Department of Family & Preventive Medicine, University of California, San Diego, CA 92093, USA
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: assisted reproduction technologies/caffeine/IVF live birth/multiple gestations/pregnancy
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Caffeine use in humans
Epidemiological studies have been advocated as the best approach to exploring any link between caffeine and reproductive outcome (Olsen, 1991). Caffeine has been implicated as a risk factor for delayed conception (Wilcox et al., 1988
; Williams et al., 1990
; Hatch and Bracken, 1993
; Stanton and Gray, 1995
; Bolumar et al., 1997
; Jensen et al., 1998
). In a cohort of 104 healthy women who attempted to become pregnant for 3 months, the high caffeine consumers (i.e.
1 cup of brewed coffee/day) were significantly less likely to become pregnant in each cycle [fecundability ratio: 0.53, 95% confidence interval (CI): 0.35, 0.79] than the lower caffeine consumers (i.e. <1 cup of coffee/day), while adjusting for age, frequency of intercourse, age at menarche, smoking, and a woman's prenatal exposure to her mothers' smoking (Wilcox et al., 1988
). The high caffeine users had an increased risk of infertility (unadjusted relative risk: 4.7, P < 0.005); there was a dose-related response to caffeine consumption when women were sub-divided into narrower categories of caffeine consumption, with women in the highest consumption group having a fecundability rate only 26% of that of the lowest consumption group. Only two studies failed to find a correlation between caffeine consumption and delayed conception (Joesoef et al., 1990
; Olsen, 1991
); both were retrospective and subject to recall bias.
Caffeine consumption has also been implicated as a risk factor for spontaneous abortions (Srisuphan and Bracken, 1986; Fenster et al., 1991
; Armstrong et al., 1992
; Infante-Rivard et al., 1993
). Some studies found no relationship (Wilcox et al., 1990
; Mills et al., 1993
), and it was suggested that nausea (occurring more frequently in the first trimester of pregnancy), may have resulted in decreased caffeine consumption, thereby creating a spurious finding of no association (Stein and Susser, 1991
). However, a study which adjusted for nausea found a doubled risk for spontaneous abortion in heavy caffeine users [adjusted odds ratio (OR): 2.10, 95% CI: 1.20, 3.70] (Fenster et al., 1991
).
There is a relationship between caffeine and birthweight; a combined analysis of mean birthweight from 22 studies resulted in a significant decrease in birthweight of nearly 43 g among newborns of the heaviest-caffeine-consuming mothers (Santos et al., 1998a).
Caffeine consumption and male reproduction
In 5476 subjects from an infertility clinic, coffee consumption was associated with increases in sperm concentration, abnormal forms, and motility. The combination of coffee drinking with smoking diminished sperm motility and increased the percentage of dead sperm (Marshburn et al., 1989).
Caffeine consumption and assisted reproductive technologies
Caffeine has been used in-vitro in an attempt to stimulate hamster sperm motility. However, the results of these studies were conflicting. In a number of studies, the addition of caffeine to medium increased the motility of cryopreserved sperm (Barkay et al., 1977; Harrison, 1978
; Aitken et al., 1983
; Hammitt et al., 1989
). In another study, an initial increase in sperm motility was reported with caffeine in the medium; however, after removing caffeine, a significant reduction in motility was observed at 48 h, as well as a reduction in the percentage of zona-free hamster oocytes penetrated, suggesting a possible toxic effect of caffeine (Hammitt et al., 1989
). Yet another study reported increased sperm motility following incubation of human semen with caffeine, but no pregnancies were achieved among the 199 artificial inseminations, compared with one pregnancy in the control group of 201 inseminations. Sperm treated with caffeine had greater morphological surface damage than placebo-treated sperm (Harrison et al., 1980
). Only one study examined the effect of caffeine on the ability of sperm to fertilize mature human oocytes; there was an improvement in sperm motility, but not in fertilization rates, with an actual decrease in fertilizing ability and embryonic development at the highest concentrations of caffeine in culture medium (Imoedemhe et al., 1992
).
There are no studies to date that have investigated the effect of caffeine consumption by men and women on success rates of IVF and gamete intra-Fallopian transfer (GIFT). The unique contributions of this study are: (i) it is prospective; (ii) it examines caffeine consumption on the success rates of all endpoints of IVF and GIFT, including live-birth deliveries (n = 41); and (iii) it includes detailed caffeine histories from both female and male partners.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The study sample is representative of the age, race, and education level of couples enrolling in IVF and GIFT programmes in Southern California. Our population's demographics reflected those of the USA, whereby the proportions of fertility-impaired females who ever received infertility services were generally highest among those females who were older, who had graduated from college, who had a high income, and who were non-Hispanic White (Stephen and Chandra, 2000).
