To compare the incidence and risk factors for gestational diabetes mellitus (GDM) between women with twin and singleton pregnancies.
Retrospective study of all women who had a twin or singleton birth in Ontario (2012–2016). Risk ratios (RR) and 95% CIs for GDM (stratified by type of treatment) were adjusted for relevant confounding variables. Multivariable Poisson regression analysis was used to identify risk factors for GDM in twin and singleton gestations.
Of 270,843 women who met inclusion criteria, 266,942 (98.6%) and 3901 (1.4%) had a singleton and a twin pregnancy, respectively. Women with twins had a significantly higher risk for overall GDM (aRR = 1.13, 95% CI 1.01–1.28) and diet-treated GDM (aRR = 1.20, 95% CI 1.01–1.42) while the association with insulin-treated GDM was not significant (aRR = 1.07, 95% CI 0.89–1.28). Maternal age ≥ 35 years, non-Caucasian ethnicity and BMI > 30 kg/m2 were independent risk factors for GDM among women with twins and singletons, and the magnitude of the association of these factors with GDM was similar.
Women with twins are at increased risk of GDM, mainly due to a higher rate of diet-treated GDM. Despite higher baseline risk of GDM in women with twins, the effect of known risk factors for GDM is similar to that observed in singletons.
Many of the common obstetrical complications such as preeclampsia, preterm birth, and fetal growth restriction, have been clearly shown to be more common in twin pregnancies when compared to singletons . However, whether the risk of gestational diabetes mellitus (GDM) is also increased in women with twins has been a matter of debate [2,3,4,5,6,7,8,9]. Given the increased placental mass, and subsequently the higher levels of placental hormones with diabetogenic effects (e.g., human placental lactogen and steroidal hormones), it may be expected that the degree of insulin resistance and consequently the rate of GDM would be higher in twins [10, 11]. However, at the same time, the presence of two fetuses and the higher maternal basal metabolic rate may be associated with increased utilization of glucose, which may counteract to some degree the increased insulin resistance [12, 13].
Indeed, while several studies have reported an increased risk for GDM in women with twins [2,3,4,5], others did not find a difference in the risk of GDM between twin and singleton pregnancies [6,7,8,9]. One possible explanation for these conflicting results may be the lack of adequate adjustment for potential confounding variables since multiple gestations are associated with several factors that are known to increase the risk of GDM, including advanced maternal age  and maternal obesity [3, 15]. In addition, many of the previously published studies did not differentiate between the type of GDM (diet vs. insulin treated) [2, 3, 16], information which is important to understand the potential clinical implications of the higher rate of GDM in twin pregnancies.
In addition, given the physiological differences between twin and singleton pregnancies and the possible difference in the baseline risk of GDM between these groups, the impact of known risk factors for GDM may vary between these groups. Information on risk factors for GDM in twin pregnancies is of interest as it may assist care providers in identifying women with twins who are at an increased risk of GDM [5, 10, 17]. However, data on risk factors for GDM in twin compared with singleton pregnancies are limited [2, 8].
Thus, the aim of the current study was to compare the incidence of GDM and risk factors for GDM between twin and singleton pregnancies using a large provincial population-based cohort with detailed information regarding relevant confounding variables.
Materials and methods
This was a retrospective population-based study of all women who had a singleton or twin hospital birth in Ontario, Canada, between April 2012 and March 2016. Data were obtained from the Better Outcomes Registry and Network (BORN) Ontario (https://www.bornontario.ca/en/about-born/). BORN Ontario is a province-wide registry of all births in Ontario, Canada. Whenever a woman is admitted to hospital to give birth, data are collected by health care providers and hospital staff from charts, clinical forms and patient interview, and then entered into the BORN Information System (either directly or by electronic upload from a hospital’s EMR system). The BORN Information System contains maternal demographics, health behaviors and reproductive history, as well as clinical information related to pregnancy, labor, birth and fetal and neonatal outcomes. An ongoing program of data verifications, quality checks, and formal training sessions for individuals collecting and entering data assures a high level of data quality is maintained.
