Serum Carotenoids and Fat-Soluble Vitamins in Women with Type 1 Diabetes and Preeclampsia a Longitudinal Study

Clinical Care/Education/Nutrition/Psychosocial Research ORIGINAL ARTICLE

Serum Carotenoids and Fat-Soluble Vitamins in Women With Type 1 Diabetes and Preeclampsia A longitudinal study

1 5,7 MADONA AZAR, MD TORE HENRIKSEN, MD pathogenesis of PE (3,4), and compared 2 8 ARPITA BASU, PHD SATISH K. GARG, MD 1,3 9 with the general population, women with ALICIA J. JENKINS, MD SAMAR M. HAMMAD, PHD 4 10 type 1 diabetes may be particularly prone ALISON J. NANKERVIS, MD JAMES A. SCARDO, MD 5,6,7 11 to and affected by impaired antioxidant KRISTIAN F. HANSSEN, MD CHRISTOPHER E. ASTON, PHD 5,7 1,11 status. Any alterations in maternal anti- HANNE SCHOLZ, PHD TIMOTHY J. LYONS, MD oxidant status in pregnancy are therefore of interest. Risk for PE has also been as- fi OBJECTIVE—Increased oxidative stress and immune dysfunction are implicated in pre- sociated with vitamin D de ciency, per- eclampsia (PE) and may contribute to the two- to fourfold increase in PE prevalence among haps mediated through sequelae such as women with type 1 diabetes. Prospective measures of fat-soluble vitamins in diabetic pregnancy immune dysfunction, inflammation, and are therefore of interest. hypertension (5,6). Several cross-sectional PE studies RESEARCH DESIGN AND METHODS— Maternal serum carotenoids (a-andb-carotene, have been performed, mostly in nondia- lycopene, and lutein) and vitamins A, D, and E (a-andg-tocopherols) were measured at first 6 6 6 6 betic women. Altered levels of carote- (12.2 1.9 weeks [mean SD], visit 1), second (21.6 1.5 weeks, visit 2), and third (31.5 a b 1.7 weeks, visit 3) trimesters of pregnancy in 23 women with type 1 diabetes who subsequently noids ( -and -carotene, lycopene, and developed PE (DM PE+) and 24 women with type 1 diabetes, matched for age, diabetes duration, lutein) have been reported (7). Mikhail et al. (8) found decreased maternal b-carotene HbA1c, and parity, who did not develop PE (DM PE2). Data were analyzed without and with adjustment for baseline differences in BMI, HDL cholesterol, and prandial status. levels in PE, but other studies were not confirmatory (9,10). Vitamin A, or retinol, RESULTS—In unadjusted analysis, in DM PE+ versus DM PE2, a-carotene and b-carotene is an essential fat-soluble vitamin with P , were 45 and 53% lower, respectively, at visit 3 ( 0.05), before PE onset. In adjusted analyses, physiologic functions in mammalian repro- b fi the difference in -carotene at visit 3 remained signi cant. Most participants were vitamin D duction (11). Cross-sectional studies in deficient (,20 ng/mL), and vitamin D levels were lower in DM PE+ versus DM PE2 throughout the pregnancy, although this did not reach statistical significance. nondiabetic women with and without PE report both lower (9) and higher (10) vita- CONCLUSIONS—In pregnant women with type 1 diabetes, low serum a-andb-carotene min A levels. Vitamin E or a-tocopherol, were associated with subsequent development of PE, and vitamin D deficiency may also be which has antioxidant functions, has been implicated. reported to be lower (8,12), higher (9,10), or similar (13) between PE and normotensive nondiabetic pregnancies. There are some longitudinal PE stud- fi reeclampsia (PE) is de ned as new- mortality, and its incidence is increased ies, although none in type 1 diabetes. Ponset hypertension and proteinuria approximately fourfold by the presence Bodnar et al. (5) found that severe vitamin during pregnancy. PE is a major cause of maternal type 1 diabetes (1,2). Placental Ddeficiency [25(OH)D ,15 ng/mL] in of maternal and infant morbidity and oxidative stress is implicated in the early pregnancy was associated with a ccccccccccccccccccccccccccccccccccccccccccccccccc fivefold increased risk of PE. Halhali et al. From the 1Harold Hamm Oklahoma Diabetes Center and Section of Endocrinology and Diabetes, University of (14) showed that serum 1, 25(OH)2D Oklahoma Health Sciences Center, Oklahoma City, Oklahoma; the 2Department of Nutritional Sciences, was not altered in pregnant women who Human Environmental Sciences, Oklahoma State University, Stillwater, Oklahoma; the 3Department of subsequently developed PE. In a longitu- Medicine, The University of Melbourne, St Vincent’s Hospital, Melbourne, Victoria, Australia; the 4Diabetes dinal study of nondiabetic women with a Service, The Royal Women’s Hospital, Melbourne, Victoria, Australia; the 5Department of Gynecology and 6 risk of high PE, Chappell et al. (15) found Obstetrics, Oslo University Hospital, Oslo, Norway; the Department of Endocrinology, Oslo University fi Hospital, Oslo, Norway; the 7Institute of Clinical Medicine, University of Oslo, Oslo, Norway; the 8Barbara no signi cant differences in a-tocopherol Davis Center for Childhood Diabetes, University of Colorado Health Sciences Center, Aurora, Colorado; concentrations. the 9Department of Regenerative Medicine and Cell Biology, Medical University of South Carolina, 10 Thus, longitudinal studies are few, Charleston, South Carolina; the Spartanburg Regional Medical Center, Spartanburg, South Carolina; and and cross-sectional studies are not infor- the 11General Clinical Research Center, Oklahoma University Health Sciences Center, Oklahoma City, Oklahoma. mative on the temporal relationship be- Corresponding author: Timothy J. Lyons, [email protected]. tween antioxidant status and subsequent Received 14 November 2010 and accepted 3 March 2011. PE. Furthermore, there are no prospective DOI: 10.2337/dc10-2145 data relating serum carotenoids, fat-soluble M.A. and A.B. contributed equally to this work. © 2011 by the American Diabetes Association. Readers may use this article as long as the work is properly vitamins, and PE in diabetes. cited, the use is educational and not for profit, and the work is not altered. See http://creativecommons.org/ We report a multicenter prospec- licenses/by-nc-nd/3.0/ for details. tive study examining differences in care.diabetesjournals.org DIABETES CARE 1 Diabetes Care Publish Ahead of Print, published online April 26, 2011 Carotenoids, vitamins, and preeclampsia