IVF and GIFT
The IVF and GIFT procedures consisted of the following steps: (i) a baseline ultrasound; (ii) ovarian stimulation with GnRH agonist followed by stimulation with gonadotrophin and HCG; (iii) after 5 days of ovarian stimulation with gonadotrophin, follicular development was monitored using two modalities: ultrasound and hormones to measure estradiol; (iv) oocyte retrieval from the follicle by transvaginal aspiration with ultrasound guidance for IVF or by laparoscopy for GIFT; (v) for IVF, the oocyte then matured in a culture medium for 58 h; (vi) collection of sperm; (vii) addition of sperm to the oocytes ~6 h after oocyte collection, and confirmation of fertilization by microscopic examination for two pronuclei 1620 h later; (viii) embryo transfer to the uterus either 4872 h after oocyte aspiration. For GIFT after the oocytes were identified from the follicular fluid, they were immediately transferred into the distal Fallopian tube along with the sperm.
All selected centres had similar technical procedures and pregnancy rates following IVF, so that these factors would not confound the relationship between caffeine and success rates. Data on embryo transfer was difficult to assess, because not all available embryos were transferred.
Data collection
Institutional approval and informed consent were obtained from all couples. Women filled out three questionnaires, one during the first clinic visit, another at the time of the procedure and the last after the pregnancy outcome. Women who never achieved or completed a pregnancy also received a third questionnaire by mail, usually 1 month after the unsuccessful outcome. Men filled out two questionnaires, one during the initial visit and the other during sperm collection. Information on demographics, obstetrical history, past infertility evaluation, medical history and medications, lifestyle habits (e.g. smoking, alcohol, illicit drug use, caffeine), environmental and occupation exposures, and psychosocial stresses, was included. Medical records were reviewed to validate medical and obstetrical information, and infant characteristics (e.g. multiple gestations, gestational age, birthweight) obtained from the questionnaires, as well as to confirm type, dose, and timing of hormones, culture medium, and number of oocytes retrieved or fertilized, number of embryos transferred, and sperm parameters.
Baseline information, including the `usual caffeine intake' in their lifetime and intake `during the week of the initial clinic visit`, was collected using the first questionnaire during step 1 of the assisted reproduction procedure. The second questionnaire established caffeine intake `during the week before the procedure' (step 2) and the `week of the procedure' (steps 38). The women's final questionnaire retrospectively assessed caffeine intake during each trimester of pregnancy; if the assisted reproduction procedure did not result in a live birth, information was ascertained retrospectively from the time of the procedure until the negative pregnancy outcome.
Definition of caffeine consumption
Male and female participants were questioned about the type (i.e. caffeinated or decaffeinated coffee, tea, soft drinks, cocoa drinks, milk chocolate, and dark chocolate) and amount (i.e. number of cups, glasses, or ounces per day and/or week) of caffeine consumed during the various time periods. Caffeine intake for both partners was converted to exact amounts in mg, based on estimates reported in several previous studies (Wilcox et al., 1988; Joesoef et al., 1990
; Grodstein et al., 1993
; Stanton and Gray, 1995
). Total caffeine consumption was estimated by assigning 100 mg per cup of coffee, 50 mg per cup of tea, and 50 mg per can of soda. Decaffeinated coffee, soda, and tea were estimated as 2 mg of caffeine per cup or glass. One cup of hot cocoa contained 4 mg of caffeine, while one ounce of milk or dark chocolate contained 7.10 or 18.6 mg respectively. For each time period, the total amount of caffeine was calculated by summing the amounts consumed from all food products containing caffeine.
In this study, caffeine consumption was evaluated as continuous and categorical predictors, with cut-off values based on previous studies (Srisuphan and Bracken, 1986; Fenster et al., 1991
), and data distributions for women and men by
or 1 cup/day increments as reported in previous studies. For women, the reference group consumed a light amount of 02 mg/day, equivalent to the amount in one decaffeinated beverage; caffeine intake of >250 mg/day was classified as moderate and >50 mg/day as high. The only exception to this categorization was during pregnancy, because very few women reported consuming caffeine (02 mg/day, >2 mg/day). For men, consumption of 050 (reference category), >50200, and >200 mg/day represented light, medium, and high caffeine intake respectively.