Women were classified into two groups based on plurality: singletons and twins. Women with any of the following conditions were excluded from both groups: gestational age at birth < 280/7 weeks (i.e., prior to routine screening for GDM) or > 420/7 weeks, high-order multiple gestations; maternal age < 19 years, pre-existing diabetes, pregnancies complicated by genetic or structural fetal anomalies, or cases with missing data. The study was approved by the Sunnybrook Health Sciences Centre Research Ethics Board.
Diagnosis of GDM
This study spans two time periods with regards to criteria for diagnosis of GDM in Ontario. Up until April 2013, the criteria for diagnosis were according to 2008 Canadian Diabetes Associations (CDA) guidelines . These guidelines recommended screening for GDM using a 50-g glucose challenge test (GCT), and when positive (> 7.8 mmol/L or 140 mg/dL), a 75-g oral glucose tolerance test (OGTT) followed (cutoff values: fasting ≥ 5.3 mmol/L or 96 mg/dL; 1 h ≥ 10.6 mmol or 191 mg/dL; 2 h ≥ 8.9 mmol/L or 160 mg/dL). GDM was defined as ≥ 2 abnormal OGTT values or a GCT result of ≥ 10.3 mmol/L or 185 mg/dL. The presence of a single abnormal OGTT value was defined as impaired glucose tolerance (IGT).
In April 2013, new CDA criteria were published . The new guidelines allowed two options for screening/testing for GDM. The “Preferred” option was essentially identical to the CDA 2008 guidelines aside from increasing the diagnostic 50-g GCT value from 10.3 mmol (185 mg/dL) to ≥ 11.1 mmol (200 mg/dL), and the 2-h 75-g OGTT threshold from 8.9 mmol/L (160 mg/dL) to 9.0 mmol/L (162 mg/dL). The distinction between IGT and GDM was eliminated in these new guidelines.
Baseline characteristics and rate of GDM were compared between the twins and singletons groups. The chi-square and Fisher’s exact tests were used for categorical variables and the Mann–Whitney U test was used for continuous variables. Modified Poisson regression model with robust error variance was used to calculate the risk ratios for GDM in twin pregnancies (using singleton pregnancies as reference) while adjusting for variables that were found to be different between twins and singletons in the bivariate analysis or variables that are known to be associated with GDM such as pre-pregnancy BMI and ethnicity.
To identify risk factors for GDM among twin and singleton pregnancies, the characteristics of women with twins and GDM were compared with those with twins without GDM. Similar separate comparison was performed for women with singleton pregnancies with and without GDM. A modified Poisson regression model was used to identify independent risk factors for GDM among twins and among singletons, while adjusting for variables that were found to be different between women with and without GDM in the bivariate analysis.
Data were analyzed using the SAS statistical software version 9.4. Significance was set at a two-sided p value < 0.05.
Characteristics of the study population
A total of 561,798 women gave birth in Ontario, Canada, during the study period. Of the 270,843 women who met the inclusion criteria, 266,942 (98.6%) and 3901 (1.4%) had a singleton and a twin pregnancy, respectively (Fig. 1).
The characteristics of women with singleton and twin gestations are presented in Table 1. Women with twins were older, were more likely to be nulliparous and to have a history of pre-existing hypertension, were less likely to smoke and had a higher BMI, although the absolute differences between the groups in all these variables were small. Although women in both groups were mostly of Caucasian ethnicity, women in the twins group were more likely to be Caucasians and less likely to be of Asian ethnicity (Table 1). Finally, women with twins were more likely to conceive following fertility treatments compared to singletons (Table 1).
Risk of GDM in twins and singletons
The incidence of GDM in the twins and singletons groups is presented in Table 2. Women with twins had a higher rate of overall GDM (8.4 vs. 6.3%, p < 0.001), diet-treated GDM (4.8 vs. 3.5%, p < 0.001), and insulin-treated GDM (3.6 vs. 2.7%, p = 0.001) compared with singletons (Table 2). When adjusted for potential confounding variables (including maternal age, parity, ethnicity, maternal pre-pregnancy obesity, and assisted reproductive technology), women with twins had a significantly higher risk for overall GDM (aRR = 1.13, 95% CI 1.01–1.28) and diet-treated GDM (aRR = 1.20, 95% CI 1.01–1.42) while the association with insulin-treated GDM became non-significant (aRR = 1.07, 95% CI 0.89–1.28) (Table 2).
