Effect of pregnancy on insulin requirements differs between type 1 and type 2 diabetes: A cohort study of 222 pregnancies

Transcription

1 Australian and New Zealand Journal of Obstetrics and Gynaecology 2016; 56: DOI: /ajo Original Article Effect of pregnancy on insulin requirements differs between type 1 and type 2 diabetes: A cohort study of 222 pregnancies Suja PADMANABHAN, 1,2 Shan JIANG, 1,2 Mark MCLEAN 1,3,4 and N. Wah CHEUNG 1,2 1 Department of Diabetes and Endocrinology, Westmead Hospital, 2 School of Medicine, University of Sydney, New South Wales, 3 Department of Diabetes and Endocrinology, Blacktown Hospital, Blacktown, New South Wales, and 4 School of Medicine, University of Western Sydney, New South Wales, Australia Background: Knowledge about expected insulin requirements during pregnancy, in women with pre-existing diabetes may assist clinicians to effectively respond to gestation-specific changes in glycemic pattern. Few studies have examined differences between type 1 (T1DM) and type 2 diabetes (T2DM). Aims: To compare patterns of insulin requirements in pregnancy for women with pre-existing T1DM and T2DM. Material and Methods: A retrospective cohort study of 222 pregnancies was conducted in women with pre-existing diabetes, (67 with T1DM, 155 with T2DM). Total daily insulin dose (TID) at the end of each trimester, recorded as units and units per kilogram (median, 25th 75th percentile) as well as percentage increase in insulin dose per trimester were compared. Results: Women with T1DM had higher insulin requirements in the first two trimesters than those with T2DM (0.69 ( ) vs 0.36 ( ) units/kg in first trimester; 0.80 ( ) vs 0.61 ( ) units/kg, P < 0.005) in second trimester), but requirements in late pregnancy were similar (0.97 ( ) vs 0.95 ( ) units/kg, P = 0.54). Women with T2DM needed much greater increases in insulin per trimester compared to T1DM (P < 0.001). Women with T1DM had a net fall in insulin requirements (3.7% in the first trimester and 4.1% in the late third trimester) while those with T2DM did not. Conclusions: This is the largest comparison study of insulin requirements in women with pre-existing diabetes, highlighting important trimester-specific differences between T1DM and T2DM to guide insulin titration during pregnancy. Our findings suggest a differential effect of pregnancy-mediated insulin resistance by type of diabetes. Key words: insulin requirements, pregnancy, type 1 diabetes, type 2 diabetes. Introduction The incidence of type 1 (T1DM) and type 2 diabetes (T2DM) in pregnancy is increasing worldwide. 1 Insulin remains the gold standard for treatment during pregnancy; however, dosages require frequent and significant adjustment in response to physiological changes. 2 A number of studies have examined insulin requirements in women with T1DM, depicting a tri-phasic pattern of insulin requirements. 3 7 There is a slight fall in requirements in late first trimester, followed by a Correspondence: Dr Suja Padmanabhan, Department of Diabetes and Endocrinology, Level 2 ICPMR Westmead Hospital, Corner Darcy and Hawkesbury Road, Westmead, NSW 2147, Australia. suja_padman@yahoo.com.au The authors report no conflict of interest. Received 13 October 2015; accepted 9 January progressive increase to 36 weeks of gestation and a slight fall again at term. However, few studies have examined insulin requirements in women with T2DM, and only one study, conducted over 20 years ago, has directly compared requirements between the two groups. 8 T2DM is now more common than T1DM in women of childbearing age and is associated with similar or greater rates of pregnancy complications. 9 Furthermore, since the underlying pathogenesis varies between T1DM and T2DM outside of pregnancy, it is plausible that the physiological changes in glucose metabolism during pregnancy may also vary. Knowledge of the differences in insulin requirements between these two groups may have important clinical significance assisting the clinician to effectively respond to gestation-specific changes in glycemic patterns, avoiding hypoglycaemia and identifying subgroups that may benefit from additional intervention. Considering the paucity of data, we conducted a retrospective review to investigate the patterns of insulin requirements in women with pre-existing diabetes during pregnancy The Royal Australian and New Zealand College of Obstetricians and Gynaecologists The Australian and New Zealand Journal of Obstetrics and Gynaecology