Originally, subjects completed questionnaires which asked questions only about coffee, decaffeinated coffee, tea, and soda. Subsequent questionnaires were updated with a distinction made between caffeinated and decaffeinated tea and soda.
Reproductive outcomes
The impact of caffeine consumption was measured at various endpoints of IVF and GIFT including the sperm profile, the number of ooyctes retrieved and fertilized, number of embryos transferred, occurrence of a pregnancy, multiple gestations, miscarriage, live-birth delivery, and infant characteristics (i.e. gestational age, birthweight). With GIFT, fertilization was inferred from a positive pregnancy test because of the inability to directly observe fertilization.
Sperm concentration (x106/ml), motility (%) and morphology (% normal) were examined as continuous, as well as dichotomous, outcomes. Low sperm concentration (<20x106/ml), low motility (<50%) and low normal morphology (<30%) were based on published guidelines (World Health Organization, 1992).
Biochemical pregnancy was diagnosed by an elevated serum HCG obtained at 1620 days after oocyte retrieval, followed by a decline and disappearance of measurable HCG prior to the gestational sac appearing on ultrasound; women with this diagnosis were not classified as pregnant (Jones et al., 1983). Clinical pregnancy was defined as an elevated serum HCG on successive occasions measured every 57 days, and ultrasound verification of at least one gestational sac (Williams, 1985
). Multiple gestations referred to more than one gestational sac or fetus (Williams, 1985
).
Infant gestational age (weeks) and birthweight (g) were also examined as continuous and dichotomous outcomes. A term delivery indicated a pregnancy of >37 weeks but <42 weeks (Williams, 1985). A low birthweight was defined as <2500 g (Medical Research International, 1989).
Statistical analyses
The relationship between caffeine intake and IVF or GIFT outcome was examined in univariate and multivariate analyses, simultaneously adjusting for potential confounders chosen from previous literature. Linear regression was used in analyses involving the following continuous outcomes: sperm parameters, number of ooyctes retrieved and fertilized, number of embryos transferred, and infant gestational age. To normalize the distributions, the square root transformation of sperm count and the natural logarithmic transformation for the number of oocytes aspirated and fertilized were used. Logistic regression was used to evaluate dichotomous outcomes (e.g. pregnancy, multiple gestation miscarriage and live-birth delivery). It was employed as a result of the need to simultaneously adjust for potential confounders (e.g. smoking, alcohol use, age, race, years of schooling, parity, type of infertility, type of procedure, number of attempts, and number of good quality embryos transferred). Because logistic regression was used for this study, odds ratios (OR) were reported; relative risks could have been used; however, it would have been difficult to calculate the CI (Woodward, 1999). Therefore, OR were provided, even though they tend to overestimate the magnitude of association when the outcomes are relatively common (Woodward, 1999
).
Confounders were included in the final models if they changed the magnitude of association between caffeine consumption and particular outcomes by 15%. Multiplicative interactions (e.g. between caffeine intake and either maternal age, number of attempts, alcohol, smoking, tubal infertility, or endometriosis) were assessed through significance testing using an
level of 0.05.
The total amounts (mg) of caffeine consumed during specified time periods were analysed first as continuous predictors, and then, if linear relationships did not exist, as categorical predictors (e.g. 02, >250 and >50 mg/day for women). Secondary analyses were performed separately for couples undergoing IVF and GIFT for the first time. In addition, women aged >35 years were analysed separately because of an increased incidence of miscarriage and a higher risk of chromosomal abnormality. Results with high measures of associations (OR >2.0) or overall P-values 0.05 were presented with corresponding 95% CI. All analyses were performed using STATA software (Stata Corp., College Station, TX, USA).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Female and male caffeine intake
Women consumed less caffeine than men during all time periods (Table II). Women reported caffeine consumption during their lifetime (94%, mean: 106 mg/day), during the week of the initial clinic visit (83%, mean: 72 mg/day), 1 week prior to the IVF or GIFT procedure (49%, mean: 66 mg/day), and during the week of the procedure (37%, mean: 40 mg/day). Of the 51 pregnant women with caffeine information, 39% (mean: 16 mg/day), 36% (mean: 17 mg/day), and 36% (mean: 28 mg/day) reported caffeine consumption during the first, second, and third trimesters respectively.
|
For all time periods, coffee was the dominant source of caffeine for both women and men.