Risk factors for GDM in twin and singleton gestations
To identify risk factors for GDM in twin and singleton gestations, we compared the characteristic of women with and without GDM in the twins group, and a similar separate comparison was performed for women in the singletons group (Table 3).
In both the twins and singletons groups, women with GDM were older, were less likely to be of Caucasian ethnicity, had a higher BMI, and were more likely to conceive by in vitro fertilization compared with women without GDM (Table 3). Among women with singletons, those with GDM were also more likely to be nulliparous, were more likely to have pre-existing hypertension and were less likely to smoke compared with those without GDM (Table 3).
In an adjusted analysis, maternal age ≥ 35 years, non-Caucasian ethnicity and BMI > 30 kg/m2 were independent risk factors for GDM in both the twin and singleton groups, and the magnitude of association between these factors and GDM was similar for both groups (Table 4). This is also demonstrated when assessing the relationship of maternal age and BMI with GDM in a continuous manner (Fig. 2). Despite the higher baseline rate of GDM in women with twins, the rate of GDM increased in a similar continuous manner in both twins and singletons with maternal age (Fig. 2a) and maternal BMI (Fig. 2b). Fertility treatments were identified as an independent risk factor for GDM among women with singleton but not for women with twins (Table 4).
Principal findings of the study
In the current study, we aimed to evaluate the association of plurality with the risk of GDM and to compare risk factors of GDM between twin and singleton gestations. Our main findings are: (1) women with twins are at a significantly higher risk for GDM compared with singletons, mainly due to an increased risk for mild (diet-treated) GDM; (2) risk factors for GDM in twin gestations are similar to those identified in singletons, and the magnitude of association of these factors with GDM is similar in both groups.
Results of the study in the context of other observations
Previous studies have shown that the rate of GDM in twin gestations ranges between 3 and 9% [2, 5, 9, 10, 14,15,16, 19,20,21,22,23,24], which is in accordance with our results (8.4%). However, with regard to the question of whether women with twins are at increased risk of GDM compared with women with a singleton pregnancy (similar to what is known for other pregnancy complications such as preterm birth and preeclampsia), data are conflicting. While some found, in agreement with our findings, that women with twins are at increased risk of GDM compared with singletons [2, 3, 5, 14, 24], other failed to detect such a difference in the risk of GDM between twins and singletons [8, 9, 15, 23]. This controversy may be attributed, at least in part, to limitations of previous studies including small sample size which may result in a type-2 error [2, 3, 8, 9, 14, 24], and lack of adjustment for potential confounding variables such as pre-pregnancy BMI and maternal ethnic group [5, 9, 15, 24]. In addition, many of the studies did not provide data on the type of GDM, information that is important to understand the potential clinical implications of the higher rate of GDM in twin pregnancies. The importance of the current study lies in our large cohort of both singleton and twin pregnancies and the adjustment for major confounders, such as BMI and maternal ethnicity, which helped to determine the true effect of plurality on the rate of GDM in general and on the rate of the subtypes of GDM (i.e., diet and insulin treated).