2 Insulin requirements in pregnancy Material and Methods We conducted a retrospective cohort study of women delivering between 1 January 2010 and 1 January 2015, who were managed in the diabetes in pregnancy clinic of two tertiary referral hospitals in Sydney, Australia. The study was approved by the district ethics committee. All women with a diagnosis of pre-existing T1DM or T2DM requiring insulin treatment during pregnancy were included. Women with T2DM were further subdivided into those who required insulin prior to pregnancy and those who did not. Women with incomplete follow up, miscarriage prior to 20 weeks and multiple gestation were excluded. Both institutions followed similar clinical practice with women reviewed at least every 2 4 weeks until 28 weeks gestation and every 1 2 weeks thereafter until delivery. Women recorded a minimum of fasting and three postprandial glucose measurements daily in a glucose diary and glucose meters were checked regularly at clinic visits to assess accuracy. Insulin dose was titrated to the same blood glucose target of 5.5 mmol (99 mg/dl) fasting and 7 mmol (126 mg/dl) 2 h postprandial, at both hospitals. The standard insulin regimen was basal bolus, with rapid acting insulin before meals and basal insulin at bedtime. Women with T1DM treated with an insulin pump prepregnancy were continued on this treatment. Information on patient demographics, history of diabetes, weight in kilograms, insulin dose, presence of hypoglycaemia (defined as documented blood sugar level <4.0 mmol or a reduction in insulin dose due to presumed hypoglycaemia) and HbA1C were obtained from the patient notes or electronic medical record for each clinic review. Daily insulin requirement at each review was recorded as total insulin dose in units (TID), including basal and prandial proportions (percentage) and was corrected for body weight as units per kilogram (units/kg). The primary outcome was: (i) insulin dose at the end of each trimester (trimester 1 (12 14 weeks), trimester 2 (24 26 weeks) and trimester 3 (delivery)); (ii) the percentage increase in insulin requirement per trimester, calculated using the formula: (Trimester end TID Trimester start TID)/Trimester Start TID 9100 (only women on insulin at the start of each trimester were included in this analysis); and (iii) the percentage decrease in insulin requirements from the peak to delivery, calculated using the formula: (Peak TID Delivery TID)/ Peak TID Secondary outcomes included the presence of one or more episodes of hypoglycaemia in each trimester, weight gain per trimester and HbA1c obtained at the end of each trimester. Statistical analysis comparing women with T1DM and T2DM was conducted using SPSS version 20. As many of the continuous variables were not normally distributed, the Mann Whitney U test was used to determine differences between groups. Pearson v 2 or Fisher exact tests were used to test for association between categorical variables. Two-tailed tests with a significance level of 5% were used throughout. Results We reviewed 277 pregnancies, in women with T1DM and T2DM, delivering over 5 years. After exclusion of women with multiple gestation (n = 8), pregnancy loss prior to 20 weeks (n = 12), incomplete follow up (n = 23) and those who did not need insulin treatment (n = 12), 222 pregnancies in 200 women were included in the study, 67 with T1DM and 155 with T2DM. Exclusion of subsequent pregnancies in women having more than one pregnancy during the study period, did not alter results; therefore all pregnancies have been included in the analysis. Baseline characteristics and pregnancy outcomes are summarised in Table 1. Women with T1DM were younger, leaner, had a longer duration of diabetes with a higher rate of complications and a higher prepregnancy HbA1c compared to those with T2DM. Almost 80% of women with T2DM were of non-caucasian background, while over 60% of women with T1DM were Caucasian. All women with T1DM were treated with insulin prepregnancy compared to 21.9% of women with T2DM. This subgroup had a longer duration of diabetes, (5 (4 16) vs 3 (1 9.5) years, P < 0.001), but were otherwise similar to the remaining T2DM cohort. Of those not on insulin prior to conception, 68.6% commenced insulin by the end