Vitamin D insufficiency, preterm delivery and preeclampsia in women with type 1 diabetes an observational study

Transcription

1 AOGS ORIGINAL RESEARCH ARTICLE Vitamin D insufficiency, preterm delivery and preeclampsia in women with type 1 diabetes an observational study MARIANNE VESTGAARD 1,2,3, ANNA L. SECHER 1,2, LENE RINGHOLM 1,4, JENS-ERIK B. JENSEN 5, PETER DAMM 1,3,6 & ELISABETH R. MATHIESEN 1,2,3 1 Center for Pregnant Women with Diabetes, Rigshospitalet, University of Copenhagen, Copenhagen, 2 Department of Endocrinology, Rigshospitalet, University of Copenhagen, Copenhagen, 3 Institute of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Copenhagen, 4 Steno Diabetes Center Copenhagen, Gentofte, 5 Department of Endocrinology, Hvidovre Hospital, Hvidovre, and 6 Department of Obstetrics, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark Key words Preeclampsia, pregnancy, preterm delivery, type 1 diabetes, vitamin D Correspondence Marianne Vestgaard, Center for Pregnant Women with Diabetes, Department of Endocrinology, Rigshospitalet, University of Copenhagen, Blegdamsvej 9, 2100 Copenhagen, Denmark. marianne.jenlev.vestgaard@regionh.dk Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article. Please cite this article as: Vestgaard M, Secher AL, Ringholm L, Jensen J-EB, Damm P, Mathiesen ER. Vitamin D insufficiency, preterm delivery and preeclampsia in women with type 1 diabetes an observational study. Acta Obstet Gynecol Scand 2017; 96: Received: 12 November 2016 Accepted: 25 May 2017 DOI: /aogs Abstract Introduction. The aim of this study was to evaluate whether vitamin D insufficiency is associated with preterm delivery and preeclampsia in women with type 1 diabetes. Material and methods. An observational study of 198 pregnant women with type 1 diabetes. 25-Hydroxy-Vitamin D and HbA1c were measured in blood samples in early (median 8 weeks, range 5 14) and late (34 weeks, range 32 36) pregnancy. Kidney involvement (microalbuminuria or nephropathy) at inclusion, smoking status at inclusion, preterm delivery (<37 weeks) and preeclampsia (blood pressure 140/ 90 mmhg and proteinuria) were registered. Vitamin D supplementation of 10 lg daily was routinely recommended. Results. Thirty-nine (20%) of the 198 women delivered preterm and 16 (8%) developed preeclampsia. Vitamin D insufficiency (<50 nmol/l) was present in 68 women (34%) in early pregnancy and in 73 women (37%) in late pregnancy. Preterm delivery occurred more frequently in women with vitamin D insufficiency in late pregnancy (27% vs. 15%, crude odds ratio 2.1; 95% confidence interval , p = 0.04). After adjustment for preexisting kidney involvement, HbA1c in late pregnancy and smoking the association became nonsignificant (adjusted odds ratio 1.8; 95% confidence interval ). Preeclampsia developed in 11% of women with vitamin D insufficiency vs. 6% of the remaining women (crude odds ratio 1.8; 95% confidence interval , p = 0.25). Conclusion. In women with type 1 diabetes, preterm delivery was twice as frequent in women with vitamin D insufficiency in late pregnancy in crude analysis, but in this small study, low vitamin D was not independently associated with preterm birth or preeclampsia. Abbreviations: HbA1c, glycosylated hemoglobin; IL-6, interleukin-6; OR, odds ratio. Introduction Preterm delivery and preeclampsia are five times more frequent in women with type 1 diabetes compared with healthy women (1,2). Low vitamin D levels have been associated with preterm delivery in healthy women (3 5) Key Message In women with type 1 diabetes, preterm delivery was twice as frequent in women with vitamin D insufficiency in late pregnancy in crude analysis, but in this small study, low vitamin D was not independently associated with preterm birth or preeclampsia. ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)

