Maternal coffee intake and the risk of threatened abortion

Background Maternal caffeine intake has an effect on fetal growth through cross the placenta freely. The objective of this study is to examine the association of pre-pregnancy coffee consumption pattern with the risk of threatened abortion. Methods A cross-sectional analysis was conducted for a total of 4,111 pregnant women who completed baseline examination for the Korean Pregnancy Outcome Study between 2013 and 2017. Coffee consumption pattern before pregnancy was measured by a questionnaire. The frequency of coffee consumption was divided into seldom (<1 cup/week), light (<1 cup/day), moderate (1 cup/day), and heavy coffee drinkers (≥2 cups/day). Threatened abortion was defined as the occurrence of vaginal bleeding in the first 20 weeks of pregnancy. Multiple logistic regression models were applied to examine the association between pre-pregnancy coffee consumption and the risk of threatened abortion, after adjusting for age, body mass index, systolic blood pressure, smoking, drinking, physical activity before and after pregnancy, stress level, history of depression, and reproductive information. Results Women with heavy coffee drinking were at increased odds for having threatened abortion in unadjusted model (odds ratio OR, 1.271; 95% confidence interval CI, 1.008 to 1.602). Compared to women with seldom coffee drinkers, the fully adjusted ORs for the risk of threatened abortion were significant in moderate (OR, 1.420; 95% CI, 1.001 to 2.014) and heavy coffee drinkers (OR, 1.750; 95% CI, 1.156 to 2.651). Conclusions Heavy coffee drinking was independently associated with a higher risk of threatened abortion among Korean pregnant women.

According to Korean National Health and Nutrition Examination Survey, the prevalence of daily coffee drinking (1 or more cups/day) greatly increased from 54.6% in 2001 to 65.3% in 2010 -2011 in Korean adults [4]. The average coffee consumption among Korean adults is high as 11.3 times per week recently, which is more than five-times greater compared to other Asia-Pacific region [3]. Coffee is composed of many physiologically active substances especially, caffeine is an important component of coffee [5,6]. Other caffeinated beverages or foods do not significantly contribute to daily caffeine intake in Korean[7]. Therefore, it is important to examine the effect of coffee consumption on health.
Caffeine is the most generally consumed xenobiotic in pregnant women, with the potential to adverse effects in the development of the fetoplacental. Maternal caffeine intake during pregnancy has an effect on fetal growth through cross the placenta freely and decrease blood flow to the placenta, although the specific effects on a fetus are inconsistent [8,9].
However, it is still unclear whether coffee consumption is an independent risk factor for threatened abortion, which is the most common complication with an incidence of 15-20% of ongoing pregnancies [10,11]. Threatened abortion may suggest an underlying placental dysfunction, which could induce pregnancy complications in later pregnancy [12,13].
There is continuous concern that coffee intake would increase particularly in pregnant women, which would result in adverse health effects. Therefore, the aim of this study is to evaluate the coffee consumption pattern before pregnancy and its association with the risk of threatened abortion among Korean pregnant women.

Study participants
Data for the present study were derived from Korean Pregnancy Outcome Study (KPOS), a prospective cohort study. Between March 2013 and January 2017, all pregnant women who visited Cheil General Hospital and CHA Hospital for antenatal care during the first trimester were asked to participate in the KPOS study. Women were excluded from enrolment if they were not Korean or were pregnant with triplets or higher-order multiple gestations. A total of 4,195 pregnant women were included in the baseline analysis after excluding 10 foreigners, 218 individuals who declined to participate in the study, and 114 individuals who were missing key variables. Gestational age was determined by the last menstrual period in naturally conceived women and was confirmed on first trimester ultrasound. After the first antenatal visit, eligible participants were requested to complete several sets of questionnaires or examinations at each visit: Visit 1 in the 1st trimester (around 12 weeks of gestation); Visit 2 in the 2nd trimester (around 24 weeks of gestation); Visit 3 in the 3rd trimester (around 36 weeks of gestation); Visit 4 at delivery; Visit 5 at 4-6 weeks after birth.
After excluding 84 individuals with missing dietary questionnaire from baseline participants, a cross-sectional analysis was conducted for 4,111 women aged 18-46 years old. Trained research nurses explained the study in detail and obtained written informed consents and completed questionnaires. All participants provided written informed consent, and the study protocol was approved by the Institutional Review Board (IRB) of Cheil General Hospital (IRB number: CGH-IRB-2013-10) and CHA Hospital (IRB number: 2013-14-KNC13-018), separately. It was emphasized to all participants that they were free to withdraw from any part of the study at any point in time.