Table 1—Clinical proﬁles of 47 type 1 diabetic and 20 nondiabetic participants

Diabetic P value: DM2 vs. P value: DM PE2 vs. Nondiabetic DM PE2 No PE (PE2) DM PE+ PE (PE+) N 20 24 23 Age (years) 31.7 6 4.6 0.17 29.9 6 3.8 0.31 28.5 6 5.6 BMI (kg/m2) 23.6 6 3.7 0.41 24.6 6 4.1 0.025 28.0 6 5.8 Alcohol use (%) None 11* 25 18 Stopped during pregnancy 680.55 580.39 68 Smoking (%) No 100* 88 91 Quit because of pregnancy 00.55 40.69 5 First pregnancy (%) 55 0.57 75 0.26 76 Gravida (n)1.76 1.0 0.24 1.3 6 0.7 0.99 1.3 6 0.7 Para (n)0.56 0.9 0.18 0.2 6 0.5 0.91 0.2 6 0.5 Abortus (n)0.26 0.4 0.86 0.1 6 0.3 0.91 0.1 6 0.4 Age at diabetes onset (years) — 15 6 8 0.065 12 6 6 Duration of diabetes (years) — 15 6 7 0.32 17 6 7 HbA1c (%) 5.3 6 0.3 ,0.0001 6.7 6 1.0 0.13 7.3 6 1.2 Blood pressure (mmHg) Systolic 112 6 90.351096 10 0.27 113 6 12 Diastolic 67 6 80.24646 8 0.27 67 6 9 Microalbumin (mg/dL) 0.47 6 0.17 0.58 0.43 6 0.20 0.12 1.03 6 1.78 Total cholesterol (mg/dL) 187 6 26 0.22 176 6 34 0.54 182 6 28 HDL cholesterol (mg/dL) 81 6 22 0.63 84 6 18 0.035 74 6 14 LDL cholesterol (mg/dL) 88 6 30 0.22 77 6 28 0.10 91 6 28 Triglycerides (mg/dL) 94 6 34 0.076 75 6 31 0.17 87 6 24 Gestational age (weeks) Visit 1 12.6 6 1.7 0.56 12.3 6 1.7 0.98 12.3 6 2.1 Visit 2 21.5 6 1.2 0.97 21.4 6 1.2 0.15 22.1 6 1.7 Visit 3 31.2 6 1.1 0.80 31.3 6 1.5 0.39 31.7 6 1.7 Term 39.2 6 1.5 0.013 38.0 6 1.4 0.036 37.2 6 1.2 Values are means 6 SD. Measurements refer to visit 1 unless otherwise indicated. P values , 0.05 are denoted in boldface. *For these data, P value refers to combined percentage (i.e., “none” and “stopped during pregnancy,” or “no” and “quit because of pregnancy”). serum carotenoids, vitamins A (retinol), enrolled during the first trimester of preg- larger cohort (16), samples from 23 were D [25(OH)D], and E (tocopherols) be- nancy and followed until delivery (16). available for this substudy because of sam- tween women with type 1 diabetes with Clinical data and specimens were col- ple attrition. Of 26 DM PE2 cases and without subsequent PE. Levels were lected at each trimester (visit 1: 12.2 6 matched on the basis of age, diabetes du- 6 measured in each trimester of pregnancy 1.9 weeks; visit 2: 21.6 1.5 weeks; ration, HbA1c, and parity, samples from and before PE onset. Our study evalu- and visit 3: 31.5 6 1.7 weeks of gestation 24 were available for this study (sample ated the temporal association of vitamin [mean 6 SD]; no overlap) and at term attrition). For reference values, 20 of 24 and antioxidant provitamin levels with PE (37.6 6 2.0 weeks). Visit 3 was before pregnant, nondiabetic, non-PE women in type 1 diabetes. PE onset. Subjects were requested to fast (DM2) were also studied (3 were previ- Because our major goal was to compare overnight, actual prandial status was ously excluded as described [16]; 1 addi- pregnant women with type 1 diabetes who recorded, and serum and urine were ob- tional exclusion due to sample attrition). subsequently developed PE with those tained before any exogenous insulin ad- who did not, our primary comparison is ministration. The study was approved by Laboratory analyses between these two groups. We also in- the institutional review boards of partici- Procedures have been detailed (16). Se- cluded healthy, normotensive, nondia- pating centers in Norway, Australia, and rum carotenoids, retinol, and tocopherols betic pregnant women to obtain reference the U.S. Exclusion criteria were renal im- were measured by high-performance liq- values during pregnancy, and for illustra- pairment (including microalbuminuria), uid chromatography (HPLC) using a tive purposes we report some secondary cardiovascular disease, hypertension, or modified combined version of methods analyses using their data. other significant medical problems pre- of Lee et al. (17) and Karppi et al. (18). pregnancy or at visit 1. PE was defined Serum vitamin D [25(OH)D] was mea- RESEARCH DESIGN AND as new-onset hypertension (.140/90 sured by HPLC/tandem mass spectrome- METHODS—This is a substudy of a mmHg) after 20 weeks’ gestation in a pre- try through Quest Diagnostics Co. (19). previously described prospective cohort viously normotensive woman, accompa- Briefly, 200 mL serum was deprotein- of 151 non-Hispanic white women with nied by proteinuria (.300 mg/24 h). Of ized with 200 mL ethanol-butylated hy- type 1 diabetes and 24 nondiabetic subjects the 26 diabetic PE (DM PE+) cases in the droxytoluene containing 50 mLinternal