Female caffeine intake and aspiration, fertilization, transfer, or the pregnancy period
For all time periods (i.e. prior to or during the procedure), there was no statistically significant effect for female caffeine consumption on oocyte retrieval, fertilization, embryo transfer, or pregnancy. Because of the inability to observe fertilization with GIFT, this may have resulted in non-differential misclassification, biasing results toward the null. Because few women consumed caffeine during pregnancy, non-significant results may be due to insufficient power.
Female caffeine intake and occurrence of miscarriages
Seventy-one women became pregnant, but knowledge of miscarriage status was known for only 62 women, 21 of whom were known to have a miscarriage. In multivariable analyses, adjusting for the most important confounders in our data (maternal age, type of infertility, type of procedure, and number of assisted reproduction attempts), `usual caffeine intake' of >250 and >50 mg/day, compared with 02 mg/day, yielded OR for miscarriage of 19.8 (95% CI: 1.3, 300.9) and 10.5 (95% CI: 0.9, 125.3) respectively. The adjusted OR for intake of >250 and >50 mg/day during the `week of the initial clinic visit' were 9.5 (95% CI: 1.4, 62.0) and 6.2 (95% CI: 0.9, 40.8) respectively. Because the sample size was small, the CI were very wide and the magnitude of association for caffeine and miscarriage may be unreliable. First trimester caffeine intake was not related to miscarriages, which may be a result of rare female caffeine intake in pregnancy. When subjects with unknown miscarriage status were classified as either (i) all experiencing a miscarriage, or (ii) all not experiencing a miscarriage, the direction of the association and significance levels were similar to the results when subjects with unknown miscarriage status were omitted from the analyses.
Female caffeine intake and live-birth delivery
In our study, `not having a live birth' resulted from either not becoming pregnant or experiencing a miscarriage. Female caffeine consumption was a strong risk factor for not achieving a live birth, even after controlling for potential confounders (Table III). `Usual intake' of >250 or >50 mg/day in their lifetime, compared with the minimal intake of 02 mg/day, increased the odds of not having a live birth by 3.1 (adjusted 95% CI: 1.0, 9.7, P = 0.05) and 3.9 (adjusted 95% CI: 1.3, 11.6, P = 0.01) times respectively. Caffeine intake of >250 or >50 mg/day during the `week of the initial clinic visit', increased the odds of no live birth by 2.9 (adjusted 95% CI: 1.1, 7.5, P = 0.01) and 3.8 (adjusted 95% CI: 1.4, 10.7, P = 0.01) times respectively. The relationship between caffeine intake and live births remained essentially unchanged when miscarriages were removed from the analyses.
|
The adjusted OR were further elevated in sub-analyses excluding women who were undergoing repeat IVF/GIFT attempts (OR: 4.9 and 8.4 respectively, for usual caffeine consumption >250 and >50 mg/day) or women 35 years old (OR: 7.2 and 7.4 respectively, for usual caffeine consumption >250 and >50 mg/day), suggesting that caffeine intake is more strongly associated with not achieving live birth among women undergoing their first assisted reproduction attempt or older women.
Female caffeine intake and gestational age
Among 41 live-birth infants, 39 had known gestational age with a mean age of 36.9 (SD: 3.4) weeks. The relationship between maternal caffeine consumption and gestational age was of borderline significance in both unadjusted and adjusted analyses (overall P-values: 0.06 for `usual intake' and 0.10 for `week of initial visit'). Infant gestational age decreased by 1.8 weeks (adjusted 95% CI: 4.9, 1.3) for mothers who `usually' (e.g. lifetime) consumed >250 mg/day of caffeine, and 3.8 weeks (adjusted 95% C.I: 6.9, 0.7) for those who consumed >50 mg/day compared with mothers who consumed a minimal amount. Those who consumed >250 or >50 mg/day of caffeine during the `week of the initial clinic visit' decreased the infant's gestational age by 0.9 weeks (adjusted 95% CI: 3.6, 1.8) and 3.5 weeks (adjusted 95% CI: 6.7, 0.3) respectively. This analysis does not include any stillbirths.
Male caffeine intake and multiple gestations
Of 71 pregnant couples, 19 had multiple gestations. In our analysis, complete information on male caffeine intake and the confounders was available for only 57 of the couples (16 with multiple gestations) in questionnaire no. 1, and 45 (13 with multiple gestations) in questionnaire no. 2. Male caffeine consumption, analysed as a linear continuous predictor, was a significant risk factor for multiple gestations in univariate and multivariate models while simultaneously adjusting for the above potential confounders.