We found that even after adjustment for known risk factors for GDM, women with twins are at an increased risk for GDM compared to women with singletons. Although the reasons for the increased risk of GDM in twins are unclear, the greater placental mass in twins [25, 26] is likely to play an important role. Previous studies have shown a positive relationship between placental mass and the risk of GDM [27, 28], most probably due to the higher levels of the diabetogenic hormone human placental lactogen (hPL) [29, 30]. However, based on our finding that the increased risk for GDM in twins is limited to diet-treated GDM, it seems that the degree of glucose intolerance in twins is only mildly impaired compared with singletons. It may be argued that this mild increase in glucose intolerance in twin pregnancies may be viewed as normal physiologic change inherent to twin pregnancies which may have a beneficial role (i.e., matching glucose availability to the increased demand for glucose in twin pregnancies) rather than a pathology that needs to be diagnosed and treated, similar to many of the other physiologic changes of pregnancy that are amplified in twin gestations . According to this argument, the apparent ‘increased risk’ of GDM in twin pregnancies is merely the result of the fact that we use the same thresholds for screening and diagnosis of GDM in twins and singletons despite the obvious physiological differences. Indeed, it has been shown that 50-g GCT has a higher false positive rate for GDM in twin compared with singleton gestations  and was suggested that different threshold should be used for screening for GDM in twin gestations . Additional support to this argument comes from the observation that women who experienced GDM in the presence of a twin pregnancy were significantly less likely to develop future type-2 diabetes compared with those who had GDM in a singleton pregnancy (HR 0.76, 95% CI 0.65–0.90) . In contrast, we have recently demonstrated that GDM and milder degrees of glucose intolerance are associated with an increased risk of asymmetric overgrowth in twin pregnancies, which may provide support to current screening and diagnosis criteria for GDM .
We found that the type and magnitude of risk factors for GDM are similar between twins and singletons. Only a few studies reported on risk factors for GDM in twins pregnancies. In concordance with our results, Lucovnik et al.  reported that high BMI was an independent risk factor for GDM in both twin and singleton gestations. Similarly, in agreement with our findings, Wang et al.  reported that assisted reproductive technology increases the likelihood of GDM in singletons (aOR = 1.26, 95% CI = 1.18–1.36), but not in twins. The reason for the discordance in association with fertility treatments with GDM between twins and singleton is unclear but we speculate that it may be related to differences in the distribution of the indications for fertility treatments in both groups, for example, differences in the proportion of women with ovulation dysfunction due to polycystic ovary syndrome (PCOS), and independent risk factor for GDM .
Strengths and limitations
The main strength of our study is the population-based nature of the study and the large sample size of both twin and singleton pregnancies, which made our study powered to detect differences between twins and singletons not only in the overall rate of GDM but also in the rate of subtypes of GDM (i.e., diet-treated vs. insulin-treated GDM). Another strength is the availability of data on important confounding variables which were not accounted for in some of the previous studies on this topic [3, 15]. Still, our study has several limitations. Due to its retrospective nature, information on several potentially confounding variables such as GDM in previous pregnancies, history of PCOS and family history of diabetes [35, 36] was unavailable for analysis. Another limitation is the potential selection bias given the fact that about 50% of the initial population was excluded due to missing data on pre-pregnancy BMI or ethnicity. Still, we believe that including those cases with missing information on BMI or ethnicity would result in an even greater bias due to the failure to adjust the analysis for these important confounding variables.
Our findings suggest that women with twins are at increased risk of GDM, mainly due to a higher rate of mild (diet-treated) GDM. The similarity in risk factors for GDM in twin and singleton pregnancies implies that GDM represents the same disease in both groups and that despite the higher baseline risk of GDM in twins, the effect of known risk factors for GDM in twin pregnancies is similar to that observed in singletons. Care providers should consider twin pregnancy as an independent risk factor for GDM when assessing patients risk for GDM, and studies on GDM should adjust for twin gestation as an independent risk factor for GDM.