2 Vitamin D in pregnancy with diabetes M. Vestgaard et al. and this seems to be the case for both spontaneous and medically indicated preterm delivery (4). Evidence points to vitamin D insufficiency increasing the risk of preeclampsia although not consistently (6 9), and vitamin D supplementation seems to reduce preeclampsia risk in healthy pregnancies (5). Approximately every fifth healthy pregnant woman in Denmark has vitamin D insufficiency and the prevalence might be even higher in other countries (10,11). One small study from the UK including 65 pregnant women with type 1 diabetes reported a very high (70 80%) incidence of vitamin D insufficiency (12). In pregnancy the prevalence of insufficient vitamin D levels appears to be increased (13). Preterm delivery has several etiologies and vitamin D insufficiency has been hypothesized to be relevant for both spontaneous as well as medically indicated preterm delivery (4). Vitamin D may play an active role in the gene regulation of the developing placenta (14) and in the regulation of the immune system by altering the production of cytokines such as interleukin-6 (IL-6) (3). Preeclampsia is characterized by hypertension and is a frequent cause of preterm delivery. Clinical studies have demonstrated an inverse relationship between vitamin D levels and the blood pressure outside pregnancy (15,16), a mechanism that has been linked to the renin angiotensin aldosterone system (16,17). Of interest is that vitamin D also seems to influence glucose metabolism (12,18), adding another aspect to the importance of investigating the association between vitamin D levels and adverse pregnancy outcomes in diabetes. The association between vitamin D levels and preeclampsia in type 1 diabetes has only been studied in one longitudinal study including 23 pregnant women with diabetes, who developed preeclampsia, pregnant women with diabetes who did not develop preeclampsia, and, for comparison, healthy pregnant women (14). A not significantly lower level of vitamin D throughout pregnancy was seen in the women with diabetes developing preeclampsia. Furthermore, vitamin D levels were generally lower in the women with type 1 diabetes compared with the healthy pregnant women (14). The aim of this study was to evaluate whether vitamin D insufficiency is associated with preterm delivery and preeclampsia in women with type 1 diabetes. Material and methods The current study consists of two study cohorts of Danish-speaking pregnant women with type 1 diabetes referred to the Center for Pregnant Women with Diabetes, Rigshospitalet University Hospital, before 14 completed weeks of gestation with a single living fetus, during the study periods of September 2004 to August 2006 (19) and February 2009 to February 2011 (20). The first cohort consists of 107 women (95% of eligible women) who had blood sampled as part of a prospective observational study evaluating the incidence and predictors of severe hypoglycemia in pregnancy (19,21). The second cohort consists of 123 women (79% of eligible women) who participated in a randomized controlled trial on the effect of continuous glucose monitoring on pregnancy outcome (20) whereof 19 women were excluded because they also figured in the first cohort (n = 15) or because they had spontaneous miscarriages (n = 4). In total, 211 women with type 1 diabetes were included (107 and 104 from the first and second cohorts, respectively). The women were included all year around, without significant seasonal variation. Twenty Danishspeaking healthy pregnant women referred to the midwifery consultation at Rigshospitalet in 2006/07 with a single living fetus were included in the present study for comparison (22). The routine clinical care of pregnant women with type 1 diabetes participating in the original study cohorts has previously been described in detail (19,20). Briefly, routine self-monitored plasma glucose was recommended seven times daily to obtain preprandial self-monitored plasma glucose of mmol/l, 90-min postprandial plasma glucose of mmol/l, prebedtime plasma glucose of mmol/l and glycosylated hemoglobin (HbA1c) 5.6% (38 mmol/mol) in the second part of pregnancy. The patients were mainly followed with clinical consultation at 2-week intervals throughout pregnancy. The presence of diabetic retinopathy at inclusion was routinely assessed by photo screening (23). Urinary albumin excretion was measured by 24-h urine collections or albumin-to-creatinine ratio in urine samples at inclusion. Based on at least two urine samples the women were classified as having normoalbuminuria (<30 mg/ 24 h or <30 mg/mmol), microalbuminuria ( mg/ 24 h or mg/mmol) or diabetic nephropathy ( 300 mg/24 h or 300 mg/mmol) (19,20). Diabetic nephropathy diagnosed before pregnancy was included in the first study population, but was an exclusion criterion in the second protocol (20). All women with type 1 diabetes were at inclusion asked about their educational level, which was classified in accordance with the International Standard Classification of Education (24) and converted to three educational levels: 10 years, years and 15 years of education. All the healthy women and most (96%) of the women with type 1 diabetes were of Caucasian origin. Oral vitamin D supplementation of 10 lg daily was recommended to all the women ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)