2 DIABETES CARE care.diabetesjournals.org Azar and Associates standard cocktail, vortexed, extracted adjustment for covariates, only b-carotene Serum vitamin A levels did not differ (1,000 mL n-hexane, 60 s), dried (nitro- at visit 3 remained significant. Serum lutein significantly between the diabetic groups gen, 10 min), and reconstituted in 200 mL levels were similar in DM PE+ versus DM at any visit (Fig. 2A), but were lower in ethanol-butylated hydroxytoluene solu- PE2 at all visits (Fig. 1C); lycopene was DM PE2 versus DM2 at visits 2 and 3 tion. Then, 50 mL was injected onto a 52% higher in DM PE+ versus DM PE2 (Fig. 2A; P , 0.05). Consideration of co- 4.6-mm C-18 Ultrasphere ODS HPLC at visit 2 (P , 0.05) (Fig. 1D). Serum lutein variates had no effect. column (Beckman Coulter, Inc., Danvers, was lower at visits 1 and 2 in DM PE2 Vitamin D [25(OH)D] insufficiency, MA) and eluted (flow rate 0.8 mL/min) versus DM2 (Fig. 1C; P , 0.05). Differ- deficiency, and severe deficiency were with an isocratic solvent (methanol, ences in lutein and lycopene were not defined as ,30, ,20, and ,15 ng/mL, 60%; acetonitrile, 20%; and dichlorome- affected by the covariate analysis. respectively (20,21). Mean 25(OH)D thane, 20%). The HPLC system included a 515 pump, 2996 photodiode array detector, and a Rheodyne 7725i manual injector (Waters, Milford, MA). Data acqui- sition was via Waters Empower data soft- ware. Interassay coefficient of variation for pooled quality samples was #10%. All assays were performed by operators masked to sample clinical status.

Statistics The primary analyses compared DM PE+ with DM PE2. A secondary analysis eval- uated differences between DM PE2 and nondiabetic non-PE control subjects (DM2). Carotene and tocopherol levels were analyzed both with and without cor- rection for total lipids (cholesterol + tri- glyceride). Minimum differences between DM PE+ and DM PE2 (n $20/group) could be detected with 80% power as fol- lows: lutein: 0.4 mmol/L; a-tocopherol: 5.4 mmol/L; vitamin A: 0.18 mmol/L; vitamin D: 3.08 mg/L. Results are reported (Table 1) as means (6 SD) or, for non-normally distributed measures (per Shapiro-Wilks test), were transformed logarithmically and reported as geometric means (95% CI) (all carotenoids [except lutein], g-tocopherol, and a-or g-tocopherol:lipid ratios). Differences between groups were analyzed by Student t test for continuous measures and x2 test for categoric measures. Analyses were performed with and without inclusion of parameters that differed between DM PE+ and DM PE2 (BMI, HDL cholesterol), and prandial status, as covariates. Microsoft Excel 2007 (Microsoft Corp., Redmond, WA) and SPSS for Windows 15.0 (SPSS Inc., Chicago, IL) were used.

RESULTS—AsshowninTable1,at baseline, the diabetic groups were compa- Figure 1—Longitudinal changes in (A) a-carotene, (B) b-carotene, (C) lutein, and (D) lycopene rable except for higher BMI and lower during gestation. Values are geometric means for a-carotene, b-carotene, and lycopene, and arithmetic means for lutein. Error bars show the 95% CI for the mean. Three groups of partic- HDL cholesterol in DM PE+ versus DM 2 PE2. ipants are shown; white circle, dashed line: DM PE (n = 24); black circle, solid line: DM PE+ (n = 23); white square, dotted lines: DM2 (nondiabetic control subjects, n = 20). Except lutein, all Serum b-carotene was 37% lower at fi fi , P , values were log-transformed before testing for signi cant differences. Values signi cantly (P visit 1 ( 0.05), and a-andb-carotene 0.05) different between DM PE2 and DM PE+ groups are indicated with an asterisk (*) and were 45 and 53% lower, respectively, at between DM PE2 and DM2 groups are indicated with a double dagger (‡). After adjustment for visit 3, in DM PE+ versus DM PE2 (P , covariates (BMI, HDL cholesterol, and prandial status), differences in a-carotene at visit 3 and 0.05) (Fig. 1A and B), although after b-carotene at visit 1 were no longer significant (P = 0.09 and 0.12, respectively). care.diabetesjournals.org DIABETES CARE 3 Carotenoids, vitamins, and preeclampsia levels did not differ significantly between and all enrolled participants were Cau- (Fig. 3). Lipid-adjusted carotenes were DM PE+ and DM PE2 groups, although casian. lower in DM PE+ than DM PE2 at visit there was a slight trend toward lower lev- Unadjusted a-tocopherol levels were 1(b-carotene only) and visit 3 (both car- els in DM PE+ throughout pregnancy ;18% higher in DM PE+ versus DM PE2 otenes) (Fig. 3A and B; P , 0.05). Again, (Fig. 2B). Also, women with type 1 diabe- at visit 2 (Fig. 2C; P , 0.05), whereas after inclusion of covariates, only b-carotene tes in general, and those who developed g-tocopherol was similar in all groups at visit 3 remained significant. Adjusted PE in particular, were more likely to be at all visits (Fig. 2D). Mean serum lycopene and lutein levels showed no dif- vitamin D deficient or severely deficient a-tocopherol was lower in DM PE2 ver- ferences between groups at any visit (data than the nondiabetic group, and the rela- sus DM2 only at visit 2 (Fig. 2C; P , 0.05). not shown). By comparing DM PE2 with tive risk of PE associated with vitamin D No differences were noted for g-tocopherol DM2, adjusted a-andb-carotenoids deficiency was increased, although not at any visit (Fig. 2D). Covariate analysis were not significantly different, but they significantlyatanyvisit(Table2).Of did not affect significance. tended to decline with gestational age relevance to vitamin D, there were no Because carotenes and tocopherols in both groups. Lipid-adjusted a-and differences between groups in the seasons circulate primarily within lipoproteins, g-tocopherols did not differ between DM of sample collection (data not shown), levels were adjusted for serum total lipids PE+ and DM PE2, or between DM2 and DM PE2,althougha-tocopherol trended downward in all groups during pregnancy (Fig. 3C). Differences in tocopherol: lipid ratios were not affected by covariate analysis.

