)
1 (8.43 for
men and 8.19
for women)

TABLE 2. Quality of nonrandomised studies in meta-analyses of The Newcastle-Ottawa Scale

Selection Comparability Outcome Total

Zhang et al **** ** *** 9
Lee et al **** ** *** 9
Gonzalez et al **** ** *** 9

DISCUSSION of insulin resistance can be used as a marker of diabetes

mellitus progression. Insulin resistance can be checked
TyG index, a measurement of the fasting triglycerides using hyperinsulinemic-euglycemic clamp, HOMA-IR and
and fasting glucose, currently can be used as a surrogate the other insulin sensitivity tests. In primary health care, the
marker because it has been reported had significant results laboratory facilities are limited, so that TyG index can be
for insulin resistance assessment in Mexican, Chinese, used as a surrogate to predict the development of diabetes
Brazilian and Korean populations [12-14]. It has also by using fasting triglyceride level and fasting glucose level.
proposed as a marker for classify metabolic health status Besides, TyG index has been proven as a better tool to
[18]. predicting the progression of diabetes mellitus compared
The association between the TyG index and glycemic to TG/HDL-C ratio and HOMA-IR [25,26].
control has been explained by several mechanisms. The sensitivity and specificity of the TyG index were
Increased of free fatty acids caused by increased of high for identifying insulin resistance in Mexico population
triglycerides level which can have an impact on increasing [27]. A study in China that conducted at nine provinces
free fatty acid flux from adipose to nonadipose tissue (Liaoning, Heilongjiang, Jiangsu, Shandon, Henan, Hubei,
thus affecting glycemic control [20]. Several studies Hunan, Guangxi, and Guizhou) reported TyG index was
reported that higher triglyceride levels in muscles and effective for predicting insulin resistance [21]. In Brazilian
liver can affect glucose metabolism in each target organ population, the study reported TyG index showed a better
[21-24]. Insulin resistance has an important role to the performance to identified insulin resistance compared to
patophysiology of type 2 diabetes, therefore examination HOMA2-IR in clinical practice [14]. From the previous

e13250-4 Association Between Triglyceride-Glucose Index (TyG Index) and Type 2 Diabetes Mellitus: A Systematic Review
ORIGINAL ARTICLES Epidemiology Biostatistics and Public Health - 2020, Volume 17, Number 1