of trimester 1, 92.6% by trimester 2 and 100% by trimester 3. Ten women with T1DM were treated with an insulin pump. The pattern of insulin requirements during pregnancy is summarised in Figure 1 and Table S1. While women with T1DM required more insulin (units/kg) in the first two trimesters than women with T2DM, both groups had similar requirements in late pregnancy (Fig. 1a,c). When analysis of women with T2DM was restricted to those requiring insulin before pregnancy, we found they had significantly higher TID throughout pregnancy compared to the other groups (Fig. 1b), although when insulin dose was corrected for weight, this remained significant in the third trimester only (Fig. 1d). Figure 2 shows the percentage change in insulin requirements from the start to end of each trimester. In trimester 1, women with T1DM needed a 3.7 (0 14.9)% reduction in insulin dose. In contrast, women with T2DM needed a 15.4 (0 40.3)% increase in dose. In trimesters 2 and 3, both groups needed an increase in dose with the percentage rise increasing progressively with advancing gestation (T1DM: 18.2 ( )% and 22.9 ( )%, T2DM: 35.7 ( )% and 44 ( )% for trimesters 2 and 3, respectively). Women with T2DM needed a significantly greater increase in dose per trimester compared to women with T1DM (P < 0.001). This remained significant after correcting for prepregnancy BMI (P 0.005). We also analysed whether there was a drop in TID from the peak reached in the third trimester, prior to delivery. Only in women with T1DM, there was a modest decrease in insulin dose prior to delivery; while those with T2DM did not require such a dose reduction (4.1 (0 18.1)% reduction vs. 0 (0 10.9)% reduction, P = 0.018) The Royal Australian and New Zealand College of Obstetricians and Gynaecologists 353

3 S. Padmanabhan et al. Table 1 Baseline characteristics and pregnancy outcomes Type 1 diabetes N = 67 Type 2 diabetes N = 155 P value* Baseline Characteristics Age (years) 29 (22 34) 34 (31 37) <0.001 Prepregnancy BMI (kg/m 2 ) 25.6 ( ) 31.1 ( ) <0.001 Ethnicity Caucasian 36 (53.7) 33 (21.3) <0.001 South Asian 12 (17.9) 41 (26.5) East and Southeast Asian 5 (7.5) 26 (16.8) Arabic 8 (11.9) 23 (14.8) Pacific Islands 3 (4.5) 12 (7.7) Other 3 (4.5) 20 (12.9) Duration of DM (years) 10 (7 18) 3 (2 6) <0.001 Pre pregnancy HbA1c % 8.5 ( ) 7.3 ( ) mmol/mol 69 (54 85) 85 (45 83) Any complication 23 (34.3) 13 (8.4) <0.001 Retinopathy 12 (17.9) 2 (1.3) <0.001 Proteinuria 14 (20.9) 12 (7.7) Hypertension 9 (13.4) 37 (23.9) Smoker 10 (14.9) 15 (9.7) Gestation at first visit (weeks) 9 (7 14) 11 (7 15) Multiparous 40 (59.7) 115 (74.3) Pregnancy outcomes Weight gain (kg) Trimester 1 (n = 205) 3.5 ( ) 1.7 ( ) Trimester 2 (n = 206) 5.3 ( ) 3.9 ( ) Trimester 3 (n = 203) 6.6 ( ) 5.1 ( ) 0.03 Total (n = 201) 15.9 ( ) 11.8 ( ) <0.001 HbA1c (%) Trimester 1 n = ( ) 6.3 ( ) <0.001 mmol/mol 58 (49 70) 45 (39 57) Trimester 2 n = ( ) 5.7 ( ) <0.001 mmol/mol 51 (44 55) 39 (36 44) Trimester 3 n = ( ) 6.1 ( ) <0.001 mmol/mol 52 (49 58) 43 (38 49) Hypoglycemic events Trimester 1 42 (62.7) 17 (11.0) <0.001 Trimester 2 23 (34.3) 25 (16.1) Trimester 3 42 (62.7) 60 (38.7) Metformin treatment during pregnancy 5 (7.5) 38 (24.5) *P value comparing T1DM and T2DM by Mann Whitney U test, Pearson s v2 or Fisher s Exact test where applicable. Prepregnancy BMI calculated from recalled prepregnancy weight or weight at first visit. Data are median (interquartile range Q1-Q3) or n (%) where applicable. Differences in the prandial and basal components of the TID are summarised in Table S1. Throughout pregnancy, women with T2DM had a significantly greater proportion of their TID as prandial insulin compared to women with T1DM. However, when analysis was restricted to women requiring insulin prepregnancy, basal and prandial proportions of TID were similar between the groups. The risk of hypoglycaemia was greater in women with T1DM throughout pregnancy with the highest risk occurring in early pregnancy and again close to term. Women with T1DM also gained more weight in each trimester and therefore had greater total weight gain than women with T2DM, which persisted after adjusting for prepregnancy BMI. They also had a higher HbA1c throughout pregnancy, although not all women in the cohort had HbA1c tested in each trimester (Table 1). Total weight gain weakly correlated with the TID at term (R 2 = 0.173, P < 0.001) The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