CONCLUSIONS—In a prospective study of pregnant women with type 1 diabetes, we demonstrated lower serum carotenoids, especially b-carotene, in affected versus unaffected women before PE onset. We also observed a high prevalence of vitamin D insufficiency, most fre- quently in women who developed PE. Relative to early pregnancy, there was also a downward trend in lipid-adjusted carotenoids and tocopherols in the third trimester, perhaps representing antioxidant reserve depletion. This decline oc- curred because total cholesterol and triglycerides increased by 1.5- and two- fold, respectively, between visits 1 and 3 (not shown) in all groups. We have previously shown that pregnant women with type 1 diabetes are already “primed” for PE by having elevated levels of endoglin (16), and in this setting, antioxidant depletion may trigger PE. We identified significantly lower serum b-carotene, and by some analyses, lower a-carotene, in pregnant women with type 1 diabetes who subsequently developed PE versus those who did not. Our longitudinal data, particularly between-group differences in serum carotenoids (DM PE+ vs. DM PE2)inthethird trimester, concur with existing cross- sectional studies, performed later in pregnancy in nondiabetic women, that show lower circulating carotenoids in PE (8,10). b-Carotene also possesses pro- — a Figure 2 Longitudinal changes in (A) retinol, (B) vitamin D (cholecalciferol), (C) -tocopherol, vitamin A activity (22), and we postulate and (D) g-tocopherol (log transformed) during gestation. Values are arithmetic means for retinol, cholecalciferol, and a-tocopherol, and geometric means (95% CI) for g-tocopherol. Group and that the low b-carotene in the third tri- statistical symbols are as in Fig. 1. Adjustment for covariates had no effect on significance. Sta- mester could be explained by increased tistically significant differences between visits were seen for (C) a-tocopherol in DM PE+ between endogenous antioxidant consumption gestational age 12 and 22 weeks, and in DM PE2 between gestational age 22 and 32 weeks due to elevated oxidative stress or in- (P , 0.05). creased conversion to vitamin A to

4 DIABETES CARE care.diabetesjournals.org Azar and Associates 7.7) 11.0) 12.4) – – – 30 ng/mL). RR, . (%) RR (95% CI) cient ( fi 2 (%) DM 2 15 ng/mL), and relative risk < cient ( fi 7.9)10.8)11.2) 96 61 39 91 52 30 75 45 30 1.5 (0.3 1.9 (0.3 2.1 (0.4 – – – RR (95% CI) DM PE+ (%) DM PE 2 20 ng/mL), or severely de < cient ( (%) DM fi 2 30 ng/mL), de < = 20). Reference group for relative risk in all cases comprises women in whom vitamin D level is suf n cient (