3 M. Vestgaard et al. Vitamin D in pregnancy with diabetes Blood samples were collected in early (median 8 weeks, range 5 14) and late (34 weeks, range 27 35) pregnancy and were available for analysis in 210 and 198 women in early and late pregnancy, respectively, among the 211 women included. Four women delivered before the planned blood sampling in late pregnancy. One woman in early and eight women in late pregnancy did not have blood samples collected for various other reasons. Among the healthy women blood was sampled at 15 (range 10 19) weeks. All blood samples were drawn with patients in the sitting position after rest. Blood samples were centrifuged at 3000 g for 10 min and stored as plasma and serum samples at 80 C until analysis. Plasma 25-hydroxy-vitamin D3+D2 was measured by competitive electrochemiluminescence-binding sassay by photon-counting on Cobas 8000 (Roche), which is standardized against the LC-MS/MS analysis. All vitamin D analyses were performed at the same time. Plasma parathyroid hormone was measured by sandwich electrochemiluminescence-binding assay by photon-counting. Plasma high-sensitivity C-reactive protein and plasma IL- 6 was measured by latex-particle-based immunoassay. The first and second cohorts were comparable regarding median vitamin D levels in early (60 in the first cohort vs. 62 nmol/l in the second cohort) and late (55 vs. 59 nmol/l) pregnancy and the prevalence of women with vitamin D levels <50 nmol/l in early (30% in the first cohort vs. 35% in the second cohort) and late (38% vs. 34%) pregnancy. The data from the two cohorts were therefore combined. Vitamin D insufficiency and deficiency were defined as 25-hydroxy-vitamin D <50 and <25 nmol/l, respectively (25). Kidney involvement at inclusion was defined as presence of microalbuminuria or diabetic nephropathy. Gestational age was estimated based on an early ultrasound scan. The total gestational weight gain (kg) was calculated as the difference between the last weight measured before delivery and the self-reported prepregnancy weight (26). Preeclampsia was defined as blood pressure 140/90 mmhg accompanied by proteinuria defined as 1+ on a sterile urinary dipstick or urinary proteinuria excretion 300 mg/24 h or albumin-to-creatinine ratio >190 mg/24 h (proteinuria of 300 mg/24 h approximates urinary albuminuria excretion of 190 mg/24 h) after 20 weeks (20,21,27). Preterm delivery was defined as delivery before 37 completed weeks of gestation. Largefor gestational-age and small-for-gestational-age were defined as a birthweight 90th or 10th centile, respectively, adjusted for gestational age and sex. Birthweight SD-score indicates how far (in standard deviations) from the mean Nordic population the infant birthweight deviates after adjustment for gender and gestational age (28). Statistical analyses Data are given as median (range) or numbers (%). Categorical variables were compared by chi-squared or Fisher s exact test, as appropriate. Continuous variables were analyzed by Mann Whitney U-test. Univariate and multivariate logistic regression analyses were conducted with preterm delivery (yes/no) or preeclampsia (yes/no) as the dependent variable and given as odds ratio (OR) and 95% CI. In the multivariate logistic regression with preterm delivery as the dependent variable, independent variables were: vitamin D insufficiency in late pregnancy (yes/no), kidney involvement at inclusion (yes/no), HbA1c (%) in late pregnancy (continuous variable) and smoking status at inclusion (yes/no). These independent variables were chosen based on significance in the univariate analyses (vitamin D insufficiency and HbA1c in late pregnancy) or of a priori significance (kidney involvement and smoking status). Due to low numbers of the dependent variable adjustment for possible confounders was not performed in the analyses concerning preeclampsia. A two-sided p-value <0.05 was regarded as statistically significant. All analyses were performed using IBM SPSS Statistics 22 (SPSS, Chicago, IL, USA). Ethical approval All women gave written informed consent and were followed until delivery. The research protocol for the current study was approved by The Regional Committee of ethics and science in the capital Region of Denmark (H , May 2016) and by the Danish Data Protection Agency ( , March 2004 and , February 2009). Results In women with type 1 diabetes, plasma vitamin D levels were 62 (11 142) nmol/l in early pregnancy and 57 (10 141) nmol/l in late pregnancy. Vitamin D insufficiency was present in 68 (32%) women in early pregnancy and in 73 (37%) women in late pregnancy. Approximately half (53%) of the women who had vitamin D insufficiency in late pregnancy also had vitamin D insufficiency in early pregnancy. In late pregnancy vitamin D deficiency (<25 nmol/l) was present in 17 (9%) women whereas a vitamin D level 75 nmol/l was present in 55 (28%) women. Preterm delivery occurred more frequently in women with vitamin D insufficiency than in women with vitamin D levels 50 nmol/l in late pregnancy (27% vs. 15%, p = 0.04). Vitamin D insufficiency in early pregnancy was ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)

4 Vitamin D in pregnancy with diabetes M. Vestgaard et al. not associated with preterm delivery. The rate of preterm delivery in women with vitamin D insufficiency in early pregnancy compared with the remaining women was 18% vs. 22% (p = 0.59). Women delivering preterm (n = 39, 20%) were characterized by higher HbA1c and lower vitamin D levels in late pregnancy compared with women delivering at term (Table 1). Preterm delivery was most often due to obstetrical indications for delivery, with only three spontaneous preterm deliveries, which were all due to preterm prelabor rupture of membranes. Among the remaining 36 preterm deliveries, most were due to pregnancy-induced hypertensive disorders and to a lesser extent due to the expectance of large fetuses based on ultrasound scans or signs of fetal distress. The prevalence of preterm delivery in women with levels of vitamin D >50 nmol/l, levels between 25 and 49 nmol/l as well as <25 nmol/l in late pregnancy are given in Figure 1. The highest prevalence of preterm delivery was seen in the women with vitamin D levels <25 nmol/l. Among the 19 women with kidney involvement in early pregnancy (6 with diabetic nephropathy and 13 with microalbuminuria), 32% (n = 6) had vitamin D deficiency, 63% (n = 12) had vitamin D insufficiency, and 32% (n = 6) delivered preterm. Univariate logistic regression analysis showed that women with vitamin D insufficiency in late pregnancy delivered preterm twice as often as women with vitamin D levels 50 nmol/l (crude OR 2.1; 95% CI ). After adjustment for kidney involvement at inclusion, HbA1c in late pregnancy and smoking, the adjusted OR for preterm delivery was 1.8 (95% CI ). HbA1c in late pregnancy was the only independent predictor of preterm delivery (OR 3.2; 95% CI ), indicating that the risk of preterm delivery is three-fold increased per 1% increase in HbA1c. Neither HbA1c in early nor in late pregnancy was associated with vitamin D levels in late pregnancy (data not shown). No differences in the development of preeclampsia were seen between the women with vitamin D insufficiency in early and late pregnancy, respectively, and the remaining women (8% vs. 8%, p = 0.93 and 11% vs. 6%, p = 0.25). Univariate logistic regression analysis showed a crude OR of 1.8 (95% CI ) for developing Table 1. Clinical data on 198 pregnant women with type 1 diabetes with preterm delivery or delivery at term. Women with preterm delivery (n = 39) Women with delivery at term (n = 156) p-value Maternal age (years) 31 (19 39) 30 (21 43) 0.71 Duration of diabetes (years) 15.0 ( ) 15.0 ( ) 0.30 Pre-pregnancy BMI (kg/m 2 ) 24.3 ( ) 24.7 ( ) 0.52 Smoking 7 (18) 18 (12) 0.28 Educational level a 10 years 6 (15) 15 (10) years 14 (36) 52 (33) 15 years 18 (46) 80 (51) 0.52 Elevated urinary albumin excretion ( 30 mg/24 h or 7 (18) 13 (8) mg/mmol) Nulliparous 26 (67) 88 (56) 0.2 Systolic/diastolic blood pressure (mmhg) Early pregnancy 121 ( )/71 (56 92) 120 (88 150)/70 (58 86) 0.60/0.91 Late pregnancy 124 (96 164)/76 (61 89) 120 (96 148)/76 (44 95) 0.52/0.59 HbA1c (%) Pregestational 7.0 ( ) 7.0 ( ) 0.98 Early pregnancy 6.7 ( ) 6.6 ( ) 0.17 Late pregnancy 6.3 ( ) 5.9 ( ) Insulin dose (IU/kg/24 h) Early pregnancy 0.93 ( ) 0.93 ( ) 0.69 Late pregnancy 1.07 ( ) 1.05 ( ) 0.78 Gestational weight gain (kg) 15.1 ( ) 14.8 ( 2.0 to 34.4) 0.77 Vitamin D level in late pregnancy (nmol/l) 47.9 ( ) 58.6 ( ) 0.03 Categorial variables were compared by Chi-square or Fisher s exact test, as appropriate. Continuous variables were analyzed by Mann-Whitney test. A p-value <0.05 was regarded statistical significant (in bold). Results are given as median (range) or numbers (%). BMI, body mass index. a Data on educational level are in total missing in 10 women ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)