Chauhan SP, Scardo JA, Hayes E, Abuhamad AZ, Berghella V (2010) Twins: prevalence, problems, and preterm births. Am J Obstet Gynecol 203:305–315
Rauh-Hain JA, Rana S, Tamez H, Wang A, Cohen B, Cohen A, Brown F, Ecker JL, Karumanchi SA, Thadhani R (2009) Risk for developing gestational diabetes in women with twin pregnancies. J Matern Fetal Neonatal Med 22:293–299
Lai FY, Johnson JA, Dover D, Kaul P (2016) Outcomes of singleton and twin pregnancies complicated by pre-existing diabetes and gestational diabetes: a population-based study in Alberta, Canada, 2005–11. J Diabetes 8:45–55
Wang YA, Nikravan R, Smith HC, Sullivan EA (2013) Higher prevalence of gestational diabetes mellitus following assisted reproduction technology treatment. Hum Reprod 28:2554–2561
Retnakaran R, Shah BR (2016) Impact of twin gestation and fetal sex on maternal risk of diabetes during and after pregnancy. Diabetes Care 39:e110–e111
Yogev Y, Eisner M, Hiersch L, Hod M, Wiznitzer A, Melamed N (2014) The performance of the screening test for gestational diabetes in twin versus singleton pregnancies. J Matern Fetal Neonatal Med 27:57–61
Spellacy WN, Handler A, Ferre CD (1990) A case-control study of 1253 twin pregnancies from a 1982–1987 perinatal data base. Obstet Gynecol 75:168–171
Buhling KJ, Henrich W, Starr E, Lubke M, Bertram S, Siebert G, Dudenhausen JW (2003) Risk for gestational diabetes and hypertension for women with twin pregnancy compared to singleton pregnancy. Arch Gynecol Obstet 269:33–36
Henderson CE, Scarpelli S, LaRosa D, Divon MY (1995) Assessing the risk of gestational diabetes in twin gestation. J Natl Med Assoc 87:757–758
Simoes T, Queiros A, Correia L, Rocha T, Dias E, Blickstein I (2011) Gestational diabetes mellitus complicating twin pregnancies. J Perinat Med 39:437–440
Committee Canadian Diabetes Association Clinical Practice Guidelines Expert (2008) Canadian diabetes association 2008 clinical practice guidelines for the prevention and management of diabetes in Canada. Can J Diabetes 32(Suppl 1):S1–S201
Shinagawa S, Suzuki S, Chihara H, Otsubo Y, Takeshita T, Araki T (2005) Maternal basal metabolic rate in twin pregnancy. Gynecol Obstet Invest 60:145–148
Marconi AM, Davoli E, Cetin I, Lanfranchi A, Zerbe G, Fanelli R, Fennessey PV, Pardi G, Battaglia FC (1993) Impact of conceptus mass on glucose disposal rate in pregnant women. Am J Physiol 264:E514–E518
Schwartz DB, Daoud Y, Zazula P, Goyert G, Bronsteen R, Wright D, Copes J (1999) Gestational diabetes mellitus: metabolic and blood glucose parameters in singleton versus twin pregnancies. Am J Obstet Gynecol 181:912–914
Luo ZC, Simonet F, Wei SQ, Xu H, Rey E, Fraser WD (2011) Diabetes in pregnancy may differentially affect neonatal outcomes for twins and singletons. Diabet Med 28:1068–1073
Foeller ME, Zhao S, Szabo A, Cruz MO (2015) Neonatal outcomes in twin pregnancies complicated by gestational diabetes compared with non-diabetic twins. J Perinatol 35:1043–1047
Guillen MA, Herranz L, Barquiel B, Hillman N, Burgos MA, Pallardo LF (2014) Influence of gestational diabetes mellitus on neonatal weight outcome in twin pregnancies. Diabet Med 31:1651–1656
Canadian Diabetes Association Clinical Practice Guidelines Expert Committee (2013) Canadian diabetes association 2013 clinical practice guidelines for the prevention and management of diabetes in Canada. Can J Diabetes 37(Suppl 1):S1–S212
Tward C, Barrett J, Berger H, Kibel M, Pittini A, Halperin I, Cohen H, Melamed N (2016) Does gestational diabetes affect fetal growth and pregnancy outcome in twin pregnancies? Am J Obstet Gynecol 214(653):e651–e658
Okby R, Weintraub AY, Sergienko R, Eyal S (2014) Gestational diabetes mellitus in twin pregnancies is not associated with adverse perinatal outcomes. Arch Gynecol Obstet 290:649–654