ﬁ Figure 3—Longitudinal changes in ratios (A) a-carotene: total lipids, (B) b-carotene: total lipids, (C) a-tocopherol: total lipids, and (D) g-tocopherol: total lipids, during gestation. Values are geometric means (95% CI). Ratios are expressed as serum tocopherol and carotenoid molarity

11.8)12.3)14.2) 95 60 30 90(m 70mol/L) 45 79 to 53 1.4 (0.3 total 26 1.8 (0.3 1.8lipid (0.3 mass (serum total cholesterol and triglycerides) (mg/dL). Group and – – – statistical symbols are as in Fig. 1. After adjustment for covariates (BMI, HDL cholesterol, and 15 ng/mL) in diabetic participants

< prandial status), differences in a-carotene at visit 3 and b-carotene at visit 1 were no longer significant (P = 0.08 and 0.07, respectively). (%) RR (95% CI) DM PE+ (%) DM PE 20, and compensate for low vitamin A levels. In Australia), were vitamin D deficient. < 2 our study, low vitamin A levels in women Levels did not differ between sites. We 30, fi < with type 1 diabetes in the rst trimester lack information on dietary intake or are comparable to those previously re- ultraviolet light exposure. The women Visit 1 Visit 2 Visit 3 (%) DM ported (9,10) and to levels in U.S. adult with type 1 diabetes as a whole had lower : nondiabetic, non-PE pregnant control subjects ( 2

2 women (23). Thus, although not statisti- levels than our nondiabetic reference cally signiﬁcant, decreasing vitamin A group (data not shown). The reason is levels between the second and third tri- unclear. Occult celiac disease, which

= 24); DM mesters in women with type 1 diabetes may cause vitamin D malabsorption, is n could promote PE. prevalent (5–10%) in type 1 diabetes Vitamin D deﬁciency, an increas- (24) and could be contributory.

:noPE( ingly prevalent condition in women of Links between vitamin D deﬁciency 2

DM PE+ (%) DM PE childbearing age (20), has been associ- and PE are unclear. Hyppönen (6) pro- ated with PE in a large prospective study posed an immune hypothesis, whereby Percentage of participants who are vitamin D [25(OH)D] insuf

