The mean prepregnancy intake of caffeine was similar in the women who had spontaneous abortion and the control subjects (346 mg and 329 mg per day, respectively; P=0.20). The women who had spontaneous abortion were significantly older than the control subjects, were more likely to have been born outside the Nordic countries (Sweden, Denmark, Norway, Finland, and Iceland), and were more likely to have had previous pregnancies and previous spontaneous abortions (P (Table 1).
During early pregnancy, the women who had spontaneous abortion had a significantly higher caffeine intake than the control subjects (P Nausea (with or without vomiting) and tiredness as symptoms of pregnancy were more prevalent and severe among the control subjects (P differences between the two groups with regard to education, prepregnancy body-mass index, participation in shift work, or use of vitamin supplements during pregnancy (data not shown).
The severity of nausea (according to the mean severity score) and the mean daily intake of caffeine in the women with spontaneous abortion and the control subjects are shown in Figure 1 as a function of the week of gestation. At four weeks after the last menstrual period, there was a marked decrease in the ingestion of caffeine in both groups, but the decline was much more pronounced in the control group. The declining intake of caffeine coincided with an increase in the proportion of women who had nausea, which was also more pronounced among the control subjects than among the women who had spontaneous abortion.
Figure 1. Mean Intake of Caffeine and Mean Severity of Nausea According to the Week of Gestation. Week 0 was the week of the last menstrual period. The case patients were women who had spontaneous abortion at 6 to 12 weeks of gestation. The control subjects did not have spontaneous abortion.
In multivariate analyses, which included caffeine intake, smoking status, age, number of previous pregnancies, history of spontaneous abortion, consumption of alcohol, and presence or absence of nausea, vomiting, and fatigue, the adjusted odds ratios for spontaneous abortion in women who ingested at least 100 mg of caffeine per day, as compared with women who ingested less than 100 mg of caffeine per day, were as follows: 100 to 299 mg per day — odds ratio, 1.3 (95 percent confidence interval, 0.9 to 1.7); 300 to 499 mg per day — odds ratio, 1.5 (95 percent confidence interval, 1.0 to 2.1); and 500 mg or more per day — odds ratio, 1.4 (95 percent confidence interval, 0.9 to 2.2) (P for trend = 0.05). For smokers as compared with nonsmokers, the adjusted odds ratio was 1.5 (95 percent confidence interval, 1.1 to 2.1). However, there was a significant interaction between caffeine ingestion and smoking with regard to the risk of spontaneous abortion (P stratified according to smoking status. Among smokers, the ingestion of caffeine was not associated with an excess risk of spontaneous abortion, whereas among nonsmokers, high intake of caffeine (at least 500 mg per day) was associated with a doubling in risk (Table 2). A repetition of the adjusted regression analysis to include control subjects in whom abortions were induced had little effect on the results (data not shown).
With regard to symptoms of pregnancy, we did not include aversion to coffee in the multivariate analysis because of its high degree of inverse correlation with caffeine ingestion. When we restricted the analysis to nonsmokers who reported no such aversion, the caffeine-related risk of spontaneous abortion increased (Table 2). Moreover, because coffee was the main source of caffeine, the possibility of confounding by other constituents in coffee also existed. When we assessed the risk of spontaneous abortion according to the level of caffeine ingestion among non–coffee drinkers and nonsmokers, the caffeine-related odds ratios increased, but the confidence intervals were wide. For smokers, there was no effect of caffeine ingestion in any of these analyses (data not shown).
The success rate of karyotyping was highly influenced by the week of gestation. In weeks 6 to 8, the fetal karyotype was known in only 25 percent of all spontaneous abortions (17 normal and 23 abnormal karyotypes), whereas in weeks 11 to 12, the corresponding figure was 57 percent (38 normal and 66 abnormal). The ingestion of caffeine was unrelated to the occurrence of spontaneous abortion of a fetus with an abnormal karyotype (Table 3). Among nonsmokers, the ingestion of a moderate or high level of caffeine was associated with an increased risk of spontaneous abortion of fetuses with normal as well as unknown karyotypes (Table 3). Among smokers, a high intake of caffeine was associated with an increased risk of spontaneous abortion of a fetus with a normal karyotype, but the confidence intervals were wide.
Our principal analyses were based on levels of caffeine ingestion that were averaged over the first trimester of pregnancy. When we disregarded caffeine intake and pregnancy-related symptoms during the last two completed weeks of gestation before the interview, the caffeine-related risk of spontaneous abortion of a fetus with a normal karyotype among nonsmokers was unchanged for moderate levels of in-gestion of caffeine but was attenuated for the highest level of ingestion (at least 500 mg per day): odds ratio, 1.4 (95 percent confidence interval, 0.5 to 3.7).