Measurements
A face-to-face interview was conducted to evaluated participants' socio-demographic profiles, medical and family history, reproductive information, health-related behaviors, and psychological health. Participants completed the clinical and laboratory measurements in pregnancy, which included anthropometric measurements, blood pressure measurements, and blood and urine laboratory tests. Socio-demographic status included age, educational level, household income, employment status, marital status, cohabiting family composition, and spouse information. The family history information included hypertension, diabetes, GDM, preeclampsia, depression, and other mental illness. Cigarette smoking, alcohol intake, and supplement intake were evaluated at each visit. Physical activity was assessed at each visit using the short form of the International Physical Activity Questionnaire (IPAQ).
Dietary intake patterns were evaluated with a food frequency questionnaire (FFQ). The level of coffee consumption pattern before the conception was estimated with the question "How often did you drink a coffee before the pregnancy?" on the FFQ. Coffee consumption was categorized into 5 groups (seldom, 2-3 cups/week, 4-6 cups/week, 1 cup/day, and 2 or more cups/day). In this analysis, participants were divided into four groups based on their reported amount of coffee consumption: ≥2 cups/day, "heavy coffee drinkers"; 1 cup/day, "moderate coffee drinkers"; <1 cup/day, "light coffee drinkers"; and <1 cup/week, "seldom coffee drinkers". The type of coffee was divided into black coffee, black coffee with sugar, black coffee with creamer, and instant coffee mix (instant coffee with creamer and sugar).
We obtained information on antenatal pregnancy complications and delivery details during pregnancy and postpartum. First trimester complications including emesis and threatened abortion were assessed at visit 1. In this study, threatened abortion was defined as the occurrence of vaginal bleeding in the first 20 weeks of pregnancy, determined via ultrasonographic examinations by physician [14,15]. The delivery outcomes included gestational age at delivery, type of labour (induced or spontaneous), type of delivery, indication for Cesarean delivery, and delivery complications.
Blood pressures were measured at every visit using the automatic oscillometric technique, but a diagnosis of hypertensive disorders of pregnancy was confirmed by manual measurements using blood pressure cuffs and auscultation. Blood samples and placenta were stored in -70 ºC freezers at a controlled temperature and humidity. All biological samples were marked with barcodes and stored in the National Biobank of Korea. We uploaded the data from all questionnaires and examinations to a web-based clinical data management system (iCReaT) managed by the Korea National Institute of Health.

Statistical analysis
We classified participants on the basis of the frequency of coffee consumption. For some analyses, the lower categories of exposure variables were combined into a single stratum because of the small numbers. We summarized the general characteristic of study participants with means with standard deviation for continuous variables and observed numbers with percentages for categorical variables. To assess the statistical differences, one-way analysis for variance and chi-square test were used for continuous and categorical variables, respectively. Multivariate logistic regression analysis was used to estimate odds ratios (OR) with 95% confidence intervals (CI) for the association of coffee consumption with threatened abortion. Age, body mass index, systolic blood pressure, cigarette smoking, alcohol consumption, parity, physical activity before pregnancy, current physical activity, stress, history of depression, and the number of livebirth, stillbirth, miscarriage, and abortion were considered as covariates in the adjusted model.
In additional sensitivity analyses, we conducted an additional multiple logistic regression analysis to estimate the association of the type of coffee and threatened abortion. All statistical analyses were performed using SAS software (version 9.4, SAS; NC, USA) and statistical significance was defined as a two-sided p-value of less than 0.05.

Results
The baseline characteristics of study participants are presented according to the frequency of coffee consumption pattern before pregnancy in Table 1. Of the 4,111 participants, there were 732 heavy coffee drinkers (≥2 cups/day, 17.8%), 1,405 moderate coffee drinkers (1 cup/day, 34.2%), 699 light coffee drinkers (<1 cup/day, 17.0%), and 1,275 seldom coffee drinkers (31.0%). The mean age of total pregnant women was 33.2 years old. Participants with heavy coffee drinking were more likely to have an older age, higher educational ascertainment, economic status, and body mass index, and higher frequencies of history of depression, former smoker, and former drinker. The overall prevalence of threatened abortion in this pregnant women was 18.1%. As shown in Figure   1, heavy coffee drinkers showed the highest prevalence of threatened abortion, especially they need a drug or inpatient treatment. Table 2 showed the result of the multiple logistic regression analysis of the association between coffee consumption and the risk of threatened abortion. The highest amount of coffee consumption group was significantly associated with high risk of threatened abortion in the unadjusted model. After additional adjustment for age, body mass index, systolic blood pressure, cigarette smoking, alcohol consumption, physical activity before pregnancy, current physical activity, stress, history of depression, and the number of parity, livebirth, stillbirth, miscarriage, and abortion, the associations remained significant. Compared to women with seldom coffee drinkers, the adjusted OR for threatened abortion was 1.101, 1.420, and 1.750 for a pre-pregnancy intake on light coffee drinkers, moderate coffee drinkers, and heavy coffee drinkers, respectively. In additional sensitivity analysis in Supplementary Table 1 and 2, there were no significant associations between the type of coffee with additives and the risk of threatened abortion.