4 Insulin requirements in pregnancy Figure 1 Total daily insulin requirements during pregnancy in type 1 and type 2 diabetes. Comparison of total daily insulin requirement measured in (units) (panel a and b) and (units/kg) (panel c and d) in type 1 and type 2 diabetes (a) and (c) and type 1 and type 2 diabetes on insulin treatment prepregnancy (b) and (d). Markers represent medians; error bars represent 25th and 75th centiles. *P < Figure 2 Percentage change in insulin requirements from the start to end of each trimester in women with type 1 and type 2 diabetes. Bars represent median and interquartile range (25th 75th centile). Error bars represent 5th and 95th centiles. P value for comparison between type 1 and type 2 diabetes in each trimester. The 43 women who were treated with metformin in addition to insulin during pregnancy did not have any differences in insulin requirements or weight gain compared to those who were treated with insulin alone (data not shown). Discussion Most previous studies have focussed on women with T1DM, but given the rising prevalence of T2DM in pregnancy, it is particularly important to develop an understanding of the therapeutic differences between the two forms of diabetes. In the largest comparison of women with pre-existing diabetes reported to date, we found that women with T1DM had higher insulin requirements up to the second trimester of pregnancy, compared to women with T2DM. However, by the end of the third trimester, insulin requirements were similar. The pattern of insulin requirements also varied between the groups, suggesting a differential effect of pregnancymediated insulin resistance. We observed that women with T1DM needed a net reduction in insulin dose in early and late pregnancy and approximately a 20% increase in each of the second and third trimesters. In contrast, women with T2DM required a much greater increase in insulin dose from the start to the end of each trimester, with the percentage change progressively increasing with advancing gestation. Insulin requirements did not fall in early and late pregnancy as they did in women with T1DM. Only one previous study has compared insulin requirements between T1DM and T2DM. 8 In this study, 63 women with T1DM and 40 women with T2DM had insulin requirements recorded every three weeks. In contrast to our findings, their cohort of women with T2DM had consistently higher TID throughout pregnancy, compared to women with T1DM, and values around 25 50% higher than we have reported. However, the majority of women with T2DM in that study were treated with insulin prepregnancy, compared to only 22% in our study. On subgroup analysis of women with 2016 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists 355

5 S. Padmanabhan et al. T2DM treated with insulin prepregnancy, we found similar results. The previous study by Langer was conducted over 20 years ago utilising regular insulin, rather than insulin analogues, for prandial treatment and tighter glycaemic targets ( mmol/l) compared to our study. This could explain the much higher insulin dose required by their T2DM group. The reduction in insulin requirement in early pregnancy, in women with T1DM, is consistent with a number of previous studies which show requirements drop by 9 12% starting as early as 7 weeks and extending up to 16 weeks. 3,5,6,10,11 This has been replicated in two recent studies examining women on insulin pumps, which found significant reductions in basal infusion rates at 8 9 weeks of gestation. 5,12 Similarly, in late pregnancy, we observed a 4.1% reduction in dose from the peak TID to delivery. A small fall in insulin requirements at term has been previously described 7,13 and does not appear to be associated with the adverse obstetric outcomes that a larger fall in requirements may signify. 14 Early and late pregnancy also corresponded to the highest risk period for hypoglycaemia in our cohort of T1DM, supporting the need for increased surveillance and greater caution with insulin dose titration during this time. Women with T2DM did not display the same insulin sensitising phenomenon in early and late pregnancy as T1DM, and required almost double the increase in dose from second trimester onwards. The first trimester fall in insulin requirements is thought to be due to a transient reduction in progesterone levels as hormonal production shifts from the corpus luteum to the placenta, as well as decreased prandial requirements due to hyperemesis. 