Moses RG, Webb AJ, Lucas EM, Davis WS (2003) Twin pregnancy outcomes for women with gestational diabetes mellitus compared with glucose tolerant women. Aust N Z J Obstet Gynaecol 43:38–40
Fox NS, Gerber RS, Saltzman DH, Gupta S, Fishman AY, Klauser CK, Rebarber A (2016) Glycemic control in twin pregnancies with gestational diabetes: are we improving or worsening outcomes? J Matern Fetal Neonatal Med 29:1041–1045
Morikawa M, Yamada T, Akaishi R, Kojima T, Nishida R, Koyama T, Cho K, Minakami H (2015) Prevalence of hyperglycaemia in singleton versus twin pregnancy. Diabetes Metab Res Rev 31:198–203
Wein P, Warwick MM, Beischer NA (1992) Gestational diabetes in twin pregnancy: prevalence and long-term implications. Aust N Z J Obstet Gynaecol 32:325–327
Wegrzyn P, Fabio C, Peralta A, Faro C, Borenstein M, Nicolaides KH (2006) Placental volume in twin and triplet pregnancies measured by three-dimensional ultrasound at 11 + 0 to 13 + 6 weeks of gestation. Ultrasound Obstet Gynecol 27:647–651
Almog B, Shehata F, Aljabri S, Levin I, Shalom-Paz E, Shrim A (2011) Placenta weight percentile curves for singleton and twins deliveries. Placenta 32:58–62
Taricco E, Radaelli T, Nobile de Santis MS, Cetin I (2003) Foetal and placental weights in relation to maternal characteristics in gestational diabetes. Placenta 24:343–347
Daskalakis G, Marinopoulos S, Krielesi V, Papapanagiotou A, Papantoniou N, Mesogitis S, Antsaklis A (2008) Placental pathology in women with gestational diabetes. Acta Obstet Gynecol Scand 87:403–407
Kazer RR, Cheng ER, Unterman TG, Glick RP (1991) Maternal plasma concentrations of insulin-like growth factor-I (IGF-I) and human placental lactogen (hPL) in twin pregnancies. Acta Genet Med Gemellol 40:383–387
Thiery M, Dhont M, Vandekerckhove D (1977) Serum HCG and HPL in twin pregnancies. Acta Obstet Gynecol Scand 56:495–497
Norwitz ER, Edusa V, Park JS (2005) Maternal physiology and complications of multiple pregnancy. Semin Perinatol 29:338–348
Rebarber A, Dolin C, Fields JC, Saltzman DH, Klauser CK, Gupta S, Fox NS (2014) Screening approach for gestational diabetes in twin pregnancies. Am J Obstet Gynecol 211(639):e631–e635
Lucovnik M, Blickstein I, Verdenik I, Steblovnik L, Trojner Bregar A, Tul N (2014) Impact of pre-gravid body mass index and body mass index change on preeclampsia and gestational diabetes in singleton and twin pregnancies. J Matern Fetal Neonatal Med 27:1901–1904
de Wilde MA, Veltman-Verhulst SM, Goverde AJ, Lambalk CB, Laven JS, Franx A, Koster MP, Eijkemans MJ, Fauser BC (2014) Preconception predictors of gestational diabetes: a multicentre prospective cohort study on the predominant complication of pregnancy in polycystic ovary syndrome. Hum Reprod 29:1327–1336
Kim C, Berger DK, Chamany S (2007) Recurrence of gestational diabetes mellitus: a systematic review. Diabetes Care 30:1314–1319
Kruse AR, Darling MS, Hansen MK, Markman MJ, Lauszus FF, Wielandt HB (2015) Recurrence of gestational diabetes in primiparous women. Acta Obstet Gynecol Scand 94:1367–1372
This study was funded by Canadian Institute of Health Research (CIHR) (Grant# 146442; Non-communicable Diseases in Obstetrics: Improving Quality of Care and Maternal-infant Outcomes Through an Obstetrical Research Network). Dr. Sarah D. McDonald is supported by a Tier II Canada Research Chair. Dr. Beth Murray-Davis is supported by a Hamilton Health Sciences Early Career Award. None of the funding agencies had any role in the idea, design, analyses, interpretation of data, writing of the manuscript or decision to submit the manuscript.
Conflict of interest
The authors declare that they have no conflict of interest.
About this article
Cite this article
Hiersch, L., Berger, H., Okby, R. et al. Incidence and risk factors for gestational diabetes mellitus in twin versus singleton pregnancies. Arch Gynecol Obstet 298, 579–587 (2018). https://doi.org/10.1007/s00404-018-4847-9
- Gestational diabetes mellitus
- Risk factors