5 M. Vestgaard et al. Vitamin D in pregnancy with diabetes Preterm delivery (%) % < Vitamin D (nmol/l) Figure 1. The prevalence of preterm delivery (<37 weeks) in 198 pregnant women with type 1 diabetes according to vitamin D levels. The cohort are divided according to vitamin D levels <25 nmol/l (n = 17), between 25 and 49 nmol/l (n = 56) or >50 nmol/l (n = 125) in late pregnancy. [Color figure can be viewed at wileyonlinelibrary.com] preeclampsia in women with vitamin D insufficiency in late pregnancy. Among the 19 women with kidney involvement in early pregnancy 26% (n = 5) developed preeclampsia. Inflammatory markers were comparable between women with and without vitamin D insufficiency in late pregnancy (Table 2). However, women with vitamin D insufficiency in early pregnancy were characterized by slightly higher plasma IL-6 levels in comparison with women with vitamin D levels 50 nmol/l (1.8 vs. 1.5 ng/l, p = 0.047). 23% 15% Among the 20 healthy pregnant women, the vitamin D level at 15 weeks was 72 (range ) nmol/l in comparison with 62 (range ) nmol/l among the women with type 1 diabetes in early pregnancy (8 weeks) (p = 0.07). Vitamin D insufficiency was present in 20% in comparison with 32% among the women with type 1 diabetes in early pregnancy (p = 0.07). The levels of parathyroid hormone, calcium, albumin, magnesium, high-sensitivity C-reactive protein and IL-6 in the healthy women were all similar to the levels in early pregnancy in the women with type 1 diabetes (data not shown). Discussion Preterm delivery occurred with almost double frequency in women with vitamin D insufficiency in late pregnancy compared with the remaining women with type 1 diabetes. After adjustments, the OR of the association was at the same level, but became nonsignificant. The number of cases with preeclampsia was small and an association between vitamin D insufficiency and preeclampsia was not identified. The association between vitamin D levels and preterm delivery is in accordance with the findings in women without diabetes, where an association between vitamin D insufficiency and preterm delivery was summarized in a systematic review with an OR of 1.6 (3) Table 2. Paraclinical data, pregnancy complications and outcome in 198 pregnant women with type 1 diabetes with vitamin D insufficiency (<50 nmol/l) or vitamin D 50 nmol/l in late pregnancy. Women with vitamin D insufficiency (n = 73) Women with vitamin D 50 nmol/l (n = 125) p-value Paraclinical data Vitamin D (nmol/l) 37 (10 49) 71 (50 141) <0.001 Parathyroid hormone (pmol/l) 2.9 ( ) 2.1 ( ) <0.001 Calcium (mmol/l) 2.2 ( ) 2.2 ( ) 0.92 Magnesium (mmol/l) 0.7 ( ) 0.7 ( ) 0.83 Albumin (g/l) 27 (21 32) 27 (22 38) 0.79 High-sensitivity C-reactive protein 3.7 ( ) 3.6 ( ) 0.52 (mg/ml) Interleukin-6 (ng/l) 2.6 ( ) 2.3 ( ) 0.87 Pregnancy complications and outcome Preeclampsia 8 (11%) 8 (6%) 0.26 Preterm delivery 20 (27%) 19 (15%) 0.04 Antihypertensive treatment a 12 (16%) 23 (18%) 0.73 Infant birthweight (grams) 3464 ( ) 3540 ( ) 0.41 Infant birthweight (SD score) 1.04 ( 1.45 to 6.71) 1.04 ( 1.90 to 4.29) 0.74 Large for gestational age 31 (42%) 60 (48%) 0.45 Small for gestational age 1 (1%) 4 (3%) 0.65 Categorial variables were compared by Chi-square or Fisher s exact test, as appropriate. Continuous variables were analyzed by Mann-Whitney test. A p-value < 0.05 was regarded statistical significant (in bold). Results are given as median (range) or numbers (%). a Antihypertensive treatment at some point during pregnancy. ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)

6 Vitamin D in pregnancy with diabetes M. Vestgaard et al. for preterm delivery in women with vitamin D insufficiency. In addition, the prevalence of preterm delivery in the present study was numerically higher in women with vitamin D <25 mmol/l in comparison with vitamin D levels of nmol/l or 50 nmol/l, respectively, suggesting a dose response relation. The findings regarding preeclampsia are in accordance with the only other study investigating the effect of the vitamin D levels on the development of preeclampsia, where there was a tendency to lower vitamin D levels among women developing preeclampsia, but no statistically significant difference was found (14). In a systematic review including nine studies