The OR were higher in the multivariable models than in the univariable models due to confounding by wife's years of schooling and type of infertility. If a man were to increase his `usual caffeine intake' by an extra 100 mg/day (i.e. equivalent to one more cup of coffee), he would thereby increase his odds of having multiple gestations 2.2 times (adjusted 95% CI: 1.1, 4.4, P = 0.03). Similarly, the OR for an increase in 100 mg/day of caffeine intake during the `week of the initial clinic visit' or during the `week before sperm collection' were 3.0 (adjusted 95% CI: 1.2, 7.4, P = 0.02) and 2.2 (adjusted 95% CI: 0.9, 5.0, P = 0.07) respectively (Table IV). The results for the latter time period may not have been significant because of the smaller sample size resulting from an increased number of missing values.
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
It is difficult to evaluate how these findings conformed to the literature, since there are currently no comparison studies on caffeine and IVF or GIFT. Our findings did not entirely parallel those of other studies; previous investigators studied the reproductive toxicity of caffeine predominantly on fecundability (Srisuphan and Bracken, 1986; Wicox et al., 1988; Williams et al., 1990
; Fenster et al., 1991
; Armstrong et al., 1992
; Hatch and Bracken, 1993
; Infante-Rivard et al., 1993
; Stanton and Gray, 1995
; Bolumar et al., 1997
; Curtis et al., 1997
; Jensen et al., 1998
; Maruoka et al., 1998) or delayed conception (Joesoef et al., 1990
; Williams et al., 1990
; Hatch and Bracken, 1993
; Stanton and Gray, 1995
; Bolumar et al., 1997
), low birthweight (Hinds et al., 1996
; Vlajinac et al., 1997
; Santos et al., 1998b
; Eskenazi et al., 1999
) and spontaneous abortions (Srisuphan and Bracken, 1986
; Wilcox et al., 1990
; Fenster et al., 1991
; Armstrong et al., 1992
; Mills et al., 1993
; Dlugosz et al., 1996
; Fernandes et al., 1998
). In our study, the risks of miscarriage were 19.8 (95% CI: 1.3300.9) and 10.5 (95% CI: 0.9125.3) respectively, for those women who consumed >250 and >50 mg/day of caffeine usually (during their lifetime) compared with 02 mg/day caffeine users, after adjusting for potential confounders. For the same intakes of caffeine during the initial clinic visit, the adjusted OR were 9.5 (adjusted 95% CI: 1.462.0) and 6.2 (adjusted 95% CI: 0.940.8) respectively. Although the relationship between female caffeine intake and miscarriage is of borderline significance, the magnitude of associations are very high, which explains the fewer live births.
There are several biological pathways by which caffeine could affect female reproduction. It could target ovulation through alterations in hormone levels. Caffeine consumption is inversely correlated with levels of estradiol in pregnant women (Hatch and Bracken, 1993) and positively correlated with levels of sex hormone-binding globulin (Hatch and Bracken, 1993
). Caffeine decreases plasma levels of prolactin in non-pregnant, healthy women (Casas et al., 1989
), and may inhibit ovulation or corpus luteum function (Bolumar et al., 1997
). In our study, the relationship during lifetime and the initial clinic visit occurred prior to the women receiving any hormones. It is possible that caffeine alters only endogenous hormone levels and not exogenous hormones. Furthermore, women may have reported drinking coffee during their lifetime but refrained from or decreased coffee drinking during the time of their procedure.
In males, caffeine consumption did not have an effect on sperm parameters (i.e. motility, morphology, sperm count). Our study is in conflict with studies on natural reproduction which reported impaired semen quality (Marshburn et al., 1989) and increased motility (Barkay et al., 1977
; Harrison, 1978
; Aitken et al, 1983
; Hammitt et al., 1989
) with caffeine. Male caffeine consumption also had no effect on fertilization, pregnancy, or live-birth delivery.
Surprisingly, when male caffeine consumption was treated as a continuous variable, caffeine appeared to be a possible risk factor for multiple gestations; an increase in male caffeine intake by an additional 100 mg/day (equivalent to 1 extra cup of coffee) significantly increased the adjusted risk of multiple gestations by 2.2 (95% CI: 1.11,4.4) for usual consumption during their `lifetime', and by 3.0 times (95% CI: 1.2,7.4) during the initial clinic visit. Another study reported OR for multiple pregnancy of 1.5 (95% CI: 0.82.8) and 2.0 (95% CI: 1.03.7) for female drinking 12 cups/day or >3 cups/day respectively, compared with non-coffee drinkers (Parazzini et al., 1996). There are no studies linking male coffee consumption and multiple gestations.
