The Association of Alanine Aminotransferase in Early Pregnancy with Gestational Diabetes

Abstract

Background:

Elevated alanine amino transferase, attributed to nonalcoholic fatty liver, is associated with later development of type 2 diabetes mellitus. We sought to determine whether maternal ALT values are associated with subsequent development of gestational diabetes.

Methods:

We performed a nested case–control study utilizing prospectively banked serum samples collected in early gestation. We excluded women with known diabetes, liver disease, or alcohol use. We included 83 cases of gestational diabetes mellitus (GDM) and 247 controls matched for prepregnancy body–mass index (BMI) and compared ALT values. We then performed a conditional logistic regression to model the adjusted odds of GDM in women with ALT ≥19 U/L, stratified by prepregnancy BMI.

Results:

The median (interquartile range) ALT in cases was 15 (12, 19) IU/L compared to 13 (11, 18) IU/L in controls (P = 0.07). Among women with a prepregnancy BMI <30 kg/m², ALT ≥19 U/L was associated with a fourfold increased odds of GDM (adjusted odds ratio [aOR] 4.56 [1.45, 14.27]), while there was no such association among obese women (aOR 0.36 [0.11, 1.20]). Similarly, each unit increase in log-transformed ALT was associated with a threefold increased odds of GDM in nonobese (aOR 3.15 [1.04,9.54]), but not obese (aOR 3.15 [0.30,3.15]) women.

Conclusion:

The association of high normal ALT and later GDM in nonobese women may reflect the role of hepatic insulin resistance and visceral obesity.

Introduction

Gestational diabetes mellitus (GDM) affects 6%–7% of all pregnancies.¹ There is ongoing debate regarding how to identify women at elevated risk for GDM who may benefit from early screening. Among the criteria supported by the American College of Obstetrics and Gynecology to select patients for consideration of early glucose screening is “known impaired glucose metabolism”²; however, women without risk factors have no reason to be screened for glucose intolerance before pregnancy.

Nonalcoholic fatty liver disease (NAFLD), considered the hepatic manifestation of metabolic syndrome, has a well-established association with type 2 diabetes mellitus (DM2) in nonpregnant adults thought to be due to fatty infiltration resulting in hepatic insulin resistance.^3
–5 A large Swedish cohort study recently identified an increased risk of GDM among women with an ICD-9 diagnosis of NAFLD.⁶ Assimilating from MRI, ultrasound, and biopsy studies, the World Gastroenterology Organisation estimates the prevalence of NAFLD among adults in the United States to be between 27% and 34%.⁷ Efforts to identify noninvasive screening tests have demonstrated a strong correlation between otherwise unexplained elevated ALT and NAFLD.^8
–10 Furthermore, unexplained elevated ALT is independently associated with insulin resistance as well as both current and future development of DM2.^11,12

Given the high estimated prevalence of NAFLD and the potential for identifying women with metabolic syndrome and thus “known impaired glucose metabolism,” we sought to determine whether first trimester ALT values predict later development of GDM.

Materials and Methods

Study population

This study is a secondary analysis of data from a longitudinal cohort study designed to evaluate the utility of angiogenic markers for the diagnosis of preeclampsia.¹³ In the original cohort, women were recruited from Brigham and Women's Hospital, an urban academic medical center in Boston, MA, from 2006 through 2013. Informed consent was obtained from all subjects. In the absence of prior data in this field to calculate a power estimate, we included all available cases of GDM in women delivering between 2006 and 2013. To maximize effect detection, three controls were selected for each case. Controls were matched 3:1 by body–mass index (BMI) within 2 kg/m² and had normal glucose tolerance based on a 50 g glucose load test ≤120 mg/dL. We excluded women with multiple gestation, known type 1 or 2 DM, self-reported prepregnancy alcohol abuse or moderate/heavy alcohol use during pregnancy, or known viral or immunologic liver disease.