Discussion
We investigated the coffee consumption pattern before pregnancy and its association with the risk of threatened abortion in Korean pregnant women. Herein, habitual coffee consumption 1 or more cup of coffee per day before pregnancy was significantly associated with an increased risk of threatened abortion during early pregnancy even after adjustment for cigarette smoking and alcohol consumption; however, the type of coffee did not significantly affect the risk of threatened abortion.
Pregnancies with the risk of threatened abortion in the first trimester occur about 15-20% of all pregnancies, however, there is still insufficient information about its risk factor and the way of prevention [16,17]. In the present of 4,111 pregnant women, although the average maternal age of participants was relatively high, the overall prevalence of threatened abortion was 18.1%, similar to previous results. Since threatened abortion has been shown to be associated with an increased risk of poor pregnancy outcomes, approximately 50% of pregnant women who experienced threatened abortion eventually suffer miscarriages, it is required to be diagnosed and managed prevent maternal of fetal mortalities and morbidities [12,18,19].
Even though the market for caffeinated beverages has increased in the past decades, the most frequently consumed caffeine beverage is coffee [1,7]. A standard cup of coffee generally expected to provide 100 mg of caffeine, however, it varies according to portion Recently, some epidemiologic studies have found a significant association between 300 mg or more of caffeine intake per day and the risk of spontaneous abortion [21,28,29].
Consistent with previous studies, we have found that pregnant women with heavy coffee drinking were significantly at increased risk for having threatened abortion. In Chinese prospective study, caffeine intake before pregnancy did not increase the risk of spontaneous abortion, but caffeine intake more than 300 mg/day during the first trimester appeared to be significantly increased the risk [21]. The UK case-control study showed that caffeine consumption more than 300 mg/day during pregnancy about doubles the risk of miscarriage and its effect is driven by coffee consumption [28]. Similarly, adjusted risk among Danish women who consumed caffeine more than 375 mg/day was 2.21[29].
One meta-analysis has evaluated the risk of pregnancy loss increased by 3% for every There are several limitations to this study that should be noted. The main limitation of the present study relates to inaccurate assessment of coffee consumption pattern because a 'cup' of coffee varies according to portion size, brewing method, and brand type. Despite the huge popularity of decaffeinated coffee in the market, we cannot measure its amount.
In future studies, objective measurements including combine biomarkers of caffeine exposure in blood, urine, and saliva with 24-hour dietary recall measurement should be needed to assess quantity of precise coffee consumption. Second, typical recall bias due to retrospective assessment of caffeine consumption should be considered. However, we may reduce the impact of recall bias since we measured assessment of coffee consumption before the onset of threated abortion. Third, coffee consumption pattern before the pregnancy at only one single time point may not reflect chronic exposure over the years because women who prepare conceive tend to keep their eating habits healthy before become pregnant. Finally, although we controlled for several potential confounders in our analysis, residual confounding by diet or other lifestyle factors effects may have been present.
As the younger women continue to experience a western coffee culture and demand high quality coffee beans, this trend is expected to continue. The properties of coffee are regard as a double-edged sword and the balance between beneficial and harmful impacts on health. In this study, a high consumption of coffee to pregnancy seems to be associated with an increased risk of threatened abortion, whereas the type of coffee does not influence the risk. Among women who experienced threatened abortion during followup, a larger fraction was habitual coffee drinkers before the pregnancy compared with women who were seldom coffee drinkers.

Conclusions
As coffee consumption is a potentially modifiable risk factor, caffeine intake before conception and during pregnancy should be reduced. Our study may provide useful  and it was emphasized to all participants that they were free to withdraw from any part of the study at any point in time.

Consent for publication
Not applicable.

Availability of data and material
The KPOS is being conducted mainly at the Cheil General Hospital & Women's Healthcare Center and CHA Gangnam Medical Center, where the staff are responsible for the collection, management, and distribution of the data. All data are stored electronically in anonymous format and are currently only available to the KPOS researchers; however, data analysis collaborations may be possible through specific research proposals. Future information can be requested by e-mailing the principal investigator (hmryu2012@naver.com).

Competing interests
The authors declare that they have no conflict of interest.   Figure 1 Prevalence and severity of threatened abortion by coffee consumption