Measuring and Reporting Nutrition Status key to Improving Public Health
Science is supposed to be a disciplined way to study the natural world. Insights from robust correlational observations are used to design experiments to test hypotheses. Ultimately, results derived through scientific methods are evaluated by peers and published in a scientific journal. Unfortunately, sometimes the quality of the research and the rigor of the review process is inconsistent. Three cases in point.
Nohr and colleagues examined reported use of dietary supplements, folate-only or multivitamin, by 35,914 women living in Denmark during a 12 week interval when they were pregnant. Biological samples were not take to measure vitamin status. The authors didn’t constrain the data collection to any part of the pregnancy although results were analyzed statistically by term of pregnancy (<20 wk or ≥20 wk). They report that folate-only supplement use was not associated with fetal death and that multivitamin supplement use was associated with modest increased risk of early fetal death. The conclusion with respect to multivitamins is not responsible. Folic acid is known to be essential for normal brain stem development of the fetus in the first trimester of pregnancy. Babies being carried by women with red blood cell (RBC) folate levels <340 nmol/L are at increased risk of neural tube defects (NTD) and miscarriage in the first trimester. Multivitamin supplements usually have less folic acid per serving than folate-only supplements. The modest increased risk among women using multivitamins likely reflects a greater likelihood of inadequate maternal folic acid intake and status during the first trimester pregnancy. The take-home message shouldn’t be an indictment against multivitamins; it should be a call for better information on the maternal folate status of women using multivitamins.
Wang and colleagues conducted a prospective cohort study of 62 patients hospitalized with Clostridium difficile associated diarrhea. They were followed for 30 days and logistic regression analyses of variables (Charleson Comorbidity Index, 25(OH)D levels, gender, white blood count, albumin level and residence type) were conducted. Medscape reports that low vitamin D levels and age (≥ 70y) are significantly associated with increased likelihood of recurrence. The odds ratio (OR) for low levels of vitamin D was ~5 fold higher (vs high 25(OH)D levels). Unfortunately vitamin D levels are not reported so it is impossible to know if participants were vitamin D deficient, suboptimal, or optimal. The take-home message is that studies reporting odds ratios without numerical definitions of vitamin status are not very useful.
The final example is a study by Pottala and colleagues. They retrospectively analyzed the relationship between RBC omega-3 fatty acid (EPA + DHA) concentrations and MRI brain volumes measured 8 years later in 1,111 postmenopausal women from the Women’s Health Initiative Memory Study. Higher RBC EPA + DHA concentrations are associated with larger brain volume. Women in the highest quartile have a RBC omega-3 index of ~ 7.5% vs 3.4% measured in the lowest quartile. This is an excellent study where valid measures of nutritional status, i.e. RBC EPA + DHA concentrations, are reported. In interviews, Dr Pottala acknowledges the limitations of the study, i.e. that it is not known if omega-3 consumption had changed over time or if the observed differences in brain volume might impact cognitive function.
Case number 3, with its rigorous focus on relating biological measures of nutritional status with functional outcomes, is an excellent model of the scientific method. More research reporting biological indicators of status, and less research restricted to measures of supplement use and/or food intake only, is needed.
Nohr EA, Olsen J, Bech BH, Bodnar LM, Olsen SF, Catov JM. Periconceptual intake of vitamins and fetal death: a cohort study on multivitamins and folate. 2014 Int J Epidemiol doi: 10.1093/ije/dyt214
Wang WJ, Gray S, Sison C, Arramraju S, John BK, Hussain SA, Kim Sang, Mehta P, Rubin M. Low vitamin D level is an independent predictor of poor outcomes in Clostridium difficile – associated diarrhea. 2014 Ther Adv Gastroenterol 10.1177/1756283X13508519
Pottala JV, Yaffe K, Robinson JG, Espeland MA, Wallace R, Harris WS. Higher RBC EPA + DHA corresponds with larger total brain and hippocampal volumes: WHIMS-MRI Study. 2014 Neurol doi:10.1212/WNL. 0000000000000080