In our study, the questionnaires established a detailed caffeine history from a variety of sources of caffeine (i.e. coffee, tea, soda, cocoa, chocolate); however, it is impossible to determine exactly how much caffeine is present in a cup of coffee or tea. The amount of caffeine depends on the mix of the brew, how it is prepared, and the size of the cup. Furthermore, serum concentrations of caffeine or its primary metabolite, paraxanthine (Klebanoff et al., 1998), were never measured during the study. Another problem was the initial design of the questionnaire. Originally, subjects completed a questionnaire which asked questions only about coffee, decaffeinated coffee, tea, and soda (n = 82). Subsequent questionnaires were updated with a distinction made between caffeinated and decaffeinated tea, soda, chocolate and cocoa (n = 138, one missing). Given that coffee is the overwhelming contributor to total caffeine intake, this change would likely not be that important and would lead to random misclassification and bias towards the null. Finally, factors other than the caffeine in coffee may present the real risk, albeit caffeine intake from other sources did not influence any other endpoints. Caffeinated drugs (e.g. Excedrin, Fiorinal), and dietary supplements were not tabulated in this data set. The second questionnaire was administered and filled out when the woman was pregnant, prior to knowledge of the pregnancy outcome. The third questionnaire was distributed at the pregnancy outcome (miscarriage, live birth). Hence, recall bias could be a potential limitation.
Our study population is predominantly Caucasian and well educated; however, this sample was representative of all women undergoing assisted reproduction in southern California. Moreover, a number of women and men (n = 28) did not complete all the questionnaires. Both females and males who completed all questionnaires did not differ from those subjects who had missing information on most characteristics, apart from race; a higher proportion of non-white participants had missing data. Because these data are not missing at random, this might affect the generalizability of the study results. It is unknown how many women were approached yet declined to be interviewed. A selection bias could have existed if only the men or women with the most healthy lifestyle habits chose to participate in the study (or if women who did not have a successful pregnancy dropped out). This is a relatively small study in terms of the number of women who delivered a healthy baby. Therefore, the lack of statistical significance may be related to low power. In addition, there is a select population of couples seeking treatment and they have a very low caffeine intake (more than half did not consume caffeine).
Finally, there were multiple endpoints and many main effects in this study, hence, some reported associations may have been spurious and attributable to chance, rather than any `real' association. There is no consensus among biostatisticians on how to correct for this problem. If all the null hypotheses being tested were independent, then a Bonferroni or Tukey type adjustment could be made (Miller, 1980; Rice, 1988
); however, the various endpoints are correlated: the number of oocytes retrieved is clearly related to the number of fertilizations and to the pregnancy outcome. This leads to many dependent null hypotheses, and hence, a stringent Bonferroni adjustment would be inappropriate (Miller, 1980
; Rice, 1988
).
The number of couples seeking assisted reproductive technologies is escalating, and the couples' ultimate goals are to give birth to healthy babies. Our study examines a modifiable habit, caffeine, a substance that is widely consumed by both men and women (as well as by 80% of pregnant women), and its effect on all reproductive endpoints of IVF and GIFT. Our findings suggest that couples undergoing IVF or GIFT might benefit from a reduced caffeine intake before the initial clinic visit. Finally, the level of caffeine intake in females is surprisingly low in this study; suggesting that women who have problems conceiving, especially those who go through the trouble and expense of IVF/GIFT, are very likely to decrease their caffeine consumption substantially. If these results are replicated in a larger population, and if doseresponse patterns hold up, this could be a very important finding, since many women consume >150 mg/day (1 cups of coffee).
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Armstrong, B.G., McDonald, A.D. and Sloan, M. (1992) Cigarette, alcohol, and coffee consumption and spontaneous abortion. Am. J. Publ. Health, 82, 8587.[Abstract]
Barkay, J., Zuckerman, H., Sklan, D. and Gordon, S. (1977) Effect of caffeine on increasing the motility of frozen human sperm. Fertil. Steril., 28, 175177.[ISI][Medline]
Bolumar, F., Olsen, J., Rebagliato, M. and Bisanti, L. (1997) Caffeine intake and delayed conception: a European multicenter study on infertility and subfecundity. Am. J. Epidemiol., 145, 324334.[Abstract]
Casas, M., Ferrer, S., Calaf, J., Rodriguez-Espinosa, J. and Herrera-Marschitz. M. (1989) Sopaminergic mechanism for caffeine-induced decreased in fertility? Lancet, 8640, 731.