Outcome assessment

The practice at our institution is to perform universal screening with a 50 g GLT at 24 weeks and if the 1-hr serum glucose value is at or above 130 mg/dL, we proceed with a 3-hr 100 g oral glucose tolerance test (GTT). We defined gestational diabetes by a 3-hr GTT meeting Carpenter–Coustan criteria or by clinical diagnosis in women who did not have a GTT but were prescribed insulin starting in the third trimester for glycemic control.

Exposure assessment

We used the earliest available samples from the original cohort, and the mean (standard deviation [SD]) gestational age of sample collection was 10.4 (2.6) weeks. No samples were fasting. Samples were stored at −80°C after the initial collection and thawed in the laboratory for assay. ALT was assayed using the ARCHITECT cSystems (Abbot Diagnostics, Abbott Park, IL) at the Massachusetts General Hospital clinical laboratory.

Covariate assessment

Medical records were abstracted by trained reviewers to obtain relevant pregnancy outcomes. Demographic (including maternal age, race, and ethnicity) and historical information (including pregnancy history, relevant medical history, history of GDM in prior pregnancy, and medication use) was collected by a participant questionnaire at the study enrollment. Maternal BMI was calculated using reported prepregnancy weight and height measured in the first trimester.

Incidences of both GDM and NAFLD are known to vary with race and ethnicity.^2,14,15 It is unclear whether ALT values vary among ethnic populations independent of liver disease. Both incidence of GDM and values of ALT vary with age.^16,17 Health insurance was used as a proxy for socioeconomic status, another minor risk factor for gestational diabetes.¹⁸ We also identified women with a history of thyroid disease or use of antiseizure medication as they have known associations with GDM or elevated liver function tests. In addition, we identified women who used narcotic pain medication during pregnancy as a proxy for risk for excessive acetaminophen use. Last, up to 50% of women who suffer from hyperemesis gravidarum can have elevated transaminases at the time of their marked nausea.¹⁹ As this condition is poorly documented, we used poor maternal weight gain, defined as less than five pounds, as a coarse marker and analyzed it as a potential covariate.

Statistical analysis

We compared baseline sociodemographics and prevalence of selected pregnancy characteristics of cases and controls using the Wilcoxon rank-sum test for continuous variables and the chi-squared, Fisher's exact, or analysis of variance (ANOVA) for categorical variables, where appropriate. Statistical testing of ALT results was performed after logarithmic transformation. Median ALT values were compared using the Wilcoxon rank-sum test. In fully adjusted multivariate conditional logistic regression models, maternal age and Asian racial status were included a priori, and additional covariates were added in a forward stepwise model selection procedure with inclusion in final models if they altered effect estimates by greater than 10%. Asian racial status was selected as the a priori variable based on the significant disproportion of Asian race presence in the GDM group (3% vs. 17%), driving the overall difference among race/ethnicity categories. Variables that were considered included covariates known to be associated with either GDM or abnormal ALT, including health insurance category, smoking status, history of GDM in prior pregnancy, history of thyroid disease, history of seizure disorders, history of preeclampsia, current preeclampsia, pregnancy weight gain less than five pounds, use of pain medication, parity, and gender of infant. We examined the association between GDM and log ALT as a continuous variable as well as between GDM and ALT ≥19U/L.²⁰ We then stratified subjects by BMI above and below 30 kg/m² to account for the strong association between ALT values and obesity.²¹

All analyses were performed with SAS version 9.4 (SAS Institute, Inc., Cary, NC).

The study was approved by the Institutional Review Board at Brigham and Women's Hospital (protocol no. 2009P000810).

Results

We identified 83 cases of GDM and 247 matched controls. As described in Table 1, cases were more likely to be Asian (17% vs. 3%, P = 0.02) and were older than controls (mean age [SD] 34 [5] vs. 32 [6]). In addition, in our cohort, subjects with GDM were more likely to have private insurance than control subjects (81% vs. 68%, P = 0.02). On average, infants in the control group delivered at slightly later in gestation than the GDM group (38.6 weeks vs. 38.4 weeks, P = 0.03).

Table 1.