= 23); PE (5). We found that the majority of our local disruption of vitamin D status pro- — n participants, regardless of geographic motes loss of maternal–fetal immune tol- location (which included sunny regions erance and immune maladaptation. 302015 96 78 39 87 65 22 84 37 2.1 (0.4 0 2.2 (0.4 2.5 (0.4 Vitamin D Table 2 PE+: PE ( relative risk. , , , of PE associated with low vitamin D levels ( level (ng/mL) such as Oklahoma, South Carolina, and Although this is intriguing, it has yet to care.diabetesjournals.org DIABETES CARE 5 Carotenoids, vitamins, and preeclampsia beestablishedasacauseforPEinhu- study of oxidative stress and antioxidant mans. Another possibility is diminished Acknowledgments—This work was sup- status in preeclampsia and normal preg- placental 1a-hydroxylase activity, which ported by a research grant from the Juvenile nancy. Free Radic Biol Med 2004;37:557– occurs in PE (25) and lowers local and Diabetes Research Foundation (1-2001-844), 570 circulating 1,25(OH) D. National Institutes of Health (NIH) (National 5. Bodnar LM, Catov JM, Simhan HN, 2 Center on Minority Health and Health Dis- Holick MF, Powers RW, Roberts JM. Ma- In agreement with Williams et al. (10), parities) Grant P20-MD-000528-05 to T.J.L., ternal vitamin D deficiency increases the our study does not support a relationship NIH (National Center for Research Resources) risk of preeclampsia. J Clin Endocrinol between serum a-andg-tocopherols Grant M01-RR-1070 to the General Clinical Metab 2007;92:3517–3522 and PE, although the decline in adjusted Research Centers at Medical University of 6. Hyppönen E. Vitamin D for the pre- a-tocopherol levels throughout the preg- South Carolina, and NIH Grant M01-RR- vention of preeclampsia? A hypothesis. nancy might have a permissive effect in 14467 to Oklahoma University Health Scien- Nutr Rev 2005;63:225–232 women already predisposed. The mean ces Center. 7. Paiva SA, Russell RM. Beta-carotene and first trimester serum a-tocopherol values Support from Novo Nordisk enabled the other carotenoids as antioxidants. J Am – m participation of the Barbara Davis Diabetes Coll Nutr 1999;18:426–433 (25 30 mol/L) in the women with type 1 fi diabetes in our study were comparable to Center for Childhood Diabetes. No other po- 8. Mikhail MS, Anyaegbunam A, Gar nkel D, tential conflicts of interest relevant to this ar- normal values in U.S. adults and above the Palan PR, Basu J, Romney SL. Preeclampsia ticle were reported. and antioxidant nutrients: decreased plasma threshold indicating vitamin E deficiency , M.A., A.B., and A.J.J. researched data, con- levels of reduced ascorbic acid, alpha- ( 11.6 mmol/L) (23). tributed to discussion, and wrote the article. tocopherol, and beta-carotene in women Our study has several limitations. A.J.N., K.F.H., H.S., T.H., S.K.G., S.M.H., and with preeclampsia. Am J Obstet Gynecol The findings merit confirmation in larger J.A.S. researched data, contributed to discus- 1994;171:150–157 prospective studies involving diabetic sion, and reviewed the article. C.E.A. researched 9. Zhang C, Williams MA, Sanchez SE, et al. and nondiabetic women. Furthermore, data and contributed to discussion. T.J.L. re- Plasma concentrations of carotenoids, reti- confounding due to certain group dif- searched data, contributed to discussion, and nol, and tocopherols in preeclamptic and ferences in this small cohort, although wrote the article. normotensive pregnant women. Am J Epi- – nonsignificant, cannot be excluded. More The skilled and dedicated assistance of the demiol 2001;153:572 580 following individuals for their scientific con- data regarding maternal diet, vitamin 10. Williams MA, Woelk GB, King IB, Jenkins L, tribution to this article is acknowledged: Mahomed K. Plasma carotenoids, retinol, supplement use, and ultraviolet exposure Yongxin Yu, Mingyuan Wu, Azar Dashti, and tocopherols, and lipoproteins in pre- would have been helpful to dissect cause Kerri May (University of