reporting on preeclampsia, the pooled OR for preeclampsia in healthy women with vitamin D insufficiency was 1.8 (95% CI ) (9). Several pathophysiological pathways for an association between vitamin D insufficiency and preterm delivery, have been proposed; most consistently that vitamin D has immunomodulatory actions that may alter the expression of proinflammatory cytokines such as IL-6 (3,29,30). The cases of preterm delivery in our cohort were mainly medically indicated for obstetrical complications and spontaneous preterm delivery was only seen in a few women. A recent, large case control study among healthy pregnant women also showed a similar association between low vitamin D levels and the rate of medically indicated and spontaneous preterm delivery (4). Furthermore, low vitamin D levels have been proposed to be associated with increased prevalence of fetal distress, as suggested in healthy women in a recent large nested case control study (31). Vitamin D is suspected to alter the fetal heart in a way that makes it more vulnerable to fetal distress. Our data lack exact data on possible fetal distress and therefore this subgroup analysis was not carried out. As expected, the parathyroid hormone levels were slightly higher in the women with vitamin D insufficiency. Our findings of a higher plasma IL-6 level in early pregnancy in women with vitamin D insufficiency is in accordance with previous reports demonstrating higher levels of plasma IL-6 in patients with diabetes and vitamin D insufficiency (32). In healthy women with preterm delivery (33) a higher level of IL-6 has been documented, indicating that proinflammatory conditions are important in the pathophysiology of preterm delivery. However, the plasma IL-6 level was only associated with vitamin D insufficiency in early and not in late pregnancy in the present study and one may therefore speculate whether factors other than increased proinflammatory state is the main mechanism for preterm delivery in these women with diabetes. In our cohort, many of the cases with preterm delivery were attributable to pregnancy-induced hypertensive disorders such as preeclampsia, and the mechanism of vitamin D insufficiency to induce preterm delivery may therefore be related to placenta development or hypertension per se. Vitamin D affects the genes responsible for the development of placenta (30), and so vitamin D insufficiency contributes to impaired growth and function of the placenta, possibly leading to preeclampsia and preterm delivery. Another potential pathophysiological pathway is that vitamin D insufficiency results in an inverse endocrine regulator of the renin angiotensin system that aggravates hypertension (17) and thereby may aggravate pregnancy-induced hypertensive disorders (7). The number of cases of preeclampsia was numerically higher among women with vitamin D insufficiency, but this was not significant. However, the numbers were small and it is not possible to exclude an association either. To support this hypothesis we would expect that blood pressure levels and the use of antihypertensive treatment during pregnancy were related to vitamin D status, which was not documented. Although it is biologically plausible that low vitamin D levels could be responsible for some of the cases of preterm delivery among women with type 1 diabetes, owing to the observational nature of the data, we cannot infer causality from these findings. Our findings of a high prevalence of vitamin D insufficiency among pregnant women with type 1 diabetes is in accordance with a study from the UK (12). However, we could not confirm an association between HbA1c in early pregnancy and the vitamin D level. The reason for this could be due to the majority of the patients having tight glycemic control in our cohort. Studies in healthy pregnancy and in women with type 1 diabetes (12) have shown a positive correlation between vitamin D levels in the mother and the newborn, which is approximately 80% of maternal concentrations (34). Therefore an even higher rate of vitamin D insufficiency in the infant of mothers with type 1 diabetes can be expected. Major strengths of our study include the use of the todate largest study exploring vitamin D insufficiency in pregnancy with type 1 diabetes with a thorough validation of all data. A post hoc power calculation was performed with the observed prevalence in preterm delivery between women with vitamin D insufficiency and women with vitamin D levels 50 nmol/l in late pregnancy (15% vs. 27%). In order to detect this difference with a significance level of 0.05 and a power of 0.8, approximately 400 women would be needed. An even larger group is needed to demonstrate a possible difference in the occurrence of preeclampsia in women with and without vitamin D 1202 ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)

