Table 1.
Characteristics of prospective studies on caffeine intake in relation to pregnancy loss
| Study | Country | Study design | No. of cases | Total no. of participants | Incidence (%) | Age | Exposure | Method of exposure assessment | Period of exposure assessed | Outcome | Adjustments |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Srisuphan and Bracken (1986)( 38 ) * | USA | Cohort | 68 | 3135 | 2·2 | ≤30 years: 71 % >30 years: 29 % | Caffeine | Interview | First trimester | Spontaneous abortion | Gestational age at interview, maternal age, prior gynaecological surgery, member of Jewish religion, last pregnancy ending with a spontaneous abortion (maternal smoking was not adjusted for as it did not significantly predict the outcome) |
| Wilcox et al. (1990)( 39 ) | USA | Cohort | 43 | 171 | 25·1 | <29 years: 49 % ≥29 years: 51 % | Caffeine | Interviewer-administered questionnaire | Pre-pregnancy | Spontaneous abortion (<42 d) | Maternal age (only minor change after adjustments for smoking, alcohol intake) |
| Mills et al. (1993)( 40 ) | USA | Cohort | 59 | 423 | 13·9 | <30 years: 48 % ≥30 years: 52 % | Caffeine | Interview | First trimester | Spontaneous abortion | Smoking, maternal age, parity, prior spontaneous abortion, alcohol use, maternal education, income |
| Dlugosz et al. (1996)( 41 ) | USA | Cohort | 135 | 2967 | 4·6 | Mean 31 years | Caffeine | Interviewer-administered questionnaire | First month of pregnancy | Spontaneous abortion | Maternal age, gestational stage at interview, smoking, alcohol intake |
| Fenster et al. (1997)( 42 ) | USA | Cohort | 499 | 5144 | 9·7 | Mean 28 years | Caffeine | Interview (telephone) | Pre-pregnancy, first trimester† | Spontaneous abortion | Maternal age, smoking, alcohol intake, gestational age, pregnancy history, ethnicity, employment, marital status, socio-economic status |
| Klebanoff et al. (1999)( 23 ) | USA | Nested case–control | 591 | 3149 | 18·8 | Cases: mean 27 years Controls: mean 25 years | Paraxanthine | Serum paraxanthine | NA but before the occurrence of spontaneous abortion | Spontaneous abortion | Smoking, maternal age, ethnicity |
| Wen et al. (2001)( 43 ) | USA | Cohort | 75 | 650 | 11·5 | Median 29 years | Caffeine | Self-administered FFQ | Pre-pregnancy, first trimester† | Spontaneous abortion | NA (only unadjusted RR presented as they found no important confounding) |
| Tolstrup et al. (2003)( 44 ) | Denmark | Nested case–control | 303 | 1684 | 18·0 | Mean age 24·5–26·0 years according to caffeine consumption | Caffeine | Self-administered FFQ | Pre-pregnancy | Spontaneous abortion | Maternal age, marital status, smoking, alcohol intake |
| Wisborg et al. (2003)( 26 ) | Denmark | Cohort | 82 | 18 478 | 0·4 | <30 years: 57 % ≥30 years: 43 % | Coffee | Self-administered questionnaire | First trimester | Stillbirth | Smoking, alcohol intake, parity, maternal age, marital status, education, employment status, maternal BMI |
| Bech et al. (2005)( 12 ) | Denmark | Cohort | 1102 | 88 482 | 1·2 | <30 years: 54 % ≥30 years: 46 % | Coffee | Interview (telephone) | First to early second trimester | All pregnancy loss†, stillbirth | Maternal age, parity, smoking, pre-pregnancy BMI, alcohol intake, socio-occupational status |
| Savitz et al. (2008)( 45 ) | USA | Cohort | 258 | 2407 | 10·7 | <30 years: 61 % ≥30 years: 39 % | Caffeine | Interview (telephone) | Pre-pregnancy, 4 weeks post LMP, and before 16 weeks’ gestation† | Spontaneous abortion | Maternal age, ethnicity, education, marital status, alcohol use, vitamin use, symptoms of nausea and vomiting during early pregnancy (excluding smoking did not change results substantially) |
| Weng et al. (2008)( 46 ) | USA | Cohort | 172 | 1063 | 16·2 | <30 years: 42 % ≥30 years: 58 % | Caffeine | Interview | Early pregnancy (soon after pregnancy is confirmed) | Spontaneous abortion | Maternal age, ethnicity, education, family income, marital status, previous miscarriage, nausea and vomiting since LMP, smoking, alcohol intake, Jacuzzi use, exposure to magnetic fields |
| Greenwood et al. (2010)( 47 ) | UK | Cohort | 28 | 2635 | 1·1 | Mean 30 years | Caffeine | Self-administered questionnaire (validated) | First trimester | Late spontaneous abortion (12–24 weeks) and stillbirth | Maternal age, parity, smoking (cotinine concentration), alcohol intake |
| Pollack et al. (2010)( 36 ) | USA | Cohort | 14 | 68 | 20·6 | NA | Caffeine | Daily diaries | Pre-pregnancy | All pregnancy loss | Maternal age, alcohol intake, smoking |
NA, not available; LMP, last menstrual period; RR, relative risk.
*
Included for qualitative review only.
†
Data used for the main analysis.