Clinical and Demographic Population Characteristics Stratified by Gestational Diabetes Diagnosis

	Control, N = 247	GDM, N = 83
Characteristics	% (n) or mean (SD)	% (n) or mean (SD)	P
Maternal age (years)	32 (6)	34 (5)	0.001
BMI (kg/m²)	29.2 (6.6)	29.7 (6.6)	0.70
Nulliparous (%)	26% (63)	35% (29)	0.10
Race/ethnicity (%)
Caucasian	38% (94)	25% (21)	0.03
Hispanic	17% (42)	20% (17)
African American	36% (88)	33% (27)
Asian	3% (7)	17% (14)
Other	6% (16)	5% (4)
Insurance type
Private Insurance/HMO	68% (163)	81% (65)	0.02
Medicaid/SSI/Mass Health	32% (77)	19% (15)
Smoked during pregnancy	10% (25)	4% (3)	0.07
Personal history of thyroid disease	6% (14)	10% (8)	0.21
Personal history of seizure disorder	2% (4)	2% (2)	0.64
History of GDM in prior pregnancy	4% (9)	1% (1)	0.18
Infant gender (% male)	55% (132)	59% (48)	0.55
Maternal weight gain ≤5 lbs	27% (66)	34% (28)	0.22
Use of pain medication	4% (10)	6% (5)	0.54
Gestational age at plasma sample (week)	10.4% (2.5)	10.9 (2.8)	0.34

BMI, body–mass index; GDM, gestational diabetes mellitus; SD, standard deviation.

The median [interquartile range] ALT was 15 [12,19] U/L among subjects who developed GDM and 13 [11,18] U/L among control subjects (P = 0.07). 19 U/L was the 75 percentile in our population. There was no significant association between each unit increase in log-transformed ALT and later diagnosis of GDM (unadjusted OR [uOR] [CI] 1.53 [0.90, 2.62], adjusted OR [aOR] 1.68 [0.85, 3.30]). There was also no significant association between ALT at or above 19 U/L and later diagnosis of GDM. (uOR 1.50 [0.88, 2.55] aOR 1.37 [0.73, 2.59]).

Nonobese versus obese effect on GDM

When stratified by maternal BMI, ALT in the first trimester was associated with GDM among nonobese women only (Table 2). Women with a BMI <30 kg/m² before pregnancy with an early ALT at or above 19 U/L had over a fourfold increased odds of developing GDM compared to those with an ALT below 19 U/L. We did not see this association among obese women. There was also a threefold increased odds of GDM for each increase in log-transformed ALT among the nonobese women.

Table 2.

ALT and GDM Stratified by Maternal BMI

	BMI <30 (n = 184)		BMI ≥30 (n = 146)
	Unadjusted OR (95% CI)	Adjusted^ OR (95% CI)*	Unadjusted OR (95% CI)	Adjusted^ OR (95% CI)*
ALT
Continuous (log)	3.03 (1.26, 7.31)	3.16 (1.02, 9.75)	0.88 (0.41, 1.91)	1.24 (0.38, 4.10)
≥19 U/L	2.89 (1.33, 6.26)	4.42 (1.38, 14.16)	0.59 (0.26, 1.37)	0.44 (0.13, 1.53)

Model adjusted for maternal age, race (Asian vs. not Asian), medical insurance, parity, smoking, history of thyroid disease, history of seizure disorders, and maternal weight gain ≤5 lbs.

Discussion

In this study, we found an association between ALT at or above 19 U/dL and later development of GDM among nonobese women that was not present among obese women. There was a nonsignificant trend in all subjects toward an increased median ALT among those later diagnosed with GDM compared to subjects with normal glucose tolerance.

Our study extends the current literature assessing the relationship between markers of hepatic insulin resistance and GDM. Few studies have previously addressed the relationship between liver enzyme patterns and subsequent GDM. Sridhar et al. examined the correlation of liver enzymes assessed before pregnancy and incidence of GDM in the subsequent pregnancy.²² They examined ALT values drawn on average 7 years before the pregnancy and found no correlation between GDM and ALT. Tan et al. checked liver enzymes at the time of the GLT and found no association between ALT values and subsequent diagnosis.²³ Neither study investigated differences by BMI, which we found to be a significant confounder.