Oklahoma); Torun eclamptic and normotensive pregnant and effect. Measures of other antioxidant Clausen and Bjørg Lorentzen (Oslo University Zimbabwean women. Am J Hypertens vitamins and enzymes would be of in- Hospital); Kathryn M. Menard (University of 2003;16:665–672 terest. Fasting overnight is challenging North Carolina); and John R. Stanley (Mercy 11. Clagett-Dame M, DeLuca HF. The role of for pregnant women with type 1 diabe- Health Center, Oklahoma). The following are vitamin A in mammalian reproduction tes, and at each visit, a similar proportion acknowledged for technical or clinical assis- and embryonic development. Annu Rev from each diabetic group (range 2–6of tance: Jill Mauldin and Mary Myers (Medical Nutr 2002;22:347–381 x 23 or 24) attended nonfasting. However, University of South Carolina); Jill Cole 12. Sagol S, Ozkinay E, Ozsener S. Impaired and Nancy Sprouse (Spartanburg Regional our conclusions were not substantially antioxidant activity in women with pre- Hospital, Spartanburg, SC); Myrra Windau eclampsia. Int J Gynaecol Obstet 1999;64: affected by statistical adjustment for (University of Colorado); Christine Knight, 121–127 prandial status (as presented) or exclu- Jennifer Conn, Peter England, Susan Hitchcock, 13. Hubel CA, Kagan VE, Kisin ER, McLaughlin sion of nonfasting visits (data not Jeremy Oats, and Peter Wein (Royal Women’s MK, Roberts JM. Increased ascorbate rad- shown). Hospital, Melbourne); and Julie Stoner and ical formation and ascorbate depletion in We believe this is the first longitudi- Lori Doyle (University of Oklahoma). At the plasma from women with preeclampsia: nal study of the relationship between PE University of Oklahoma, Kenneth Wilson implications for oxidative stress. Free Radic in pregnant women with type 1 diabetes and Gennadiy Moiseyev assisted with sample Biol Med 1997;23:597–609 and antecedent serum levels of carote- processing and the conduct of laboratory 14. Halhali A, Villa AR, Madrazo E, et al. noids and vitamins A, D, and E. We found work. Longitudinal changes in maternal serum that women with type 1 diabetes who 1,25-dihydroxyvitamin D and insulin like developed PE had lower antecedent se- growth factor I levels in pregnant women References who developed preeclampsia: compari- rum a-andb-carotene concentrations 1. Zhu L, Nakabayashi M, Takeda Y. Statis- son with normotensive pregnant women. than those who did not. Vitamin D insuf- tical analysis of perinatal outcomes in J Steroid Biochem Mol Biol 2004;89–90: ficiency affected most women with type 1 pregnancy complicated with diabetes mel- 553–556 diabetes, and levels tended to be lower in litus. J Obstet Gynaecol Res 1997;23:555– 15. Chappell LC, Seed PT, Briley A, et al. A those who developed PE. Lipid-adjusted 563 longitudinal study of biochemical vari- levels of many of the fat-soluble antioxi- 2. Hanson U, Persson B. Outcome of preg- ables in women at risk of preeclampsia. dants declined during pregnancy. Our nancies complicated by type 1 insulin- Am J Obstet Gynecol 2002;187:127– study sample, although small, repre- dependent diabetes in Sweden: acute 136 sented women with type 1 diabetes from pregnancy complications, neonatal mortality 16. Yu Y, Jenkins AJ, Nankervis AJ, et al. Anti- and morbidity. Am J Perinatol 1993;10: angiogenic factors and pre-eclampsia in different geographic locations. Further – fi 330 333 type 1 diabetic women. Diabetologia 2009; studies are needed to de ne whether op- 3. Hubel CA. Oxidative stress in the patho- 52:160–168 timizing fat-soluble antioxidant and vita- genesis of preeclampsia. Proc Soc Exp Biol 17. Lee BL, New AL, Ong CN. Simultaneous min status throughout pregnancy can Med 1999;222:222–235 determination of tocotrienols, tocopherols, reduce the high incidence of PE in those 4. Llurba E, Gratacós E, Martín-Gallán P, retinol, and major carotenoids in human with type 1 diabetes. Cabero L, Dominguez C. A comprehensive plasma. Clin Chem 2003;49:2056–2066