Curtis, K.M., Savitz, D.A. and Arbuckle, T.E. (1997) Effects of cigarette smoking, caffeine consumption, and alcohol intake on fecundability. Am. J. Epidemiol., 146, 3241.[Abstract]
Dlugosz, L., Belanger, K., Hellenbrand, K., Holford, T.R., Leaderer, B. and Bracken, M.B. (1996) Maternal caffeine consumption and spontaneous abortion: a prospective cohort study. Epidemiology, 3, 250255.
Eskenazi, B. (1993) Caffeine during pregnancy: grounds for concern? J. Am. Med. Assoc., 270, 29732974.[ISI][Medline]
Eskenazi, B., Stapleton, A.L., Kharrazi, M. and Chee, W.Y. (1999) Associations between maternal decaffeinated and caffeinated coffee consumption and fetal growth and gestational duration. Epidemiology, 10, 242249.[ISI][Medline]
Fernandes, O., Sabharwal, M., Smiley, T., Pastuszak, A., Koren, G. and Einarson, T. (1998) Moderate to heavy caffeine consumption during pregnancy and relationship to spontaneous abortion and abnormal fetal growth: a meta-analysis. Reprod. Toxicol., 12, 43544.[ISI][Medline]
Fenster, L., Eskenazi, B., Windham, G. and Swan, S.H. (1991) Caffeine consumption during pregnancy and spontaneous abortion. Epidemiology, 2, 168174.[Medline]
Grodstein, F., Goldman, M.B., Ryan, L. and Cramer, D.W. (1993) Relation of female infertility to consumption of caffeinated beverages. Am. J. Epidemiol., 137, 13501360.
Hammitt, D.G., Bedja, E., Rogers, P.R., Syrop, C.H., Donovan, J.F. and Williamson, R.A. (1989) Comparison of motility stimulants for cryopreserved human semen. Fertil. Steril., 52, 495502.[ISI][Medline]
Harrison, R.F. (1978) Insemination of husband's semen with and without the addition of caffeine. Fertil. Steril., 291, 532534.
Harrison, R.F., Sheppard, B.L. and Kaliszer, M. (1980) Observations on the motility, ultrastructure and elemental composition of human spermatozoa incubated with caffeine. Andrologica, 12, 2429.
Hatch, E.E. and Bracken, M.B. (1993) Association of delayed conception with caffeine consumption. Am. J. Epidemiol., 138, 10821092.[Abstract]
Hinds, T.S., West, W.L., Knight, E.M. and Harland, B.F. (1996) The effect of caffeine on pregnancy outcome variables. Nutr. Rev., 54, 203207.[ISI][Medline]
Imoedemhe, D.A., Sigue, A.B., Pacpaco, E.L. and Olazo, A.B. (1992) The effect of caffeine on the ability of spermatozoa to fertilize mature human oocytes. J. Assist. Reprod. Genet., 9, 155160.[ISI][Medline]
Infante-Rivard, C., Fernandez, A., Gauthier, R., David, M. and Rivard, G.E. (1993) Fetal loss associated with caffeine intake before and during pregnancy. J. Am. Med. Assoc., 270, 29402943.[Abstract]
Jensen, T.K., Henriksen, T.B., Hjollen, N.H., Scheike, T., Kolstad, H., Giwercman, A., Ernst, E., Bonde, J.P., Skakkbaek, N.E. and Olson, J. (1998) Caffeine intake and fecundability: a follow-up study among 430 Danish couples planning their first pregnancy. Reprod. Toxicol., 12, 289295.[ISI][Medline]
Joesoef, M.F., Beral, V., Rolfs, R.T., Aral, S.O. and Cramer, D.W. (1990) Are caffeinated beverages risk factors for delayed conception? Lancet, 335, 136137.[ISI][Medline]
Jones, H.W., Acosta, A.A., Andrews, M.C., Garcia, J.E., Jones, G.S., Mantzavinos, T., McDowell, J., Sandow, B.A., Veeck, L. and Whibley, T.W. (1983) What is a pregnancy? A question for programs of in vitro fertilization. Fertil. Steril., 40, 728.[ISI][Medline]
Klebanoff, M.A., Levine, R.J., Dersimonian, R., Clemens, J.D. and Wilkins, D.G. (1988) Serum caffeine and paraxanthine as markers for reported caffeine intake in pregnancy. Ann. Epidemiol., 8, 107111.