The Swedish birth registry cited earlier found an analogous pattern of NAFLD and GDM with a strong risk among women with a BMI less than 30 kg/m² (adjusted risk ratio [aRR] 12.50 [4.76, 32.66]) and no significant risk among women with a BMI above 30 kg/m² (aRR 1.67 [0.56, 4.97]).⁶ The rarity of NAFLD observed in their population, one case for every 20,000 pregnancies, as well as the reliance on diagnostic codes, suggests they may have selectively captured women with advanced disease. Our study sought to utilize a soft maker for NAFLD that is significantly less expensive and less morbid to obtain than liver biopsy.

Our study has several strengths: all the data were prospectively collected according to standard definitions; all eligible pregnancies were included and managed according to standardized protocols during the whole study period; there was no loss of follow-up; and logistic regression analysis was performed to adjust for potential confounders. Our findings must of course be interpreted in the context of our study design. Our sample size, dictated by availability, may have been underpowered to detect associations in the overall population and in obese women in particular. The strengths of our subgroup analysis are that only two subgroups were evaluated and they were prespecified based on the differences in the rate of elevated ALT and NAFLD in strata of BMI described earlier.^24,25 Our results must be followed by a prospective study designed to address ALT at or above 19 U/L as a risk factor for later diagnosis of GDM targeting lean and overweight women.

A final limitation is the presumptive association of NAFLD, a histologic diagnosis, with serum ALT activity. We attempted to recreate the paradigm of other researchers by eliminating other common causes of elevated ALT, namely alcohol abuse, viral hepatitis, and immunologic liver disease.^21,25 However, our efforts may have been limited with regard to identifying alcohol use, which carries a stigma in pregnancy and thus might motivate incomplete disclosure on the part of the patient. While we may have been unsuccessful in our attempt to completely exclude subjects who used moderate or greater alcohol during pregnancy, inclusion of such subjects would have biased our results toward the null and do not detract from our positive finding. We also evaluated ALT independent of other features of metabolic syndrome. Hypertriglyceridemia and increased waist circumference, both closely associated with visceral adiposity, may be associated with GDM.^26
–28 We could not reliably assess triglycerides as our serum samples were not fasting nor did we have measurements of maternal waist circumference.

There is no agreed upon threshold for ALT at which investigation of NAFLD is mandatory. A recent meta-analysis observed a dose–response relationship between ALT in the normal range and risk of metabolic syndrome.²⁹ The limit of 19U/L was proposed by Prati et al. more than 10 years ago as a threshold for women specifically.²⁰ They found that by reducing the lower limit of normal for ALT to 19 U/L in that group, they improved the sensitivity of identification of hepatocellular injury from 56% to 76%, mostly because of increased detection of fatty liver. Their finding has been cited extensively in the hepatology literature and the debate regarding an appropriate upper limit of normal is ongoing. These data support our hypothesis that the relationship between high normal ALT and glucose intolerance may be related to fatty liver physiology.

The novel finding of association between early ALT and GDM in the nonobese population may represent insulin resistance mediated by visceral adiposity rather than excessive peripheral adiposity. The association of visceral obesity with NAFLD is well described.³⁰ While the clinically obese population is at elevated risk due to peripheral insulin resistance (e.g., adipose tissue), the nonobese population may be at risk due to hepatic insulin resistance, consequent to visceral adiposity. However, in the development of GDM, the contribution of hepatic insulin resistance, reflected in elevated ALT, may be diminished among obese women by the impact of excess peripheral adipose tissue. Furthermore, a recent study found that patients with biopsy-proven NAFLD who had a normal BMI had higher median levels of ALT compared to NAFLD patients with elevated BMI.³¹ This phenomenon may also contribute to the strength of association between ALT values and insulin resistance in the normal-overweight group that was not observed in the obese group. Our findings support the idea that GDM in nonobese women may be related to hepatic insulin resistance.