4 Nielson described a pregnancy-induced increase in C peptide concentration in women with T1DM with advancing gestation, suggesting that there are improvements in residual beta cell function and endogenous insulin production during pregnancy. 15 If true, this could potentially explain the late third trimester fall in insulin requirements we observed. Although C peptide levels and changes in endogenous insulin production have not been studied in women with T2DM during pregnancy, it is likely that any pregnancy-mediated improvement in beta cell function is offset by greater increases in insulin resistance. Similar to Langer s study, 8 our findings suggest that women with T1DM and T2DM both manifest a rise in insulin resistance with advancing gestation; however, they do so to differing degrees. In T1DM, hyperinsulinemic euglycemic clamp techniques confirm a 50 60% decrease in insulin sensitivity at weeks of gestation; 16 however, this has not been well studied in women with T2DM. Outside of pregnancy, it is known that women with T2DM have greater insulin resistance than T1DM, so it is possible that the magnitude of physiological insulin resistance in pregnancy is also greater in women with T2DM. Human and animal studies have shown that the placental production of hormones, including human placental lactogen, progesterone, cortisol and placental growth hormone, are responsible for the increase in insulin resistance in the second half of pregnancy through a postreceptor effect Whether women with T1DM and T2DM produce different amounts of placental hormones leading to these differences in insulin requirements is a topic for future research. BMI and weight gain may also contribute to the differences in insulin requirements observed in our study. Women with T2DM had significantly higher BMI at conception; however, the greater increases in insulin dose required per trimester persisted after correction for this. We also found an association between total weight gain in pregnancy and final insulin dose. Increasing adiposity is associated with greater production of pro-inflammatory cytokines and adipokines, such as tumor necrosis factor alpha and leptin, which in turn correlates with changes in insulin sensitivity. 17,22,23 Although women with T2DM gained less weight in pregnancy overall, this weight gain may have a greater metabolic impact, as their baseline BMI was higher, thus accounting for the differential degrees of insulin resistance during pregnancy. Furthermore, we found that women with T2DM had a greater proportion of their TID as prandial insulin compared to T1DM, although this was primarily in women who were NOT on insulin prepregnancy. In nondiabetic women, the increase in insulin resistance with advancing gestation is almost twofold greater than increases in basal hepatic glucose production. 24 It is likely that women with T2DM, who do not need insulin preconception, have better endogenous reserves accounting for lower requirements in the fasting state. However, after a glucose load, compensation to overcome the increased insulin resistance is inadequate, resulting in relatively higher prandial requirements. Alternatively, higher carbohydrate intake could potentially explain the difference in prandial needs in our study. We acknowledge our study has some limitations. The data were collected retrospectively; therefore, we relied on documentation of compliance with therapy, insulin doses, metformin therapy, weight measurement, hypoglycaemic events and regular HbA1c testing by the treating clinician, resulting in some missing data. Furthermore as the titration of insulin was not based on a predetermined protocol, individual clinical practice may have varied between treating doctors. However, as all patients were treated to the same glycaemic targets, this is unlikely to significantly bias the results. Finally, we were unable to quantify caloric intake, in particular carbohydrate intake, which could have influenced prandial insulin requirements. Nevertheless, this is the largest comparison study of insulin requirements in women with pre-existing diabetes, including for the first time, a large number of women with T2DM. As the prevalence of T2DM increases, our study provides important insights to guide the clinician on the expected changes in insulin requirements by type of diabetes, to help anticipate and respond to gestationspecific changes in glycemia The Royal Australian and New Zealand College of Obstetricians and Gynaecologists