Marshburn, P.B., Sloan, C.S. and Hammond, M.G. (1989) Semen quality and association with coffee drinking, cigarette smoking, and ethanol consumption. Fertil. Steril., 52, 162165.[ISI][Medline]
Maruoka, K., Yagi, M., Akazawa, K., Kinukawa, N., Ueda, K. and Nose, Y. (1988) Risk factors for low birthweight in Japanese infants. Acta Paediat., 87, 304309.
Medical Research International and the Society of Assisted Reproductive Technology (1989) The American Fertility Society: In Vitro Fertilization/Embryo Transfer in the United States: 1987 Results From the National IVFET Registry. Fertil. Steril., 51, 1318.[ISI][Medline]
Miller, R.G. (1980) Simultaneous Statistical Inference. Springer-Verlag, New York.
Mills, J.L., Holmes, L.B., Aarons, J.H., Simpson, J.L., Brown, Z.A., Jovanic-Peterson, L.G., Conley, M.R., Craubard, B.I., Knopp, R.H. and Metzger, B.E. (1993) Moderate caffeine use and the risk of spontaneous abortion and intrauterine growth retardation. J. Am. Med. Assoc., 269, 593597.[Abstract]
Olsen, J. (1991) Cigarette smoking, tea and coffee drinking, and subfecundity. Am. J. Epidemiol., 133, 734739.[Abstract]
Parazzini, F., Chatenoud, L, Benzi, G., Di Cintio, E., Dal Pino, D., Tozzi, L. and Fedele, L. (1996) Coffee and alcohol intake, smoking and risk of multiple pregnancy. Hum. Reprod., 11, 23062309.[Abstract]
Pollard, I. and Smallshaw, J. (1988) Male mediated caffeine effects over two generations of rats. J. Dev. Physiol., 10, 271281.[ISI][Medline]
Rice, J.A. (1988) Mathematical Statistics and Data Analysis. Wadsworth & Brooks/Cole, Monterey, CA.
Santos, I.S., Victora, C.G., Huttly, S. and Morris, S. (1998a) Caffeine intake and pregnancy outcomes: a meta-analytic review. Cadernos de Sude Publica, 14, 523530.
Santos, I.S., Victora, C.G., Hittly, S. and Carvalhal, J.B. (1998b) Caffeine intake and low birth weight: a population-based casecontrol study. Am. J. Epidemiol., 147, 620627.[Abstract]
Srisuphan, W. and Bracken, M.B. (1986) Caffeine consumption during pregnancy and association with late spontaneous abortion. Am. J. Obstet. Gynecol., 154, 1420.[ISI][Medline]
Stanton, C.K. and Gray, R.H. (1995) Effects of caffeine on delayed conception. Am. J. Epidemiol., 142, 13221329.[Abstract]
Stein, A. and Susser, M. (1991) Miscarriage, caffeine and the epiphenomena of pregnancy: the causal model. Epidemiology, 2, 163167.[Medline]
Stephen, E.H. and Chandra, A. (2000) Use of infertility services in the United States, 1995. Fam. Plann. Perspect., 32, 132137.[ISI][Medline]
Vlajinac, H.D., Petrovic, R.R., Marinkovic, J.M., Sipetic, S.B. and Adanja, B.J. (1997) Effect of caffeine intake during pregnancy on birth weight. Am. J. Epidemiol., 145, 335338.[Abstract]
Wilcox, A., Weinberg, C. and Baird, D. (1988) Caffeinated beverages and decreased fertility. Lancet, 2, 14531456.[Medline]
Wilcox, A.J., Weinberg, C.R. and Baird, D.D. (1990) Risk factors for early pregnancy loss. Epidemiology, 1, 382385.[Medline]
Williams, M.A., Monson, R.R., Goldman, M.B., Mittendorf, R. and Ryan, K.J. (1990) Coffee and delayed conception. Lancet, 335, 1603.
Williams, J. (1985) Obstetrics, 17th edn. Appleton, New York.
World Health Organization (1992) Report from Committee on Sperm Fertility Parameters.
Woodward, M. (1999) Epidemiology, Study Design and Data Analysis. Chapman & Hall, Boca Raton, FL.
Submitted on July 11, 2001; resubmitted on November 22, 2001; accepted on March 8, 2002.