While we controlled for Asian and non-Asian race as a categorical variable in this analysis, the difference in distribution of Asian subjects in the obese and nonobese groups cannot be overlooked. If we categorized these patients as obese using the BMI threshold of 27.5 kg/m² suggested by the World Health Organization, 9 of our 23 Asian subjects would have been categorized as obese rather than the two with BMI >30 kg/m².³² Given the higher prevalence of both GDM and NAFLD in the Asian populations described earlier, it is possible that the additional seven subjects included in the nonobese group contributed a disproportionate amount of pathology.

In conclusion, we found a significant positive association between ALT at the upper limit of normal and the subsequent development of GDM among nonobese women. This may reflect early stages of NAFLD. This finding may have implications not only for the health of a woman's pregnancy—that is, increased risk of GDM—but also for her future health. In women whose BMI does not put them at elevated risk for gestational diabetes, screening with ALT in the first trimester may be beneficial in identifying insulin resistance. Further study is needed to assess the utility and cost-effectiveness of screening women with BMI below 30 kg/m² with an early ALT as a marker for future GDM risk.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Moyer

, U.S. Preventive Services Task Force. Screening for gestational diabetes mellitus: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med, 2014; 160:414.

Gestational diabetes mellitus. Practice Bulletin N. 137. American College of Obstetricians and Gynecologists. Obstet Gynecol, 2013; 122:406–416.

Ohlson

, Larsson

, Bjorntorp

, et al. Risk factors for type 2 (non-insulin dependent) diabetes mellitus. Thirteen and one-half years of follow-up of the participants in a study of Swedish men born in 1913. Diabetologia, 1988; 31:798–805.

Schindhlem

, Diamant

, Dekker

, et al. Alanine aminotransferase as a marker of non-alcoholic fatty liver disease in relation to type 2 diabetes mellitus and cardiovascular disease. Diabetes Metab Res Rev, 2006; 22:437–443.

Utzschnieder

, Kahn

. Review: The role of insulin resistance in nonalcoholic fatty liver disease. J Clin Endocrinol Metab, 2006; 91:4753–4761.

Hagström

, Höijer

, Ludvigsson

, et al. Adverse outcomes of pregnancy in women with non-alcoholic fatty liver disease. Liver Int, 2016; 36:268–274.

World Gastroenterology Organisation.. Global guidelines: Nonalcoholic fatty liver disease and nonalcoholic steatohepatitis,. 2012. Accessed at www.worldgastroenterology.org/assets/export/userfiles/2012_NASH%20and%20NAFLD_Final_long.pdf on March 2, 2015.

Fraser

, Ebrahim

, Smith

. Lawlor

. A comparison of associations of alanine aminotransferase and gamma-glutamyltransferase with fasting glucose, fasting insulin and glycated hemoglobin in women with and without diabetes. Hepatology, 2007; 46:158–165.

Zhang

, Lu

, Hong

, et al. Positive correlations of liver enzymes with metabolic syndrome including insulin resistance in newly diagnosed type 2 diabetes mellitus. Endocrine, 2010; 38:181–187.

10.

Williams

, Stengel

, Asike

, et al. Prevalence of nonalcoholic fatty liver disease and nonalcoholic steatohepatitis among a largely middle-aged population utilizing ultrasound and liver biopsy: A prospective study. Gastroenterology, 2011; 140:124–131.

11.

Vozarova

, Stefan

, Lindsay

, et al. High alanine aminotransferase is associated with decreased hepatic insulin sensitivity and predicts the development of type 2 diabetes. Diabetes, 2002; 51:1889–1895.

12.

Hanley

, Wagenknecht

, Festa

, et al. Alanine aminotransferase and directly measured insulin sensitivity in a multiethnic cohort: The Insulin Resistance Atherosclerosis Study. Diabetes Care, 2007; 30:1819–1827.

13.