6 Insulin requirements in pregnancy References 1 Guariguata L, Linnenkamp U, Beagley J et al. Global estimates of the prevalence of hyperglycaemia in pregnancy. Diabetes Res Clin Pract 2013; 103: McElduff A, Cheung NW, McIntyre HD et al. The Australasian Diabetes in Pregnancy Society consensus guidelines for the management of type 1 and type 2 diabetes in relation to pregnancy. Med J Aust 2005; 183: Garcıa-Patterson A, Gich I, Amini SB et al. Insulin requirements throughout pregnancy in women with type 1 diabetes mellitus: three changes of direction. Diabetologia 2010; 53: Lois J, Joe LS, Robert HK et al. Declining insulin requirement in the late first trimester of diabetic pregnancy. Diabetes Care 2001; 24: Roeder HA, Moore TR, Ramos GA. Insulin pump dosing across gestation in women with well-controlled type 1 diabetes mellitus. Am J Obstet Gynecol 2012; 207: Steel JM, Johnstone FD, Hume R, Mao JH. Insulin requirements during pregnancy in women with type I diabetes. Obstet Gynecol 1994; 83: McManus RM, Ryan EA. Insulin requirements in insulindependent and Insulin-requiring GDM women during final month of pregnancy. Diabetes Care 1992; 15: Langer O, Anyaegbunam A, Brustman L et al. Pregestational diabetes: insulin requirements throughout pregnancy. Am J Obstet Gynecol 1988; 159: McElduff A, Cheung NW, Ross GP et al. Pregestational diabetes and pregnancy: an Australian experience. Diabetes Care 2005; 28: Rayburn W, Piehl E, Lewis E et al. Changes in insulin therapy during pregnancy. Am J Perinatol 1985; 2: Weiss PA, Hofmann H. Intensified conventional insulin therapy for the pregnant diabetic patient. Obstet Gynecol 1984; 64: Mathiesen JM, Secher AL, Ringholm L et al. Changes in basal rates and bolus calculator settings in insulin pumps during pregnancy in women with type 1 diabetes. J Matern Fetal Neonatal Med 2014; 27: Fuglsang J, Lauszus F, Flyvbjerg A, Ovesen P. Human placental growth hormone, insulin-like growth factor I and -II, and insulin requirements during pregnancy in type 1 diabetes. J Clin Endocrinol Metab 2003; 88: Padmanabhan S, McLean M, Cheung NW. Falling insulin requirements are associated with adverse obstetric outcomes in women with preexisting diabetes. Diabetes Care 2014; 37: Nielsen LR, Rehfeld JF, Pedersen-Bjergaard U et al. Pregnancy-induced rise in serum C-peptide concentrations in women with type 1 diabetes. Diabetes Care 2009s; 32: Schmitz O, Klebe J, Møller J et al. In vivo insulin action in type 1 (insulin-dependent) diabetic pregnant women as assessed by the insulin clamp technique. J Clin Endocrinol Metab 1985; 61: Barbour LA, McCurdy CE, Hernandez TL et al. Cellular mechanisms for insulin resistance in normal pregnancy and gestational diabetes. Diabetes Care 2007; 30(Suppl 2): S112 S Newbern D, Freemark M. Placental hormones and the control of maternal metabolism and fetal growth. Curr Opin Endocrinol Diabetes Obes 2011; 18: Ryan EA, Enns L. Role of gestational hormones in the induction of insulin resistance. J Clin Endocrinol Metab 1988; 67: Spellacy WN, Cohn JE. Human placental lactogen levels and daily insulin requirements in patients with diabetes mellitus complicating pregnancy. Obst Gynecol 1973; 42: Stewart MO, Whittaker PG, Persson B et al. A longitudinal study of circulating progesterone, oestradiol, hcg and hpl during pregnancy in type 1 diabetic mothers. Br J Obstet Gynaecol 1989; 96: Kirwan JP, Hauguel-De Mouzon S, Lepercq J et al. TNFalpha is a predictor of insulin resistance in human pregnancy. Diabetes 2002; 51: McIntyre HD, Chang AM, Callaway LK et al. Hormonal and metabolic factors associated with variations in insulin sensitivity in human pregnancy. Diabetes Care 2010; 33: Catalano PM, Tyzbir ED, Wolfe RR et al. Longitudinal changes in basal hepatic glucose production and suppression during insulin infusion in normal pregnant women. Am J Obstet Gynecol 1992; 1: Supporting Information Additional Supporting Information may be found in the online version of this article: Table S1. Insulin requirements during pregnancy The Royal Australian and New Zealand College of Obstetricians and Gynaecologists 357