6 DIABETES CARE care.diabetesjournals.org Azar and Associates

18. Karppi J, Nurmi T, Olmedilla-Alonso B, pregnant and nonpregnant women of 23. Tohill BC, Heilig CM, Klein RS, et al. Granado-Lorencio F, Nyyssönen K. Si- childbearing age in the United States. Am J Nutritional biomarkers associated with multaneous measurement of retinol, alpha- Obstet Gynecol 2010;202:436.e1–e8 gynecological conditions among US women tocopherol and six carotenoids in human 21. Bischoff-Ferrari HA, Giovannucci E, with or at risk of HIV infection. Am J Clin plasma by using an isocratic reversed-phase Willett WC, Dietrich T, Dawson-Hughes B. Nutr 2007;85:1327–1334 HPLC method. J Chromatogr B Analyt Estimation of optimal serum concentra- 24. Crone J, Rami B, Huber WD, Granditsch G, Technol Biomed Life Sci 2008;867:226– tions of 25-hydroxyvitamin D for multiple Schober E. Prevalence of celiac disease and 232 health outcomes. Am J Clin Nutr 2006;84: follow-up of EMA in children and adoles- 19. Quest Diagnostics. Available from http:// 18–28 cents with type 1 diabetes mellitus. J Pediatr www.questdiagnostics.com/hcp/topics/ 22. Strobel M, Tinz J, Biesalski HK. The im- Gastroenterol Nutr 2003;37:67–71 endo/vitamin_d.html. Accessed 17 March portance of beta-carotene as a source of 25. Mulligan ML, Felton SK, Riek AE, Bernal- 2011 vitamin A with special regard to pregnant Mizrachi C. Implications of vitamin D 20. Ginde AA, Sullivan AF, Mansbach JM, and breastfeeding women. Eur J Nutr deﬁciency in pregnancy and lactation. Am Camargo CA Jr. Vitamin D insufﬁciency in 2007;46(Suppl. 1):I1–I20 J Obstet Gynecol 2010;202:429.e1–e9

care.diabetesjournals.org DIABETES CARE 7