7 M. Vestgaard et al. Vitamin D in pregnancy with diabetes insufficiency. Hence, our study lacks sufficient statistical power. All vitamin D analyses were performed at the same time using the same methods after insuring that storing of plasma samples does not influence the stability of vitamin D (35). Furthermore, women were mainly of Caucasian origin, and differences in skin type should therefore not influence the results. A limitation of our study is that we do not have data on known potential confounders for the vitamin D level such as dietary habits, dietary supplementation, sun exposure and seasonal change. However, inclusion of the women was evenly distributed over the year. In addition, four women delivered before the blood sample was obtained in late pregnancy, which could have biased the results. Our data lacked further details on the indications behind preterm delivery. Therefore we were unable to perform subgroup analyses, which would have contributed to further insight on underlying etiologies. Even though this is the to-date largest cohort of its kind, limited numbers might impair a demonstration of an association between preterm delivery and vitamin D levels when taking several independent variables into account. The small number of women developing preeclampsia makes the risk of a type 2 error in the estimation of the association between vitamin D levels and preeclampsia considerable in this study. A systematic Cochrane review from 2016 including data from three trials involving 477 women suggests that vitamin D supplementation during pregnancy reduces the risk of preterm birth compared with no intervention or placebo (8.9% vs. 15.5%; relative risk 0.36; 95% CI ). However, studies including women with diabetes were not available (5). In conclusion, among women with type 1 diabetes, preterm delivery was twice as frequent in women with vitamin D insufficiency in late pregnancy in crude analysis, but the increase was nonsignificant after adjustments. Low levels of vitamin D were not associated with preeclampsia. Larger studies investigating whether low vitamin D levels are a risk factor for preterm delivery in women with diabetes are needed. Funding MV has received financial support from Rigshospitalet s Research Foundation. ALS received financial support from the European Foundation for the Study of Diabetes and LifeScan. ERM received financial support from the Novo Nordisk Foundation. References 1. Colstrup M, Mathiesen ER, Damm P, Jensen DM, Ringholm L. Pregnancy in women with type 1 diabetes: have the goals of St. Vincent declaration been met concerning foetal and neonatal complications? J Matern Fetal Neonatal Med. 2013;26: Persson M, Norman M, Hanson U. Obstetric and perinatal outcomes in type 1 diabetic pregnancies: a large, population-based study. Diabetes Care. 2009;32: Wei SQ, Qi HP, Luo ZC, Fraser WD. Maternal vitamin D status and adverse pregnancy outcomes: a systematic review and meta-analysis. J Matern Fetal Neonatal Med. 2013;26: Bodnar LM, Platt RW, Simhan HN. Early-pregnancy vitamin D deficiency and risk of preterm birth subtypes. Obstet Gynecol. 2015;125: De-Regil LM, Palacios C, Lombardo LK, Pe~na-Rosas JP. Vitamin D supplementation for women during pregnancy. Cochrane Database Syst Rev. 2016;(1):CD Wei SQ, Audibert F, Hidiroglou N, Sarafin K, Julien P, Wu Y, et al. Longitudinal vitamin D status in pregnancy and the risk of pre-eclampsia. BJOG. 2012;119: Bakacak M, Serin S, Ercan O, K ost u B, Avci F, Kılıncß M, et al. Comparison of Vitamin D levels in cases with preeclampsia, eclampsia and healthy pregnant women. Int J Clin Exp Med. 2015;8: van Weert B, van den Berg D, Hrudey EJ, Oostvogels AJ, de Miranda E, Vrijkotte TG. Is first trimester vitamin D status in nulliparous women associated with pregnancy related hypertensive disorders? Midwifery. 2016;34: Aghajafari F, Nagulesapillai T, Ronksley PE, Tough SC, O Beirne M, Rabi DM. Association between maternal serum 25-hydroxyvitamin D level and pregnancy and neonatal outcomes: systematic review and meta-analysis of observational studies. BMJ. 2013;346:f Milman N, Hvas AM, Bergholt T. Vitamin D status during normal pregnancy and postpartum. A longitudinal study in 141 Danish women. J Perinat Med. 2011;40: Mulligan ML, Felton SK, Riek AE, Bernal-Mizrachi C. Implications of vitamin D deficiency in pregnancy and lactation. Am J Obstet Gynecol. 