McElrath

, Lim

, Rich-Edwards

, et al. Longitudinal evaluation of predictive value for preeclampsia of circulating angiogenic factors through pregnancy. Am J Obstet Gynecol, 2012; 207:407.e1–e7.

14.

Hou

, Shu

, Lu

, et al. Non-alcoholic fatty liver disease's prevalence and impact on alanine aminotransferase associated with metabolic syndrome in the Chinese. J Gastroenterol Hepatol, 2011; 26:722–730.

15.

Schwimmer

, McGreal

, Deutsch

, et al. Influence of gender, race, and ethnicity on suspected fatty liver in obese adolescents. Pediatrics, 2005; 115:e561–e565.

16.

McFarland

, Case

. The relationship of maternal age of gestational diabetes. Diabetes Care, 1985; 8:598–600.

17.

Moranaska

, Jurczyk

, Kloniska

, et al. [Relationship of serum alanine aminotransferase (ALT) to body weight, age, and sex in blood donors population]. Wiad Lek, 2004; 57:427–430.

18.

Anna

, van der Ploeg

, Cheung

, et al. Sociodemographic correlates of the increasing trend in prevalence of gestational diabetes mellitus in a large population of women between 1995 and 2005. Diabetes Care, 2008; 31:2288–2293.

19.

Abell

, Riley

. Hyperemesis gravidarum. Gastroenterol Clin North Am, 1992; 21:835–849.

20.

Prati

, Taioli

, Zanella

, et al. Updated definitions of healthy ranges for serum alanine aminotransferase levels. Ann Intern Med, 2002; 137:1–9.

21.

Clark

, Brancati

, Diehl

. The prevalence and etiology of elevated aminotransferase levels in the United States. Am J Gastroenterol, 2003; 98:961–967.

22.

Sridhar

, Xu

, Darbinian

, et al. Pregravid liver enzyme levels and risk of gestational diabetes mellitus during a subsequent pregnancy. Diabetes Care, 2014; 37:1878–1884.

23.

Tan

, Aziz

, Ismail

, et al. Gamma-glutamyltransferase, alanine transaminase and aspartate transaminase levels and the diagnosis of gestational diabetes mellitus. Clin Biochem, 2012; 45:1192–1196.

24.

Szczepaniak

, Nurenberg

, Leonard

, et al. Magnetic resonance spectroscopy to measure hepatic triglyceride content: Prevalence of hepatic steatosis in the general population. Am J Physiol Endocrinol Metab, 2005; 288:E462–E468.

25.

Graham

, Burke

, Stettler

. Ethnic and sex differences in the association between metabolic syndrome and suspected nonalcoholic fatty liver disease in a nationally representative sample of US adolescents. J Pediatr Gastroenterol Nutr, 2009; 49:442–449.

26.

Baumfeld

, Novack

, Wiznitzer

, et al. Pre-conception dyslipidemia is associated with development of preeclampsia and gestational diabetes mellitus. PLoS One, 2015; 10:e0139164.

27.

Ryckman

, Spracklen

, Smith

, et al. Maternal lipid levels during pregnancy and gestational diabetes: A systematic review and meta-analysis. BJOG, 2015; 122:643–651.

28.

Brisson

, Perron

, Simon-Pierre

, et al. The “hypertriglyceridemic waist” phenotype and glucose intolerance in pregnancy. CMAJ, 2010; 182:E722–E725.

29.

Kunutsor

, Seddoh

. Alanine aminotransferase and risk of the metabolic syndrome: A linear dose-response relationship. PLoS One, 2014; 9:e96068.

30.

Van der Poorten

, Milner

, Hui

, et al. Visceral fat: A key mediator of steatohepatitis in metabolic liver disease. Hepatology, 2008; 48:449–457.

31.

Margariti

, Deutsch

, Manalokopoulos

, et al. Non-alcoholic fatty liver disease may develop in individuals with normal body mass index. Ann Gastroenterol, 2012; 25:45–51.

32.

WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet, 2004; 363:157–163.