studies showed that TyG index can be used to estimate the 10. Khan SH, Sobia F, Niazi NK, Manzoor SM, Fazal N, Ahmad F.
incidence on T2DM. Metabolic clustering of risk factors: evaluation of triglyceride-glucose
This study has collected the data from previous study index (TyG index) for evaluation of insulin resistance. Diabetology &
about the association between TyG index and incidence metabolic syndrome. 2018;10(1):74.
of T2DM to provide the evidence of its association. 11. Wan, K., Zhao, J., Huang, H., et al., 2015. The association
From 3 different populations, they showed the same cut between triglyceride/high-density lipoprotein cholesterol ratio and
off point with different cut off values based on gender. all-causemortality in acute coronary syndrome after coronary
The association between higher TyG index and risk revascularization. PLoS ONE 10 (4), e0123521. http://dx.doi.
of development T2DM is significant, it’s showed by org/10.1371/564journal.pone.0123521.
the increasing of quartiles followed by the increasing 12. Simental-Mendía, L.E., Rodríguez-Morán, M., Guerrero-Romero,
incidence of diabetes. F. The product of fasting glucose and triglycerides as surrogate for
identifying insulin resistance in apparently healthy subjects. Metab.
Syndr. Relat. Disord. 2008; 6 (4), 299–304. http://dx.doi.
CONCLUSION org/10.1089/met.2008.0034.
13. Abbasi, F., Reaven, G.M. Comparison of two methods using
This systematic review provides further evidence about plasma triglyceride concentration as a surrogate estimate of
higher TyG index is related to higher risk of development insulin action in nondiabetic subjects: triglycerides × glucose
T2DM. Increasing of TyG index followed by increasing risk versus triglyceride/high-density lipoprotein cholesterol. Metabolism.
of development T2DM. It represents that TyG index can 2011; 60 (12), 1673–1676. http://dx.doi.org/10.1016/j.
predicting the risk of incident T2DM. metabol.2011.04.006.
14. Vasques, A.C.J., Novaes, F.S., MdS, De Oliveira, et al., 2011.
TyG index performs better than HOMA in a Brazilian population:
References a hyperglycemic clamp validated study. Diabetes Res. Clin.
Pract. 93 (3), e98–e100. http://dx.doi.org/10.1016/j.
1. Global Burden of Metabolic Risk Factors for Chronic Diseases C. diabres.2011.05.030
Cardiovascular disease, chronic kidney disease, and diabetes 15. Wells GA, Shea B, O’connell D, et al., The Newcastle-Ottawa
mortality burden of cardiometabolic risk factors from 1980 to Scale (NOS) for assessing the quality if nonrandomized studies
2010: a comparative risk assessment. Lancet Diabetes Endocrinol. in meta-analyses. 2006. Epub Available from: URL: http://www.
2014;2:634–47. ohri. ca/programs/clinical_epidemiology/oxford. htm. Accessed
2. Seuring T, Archangelidi O, Suhrcke M. The economic costs of 1 August 2019.
type 2 diabetes: a global systematic review. Pharmacoeconomics. 16. Brody T. Clinical trials: study design, endpoints and biomarkers,
2015;33:811–31. drug safety, and FDA and ICH guidelines. Academic press; 2016.
3. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence 203-226
of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 17. Zhang M, Wang B, Liu Y, Sun X, Luo X, Wang C, Li L, Zhang
2010;87:4–14. L, Ren Y, Zhao Y, Zhou J. Cumulative increased risk of incident
4. Lozano R, Naghavi M, Foreman K, et al. Global and regional type 2 diabetes mellitus with increasing triglyceride glucose
mortality from 235 causes of death for 20 age groups in 1990 and index in normal-weight people: the rural Chinese cohort study.
2010: a systematic analysis for the Global Burden of Disease Study Cardiovascular diabetology. 2017 Dec;16(1):30.
2010. Lancet. 2012;380:2095–128. 18. Lee ES, Kim JH, Park SE, Park CY, Oh KW, Park SW, Rhee EJ, Lee
5. Federation ID. IDF diabetes atlas. 7th ed. 2015. http://www. WY. Predictive value of triglyceride glucose index for the risk of
diabetesatlas. org/key-messages.html. Accessed 4 July 2019. incident diabetes: a 4-year retrospective longitudinal study. PLoS
6. Friedewald, W.T., Levy, R.I., Fredrickson, D.S. Estimation of the One. 2016;11(9):e0163465.
concentration of 485 low-density lipoprotein cholesterol in plasma, 19. Navarro-González D, Sánchez-Íñigo L, Pastrana-Delgado J,
without use of the preparative ultra-486 centrifuge. Clin Chem. Fernández-Montero A, Martinez JA. Triglyceride–glucose index (TyG
1972; 18 (6), 499–502. index) in comparison with fasting plasma glucose improved diabetes
7. Knowler, W.C., Barrett-Connor, E., Fowler, S.E., et al. Reduction prediction in patients with normal fasting glucose: the vascular-
in the incidence of 501 type 2 diabetes with lifestyle intervention metabolic CUN cohort. Preventive medicine. 2016;86:99-105.
or metformin. N. Engl. J. Med. 2002; 346 (6), 502, 393–403. 20. Guerrero-Romero F, Simental-Mendıa LE, Gonzalez-Ortiz M, et al.
8. Muniyappa R, Madan R, Quon MJ. Assessing insulin sensitivity The product of triglycerides and glucose, a simple measure of insulin
and resistance in humans. In: De Groot LJ, Chrousos G, Dungan K, sensitivity. Comparison with the euglycemic-hyperinsulinemic clamp.
Feingold KR, Grossman A, Hershman JM, et al. editors. Endotext J Clin Endocrinol Metab 2010;95:3347–51.
[Internet]. South Dartmouth (MA): MDText.com, Inc.; 2000–2015. 21. Lee SH, Kwon HS, Park YM, Ha HS, Jeong SH, Yang HK, Lee JH,
9. Oddoze C, Lombard E, Portugal H. Stability study of 81 analytes Yim HW, Kang MI, Lee WC, Son HY. Predicting the development
in human whole blood, in serum and in plasma. Clin Biochem. of diabetes using the product of triglycerides and glucose: the
2012;45(6):464–9. https ://doi.org/10.1016/j.clinb ioche Chungju Metabolic Disease Cohort (CMC) study. PLoS One.
m.2012.01.012. Accessed 4 July 2019. 2014;9(2):e90430.

Association Between Triglyceride-Glucose Index (TyG Index) and Type 2 Diabetes Mellitus: A Systematic Review e13250-5
 Epidemiology Biostatistics and Public Health - 2020, Volume 17, Number 1 ORIGINAL ARTICLES

22. Parhofer KG. Interaction between glucose and lipid metabolism: more diabetology. 2014; 13(1): 146.
than diabetic dyslipidemia. Diabetes Metab J. 2015;39(5):353e62. 26. Low S, Khoo KC, Irwan B, et al. The Role of Triglyceride Glucose
23. Kelley DE, Goodpaster BH. Skeletal muscle triglyceride an aspect Index in Development of Type 2 Diabetes Mellitus. Diabetes
of regional adiposity and insulin resistance. Diabetes Care. Research and Clinical Practice. 2018; 143: 43-49.
2001;24(5):933e41. 27. Salazar, J., Bermúdez, V., Calvo, M., Olivar, L. C., Luzardo, E.,
24. Nagle CA, Klett EL, Coleman RA. Hepatic triacylglycerol Navarro, C., … Rojas, J. (2017). Optimal cutoff for the evaluation
accumulation and insulin resistance. J Lipid Re.s 2009; 50. of insulin resistance through triglyceride-glucose index: A cross-
25. Du T, Yuan G, Zhang M, Zhou X, Sun X, Yu X. Clinical usefulness of sectional study in a Venezuelan population. F1000Research, 6,
lipid ratios, visceral adiposity indicators, and the triglycerides and 1337. doi:10.12688/f1000research.12170.3
glucose index as risk markers of insulin resistance. Cardiovascular

e13250-6 Association Between Triglyceride-Glucose Index (TyG Index) and Type 2 Diabetes Mellitus: A Systematic Review