2010;202:429.e Bennett SE, McPeake J, McCance DR, Manderson JG, Johnston P, McGalliard R, et al. Maternal vitamin D status in type 1 diabetic pregnancy: impact on neonatal vitamin D status and association with maternal glycaemic control. PLoS ONE. 2013;8:e Tamblyn JA, Hewison M, Wagner CL, Bulmer JN, Kilby MD. Immunological role of vitamin D at the maternalfetal interface. J Endocrinol. 2015;224:R Azar M, Basu A, Jenkins AJ, Nankervis AJ, Hanssen KF, Scholz H, et al. Serum carotenoids and fat-soluble vitamins in women with type 1 diabetes and preeclampsia: a longitudinal study. Diabetes Care. 2011;34: Zhu N, Wang J, Gu L, Wang L, Yuan W. Vitamin D supplements in chronic kidney disease. Ren Fail. 2015;37: ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)

8 Vitamin D in pregnancy with diabetes M. Vestgaard et al. 16. Min B. Effects of Vitamin D on blood pressure and endothelial function. Korean J Physiol Pharmacol. 2013;17: Andersen LB, Przybyl L, Haase N, von Versen-H oynck F, Qadri F, Jørgensen JS, et al. Vitamin D depletion aggravates hypertension and target-organ damage. J Am Heart Assoc. 2015;4:pii: e Pittas AG, Dawson-Hughes B. Vitamin D and diabetes. J Steroid Biochem Mol Biol. 2010;121: Nielsen LR, Pedersen-Bjergaard U, Thorsteinsson B, Johansen M, Damm P, Mathiesen ER. Hypoglycemia in pregnant women with type 1 diabetes: predictors and role of metabolic control. Diabetes Care. 2008;31: Secher AL, Ringholm L, Andersen HU, Damm P, Mathiesen ER. The effect of real-time continuous glucose monitoring in pregnant women with diabetes: a randomized controlled trial. Diabetes Care. 2013;36: Nielsen LR, Damm P, Mathiesen ER. Improved pregnancy outcome in type 1 diabetic women with microalbuminuria or diabetic nephropathy: effect of intensified antihypertensive therapy? Diabetes Care. 2009;32: Nielsen LR, Pedersen-Bjergaard U, Thorsteinsson B, Boomsma F, Damm P, Mathiesen ER. Severe hypoglycaemia during pregnancy in women with type 1 diabetes: possible role of renin-angiotensin system activity? Diabetes Res Clin Pract. 2009;84: Vestgaard M, Ringholm L, Laugesen CS, Rasmussen KL, Damm P, Mathiesen ER. Pregnancy-induced sightthreatening diabetic retinopathy in women with Type 1 diabetes. Diabet Med. 2010;27: UNESCO, United Nations educational saco. International Standard Classification of Education. ISCED, Ross AC. The 2011 report on dietary reference intakes for calcium and vitamin D. Public Health Nutr. 2011;14: Secher AL, Parellada CB, Ringholm L, Asbj ornsdottir B, Damm P, Mathiesen ER. Higher gestational weight gain is associated with increasing offspring birth weight independent of maternal glycemic control in women with type 1 diabetes. Diabetes Care. 2014;37: Ringholm L, Pedersen-Bjergaard U, Thorsteinsson B, Boomsma F, Damm P, Mathiesen ER. A high concentration of prorenin in early pregnancy is associated with development of pre-eclampsia in women with type 1 diabetes. Diabetologia. 2011;54: Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr. 1996;85: Bodnar LM, Krohn MA, Simhan HN. Maternal vitamin D deficiency is associated with bacterial vaginosis in the first trimester of pregnancy. J Nutr. 2009;139: Liu NQ, Hewison M. Vitamin D, the placenta and pregnancy. Arch Biochem Biophys. 2012;523: Lindqvist PG, Silva AT, Gustafsson SA, Gidl of S. Maternal vitamin D deficiency and fetal distress/birth asphyxia: a population-based nested case control study. BMJ Open. 2016;6:e Tiwari S, Pratyush DD, Gupta SK, Singh SK. Vitamin D deficiency is associated with inflammatory cytokine concentrations in patients with diabetic foot infection. Br J Nutr. 2014;112: Osanyin GE, Adegbola O. Maternal serum interleukin 6 levels and fetomaternal outcomes in women with preterm premature rupture of membranes in Lagos, South-western Nigeria. J Matern Fetal Neonatal Med. 2016;29: Bowyer L, Catling-Paull C, Diamond T, Homer C, Davis G, Craig ME. Vitamin D, PTH and calcium levels in pregnant women and their neonates. Clin Endocrinol (Oxf). 2009;70: Hollis BW. Measuring 25-hydroxyvitamin D in a clinical environment: challenges and needs. Am J Clin Nutr. 2008;88:507S 10S ª 2017 Nordic Federation of Societies of Obstetrics and Gynecology, Acta Obstetricia et Gynecologica Scandinavica 96 (2017)