LAUSR

We appreciate Dr. Camacho’s interest in our article (Desrosiers, Siega-Riz, Mosley, Meyer, & National Birth Defects Prevention Study, 2018), and welcome the opportunity to respond. As Dr. Camacho emphasizes, the association between folic acid and neural tube defects (NTDs) is well established (Viswanathan et al., 2017). What is less certain are factors potentially related to folate insufficiency among women of reproductive age, and even more so, among women who meet the U.S. Preventative Health Task Force’s recommendation of 0.4–0.8 mg/day of supplemental folic acid (Tinker, Hamner, Qi, & Crider, 2015; US Preventive Services Task Force, 2017). Understanding the reasons for folate insufficiency could help prevent some cases of birth defects in the future and is thus a worthy research pursuit. One theory expressed in the literature is whether avoidance of carbohydrate-rich foods such as enriched grains (fortified with folic acid) and beans (high in natural folate) could lead to meaningful reductions in folate status, which could in turn lead to an increased risk for some women of having an NTDaffected pregnancy (Mills, 2017; Quinlivan & Gregory, 2007). Dr. Camacho asserts that the title Low carbohydrate diets may increase risk of neural tube defects implies that carbohydrate intake “must be encouraged in order to avoid NTDs, or that even a high [carbohydrate] intake could be protective.” In general, we caution against ever drawing conclusions about the substance and implications of any article without reading beyond its title. In the case of our article, a reader would discover that the observed association between carbohydrate intake and NTDs in our data was not monotonic. There is in fact no evidence in our study that a high carbohydrate intake could be protective. Rather, the primary finding of our casecontrol study was that women with low carbohydrate intake (defined as 10th percentile of intake among controls in our study population, approximately 95 g/day) were slightly more likely to have a pregnancy affected with a NTD (AOR5 1.30; 95% CI: 1.02, 1.67). We would like to believe that we and Dr. Camacho share the goal of improving maternal and child health. With that goal in mind, rather than debating the merit of making generalized inferences based on an article title alone, a more productive effort could be to find common ground about the implications of this study. As we state in the article, further research is needed to understand whether the observed association between carbohydrate intake and NTDs is mediated by low folate status versus other pathways. Understanding the reasons why women may have low carbohydrate intake around the time of pregnancy and how those preferences influence NTD risk is another avenue for future research. Finally, we believe our findings serve to underscore the importance for health care providers and women who are or may become pregnant to have routine conversations about any special diets or food restrictions a woman may be practicing. This conversation would ideally consider the risks and benefits of a particular dietary pattern (including low carbohydrate) in the context of pregnancy and pre-existing medical conditions for an individual patient, and help ensure that critical nutritional needs (including folic acid) are met for both mother and baby. Tania Desrosiers, Anna Maria Siega-Riz, Bridget Mosley, and Robert Meyer