Anne Marte W. Johansen, Allen J Wilcox, Rolv T. Lie, Lene F. Andersen, and Christian A. Drevon
Objectives
In this study, Anne Johansen and colleagues used a population-based case-control study to evaluate the association of maternal consumption of coffee and caffeinated beverages in early pregnancy, the with risk of delivering an infant with an orofacial cleft. The study was based in Norway, where orofacial clefts show a high prominence. Coffee and caffeine-containing beverages were selected variables for the analysis, as coffee tends to be a common beverage for consumption by pregnant women, and thus even a small association between it and malformation risk could be relevant for prevention practices.
Procedures
The study was completed from 1996-2001 on mothers of babies born with facial clefts, as well as on control mothers. 573 mothers participated who had delivered infants with orofacial clefts, and 763 control mothers were included in the study. Those participants which delivered infants with orofacial clefts were divided into two groups, those who suffered with Cleft lip (with or without Cleft palate), CLP, and those who had suffered with Cleft palate only, COP. Medical records were used, in conjunction with the questionnaires, to identify any additional defects which occurred in the cleft cases. Those with no additional defects were termed isoloated cases.
The mothers completed a 32 page questionnaire, at a median time of 14 weeks after delivery (cleft cases) and 15 weeks after delivery (control cases). These questionnaires covered demographic characters, reproductive history, and exposures during pregnancy (smoking, alcohol consumption, coffee intake, medication use, and work/household exposures). Items containing maternal caffeine consumtion were based on coffee, tea, and soft-drinks consumed during the first three months of pregnancy.
The risk of delivering infants with an orofacial cleft was approximated using odds-ratios. Coffer beverages were analysed based on a "cups per day" CPD measurement, and then the women were grouped based on a 3-category variable of 0 CPD, 1-2 CPD, and 3 or more CPD. Caffeine content of coffee, tea, and soft drinks was estimated based on Norway nutrition guides, such that the risk of cleft could be evaluated per 100 mg increase in caffeine intake, as well as in categories of women who used >100-<500> or = to 500 mg. Johansen and colleagues also made adjustments in data, to minimize the impact of confounding or unrelated factors, which possibly could cause a cleft risk. CLP and COP were separately analysed. In addition, isolated clefts (those without accompanying defects) were also considered as a separate type of cleft when considering caffeine risk. Statistical analysis was conducted using SPSS 14.0 software.
Results
The results obtained showed an association between maternal coffee consumption in the firs trimester and an increased risk of CLP. No evidence for an association between maternal coffee consumption and COP was observed. The intake of coffee during the first three months of pregnancy was associated in a dose-response manner with risk of delivering an infant with CLP. In addition, the results also showed caffeine containing tea to be associated with with a decreased risk of CLP and COP. Soft drinks containing caffeine showed an increase in risk for both types of facial clefts, however the relationship recorded did not prove to be significant.
From the results, Johansen et al. deduced that caffeine is not a risk factor associated with the delivery of a babies with oral cleft. They did state however, that coffee may potentially impact the infants during the first trimester of their development, adding an increased risk to CLP development. They suggested a pathway by which this defect occurs through coffee influence, based on the fact that coffee, like smoking (a known CLP risk factor), increases plasma homocysteine, which has been suggested as a potential cause of CLP.
Authors critiques
Throughout the discussion, Johansen et al. discussed a number of critques for their experimental procedure, and even ruled out some of the conclusions which could be drawn from the obtained results. They mentioned that the declined risk of cleft with tea consumption was most likely due to unmeasured confounding factors, as tea drinkers showed healthier overall life habits on the questionnaire. In addition, they commented on the fact that it is unlikely for the risk of both cleft types to be reduced by the same specific factor, and thus determined that it was most likely not the tea which was reducing the risk. The discussion also stated that adjustments do not always rule out the possibility that unrelated factors may impact the results. For instance, smoking may infact be creating a bigger increase in risk than was being accounted for. Even the correlation between CLP and coffee was questioned by the authors, as confounding factors associated with the habit of coffee drinking may not have been controlled and thus may have been the direct link creating the relationship between the two variables.
Personal Critique
Many aspects of this paper presented issues in terms of result interpretation. Firstly, 14 and 15 weeks after infant delivery is a substantial amount of time from the first trimester of pregnancy, and thus it may have caused the survey results to be inaccurate. Secondly, although the experiment does not specify cup size, this should not be left to be implied or assumed, as one person's judgement of a cup could be quite different from anothers, therefore possibly resulting in incorrect measures of caffeine intake. Also, non-Norwegian speaking women were omitted from the women selected for the study. Even though this study is directed toward the Norway population (with its high cleft incidence and high coffee intake), those giving birth in Norway and speaking another language should also be included in the study, as they too could provide insight on the impacts of caffeine/coffee. Throughout the discussion, Johansen et al. made reference to 'data not shown'. These missing/not shown results make understanding the author's conclusions difficult to follow, and are needed for a clear understanding of the studies outcomes. Lastly, the presentation of the survey's findings into tables full of numbers was perhaps not the best choice. All of the abbreviations, and statistical data was bunched together into large tables, making the findings difficult to follow. A potential improvement would be to provide figures, showing general trends, in additon to the tables.
