The Nutrition Labeling and Education Act (NLEA), which came to pass in 1994, required most processed foods to be properly labeled displaying the very familiar Nutrition Facts panel we see on foods today. In this study, Jayachandran N. Variyam and John Cawley set out to examine the impact of the NLEA on obesity in America. Survey data is gathered on those who mind labels when shopping and those who ignore them, and weights are examined for the years preceding and following the institution of the NLEA. From the data it was observed that among non-Hispanic white women, there was a decline in average bodyweight and the probability of obesity the after the NLEA took effect.
The potential effect of the new label law on obesity was explored by observing body weight changes of individuals before and after the law took effect and comparing these changes for those who say they use labels and those who don’t. The sample data to implement this study came from the National Health Interview Survey (NHIS), which is a multipurpose survey of US citizens conducted yearly since 1958. Anywhere from 36,000 to 47,000 households are interviewed yearly, which includes 92,000 to 125,000 individuals.
The survey included a question requiring individuals had to rate their level of nutrition label usage: always, often, sometimes, rarely, or never. Those who chose always, often, or sometimes were placed into the label user group and those who chose rarely and never were placed into the label nonuser group. The data concerning obesity before the NLEA was enforced is gathered from the 1991 to 1993 survey data, and the “after” data is gathered from the 1995 to 1998 surveys. Because the sample is so large the data can also be broken down into subgroups so that any effect can be compared across different demographics, as weight outcomes have a definite genetic component influencing the data separate from changes in label requirements. As a measurement for obesity the Body Mass Index (BMI) was used (a ratio of weight in kilos to the square of height in meters), calculated using individuals self-reported height and body weight from the survey. Those considered obese are those with a BMI of greater than or equal to 30.
Results concerning weight change show that in most cases the weight of the label users increased from before to after the enforcement of the NLEA. Many other factors could be influencing this weight gain, which is why data was collected on those who don’t use labels. Those who pay no attention to labels are subject to the same outside factors influencing weight gain, so to measure the impact of solely the NLEA, the relative weight change in label users to label nonusers is compared. This allows for the amount of weight gained independent of any nutrition label effects to be observed and thus removed, leaving solely the change in weight caused by the NLEA. When these non-label effects are removed, the body weight of the overall population actually fell after the NLEA began, both as a measure of BMI and by percent considered obese. More specifically, the largest label effect was found for non-Hispanic white women, with a net reduction of .52kg/m^2 in their BMI and a 3.36% reduction in the number of obese. Obesity in black women declined as well by 5%, but the decline in BMI was not statistically significant. For white men no effect was found, but for black men, post NLEA enforcement their body weight actually increased. However, the data is reevaluated under two regression models in order to control for differences in sociodemographic characteristics between label users and nonusers. Based on the full regression analysis the effect on obesity prevalence and BMI is insignificantly different from zero for all other genders/races besides that of non-Hispanic white women, who based on the full covariate specification saw on average a .3 kg/m^2 decrease in their BMI and a 2.36% decrease in obesity prevalence as a result of the new nutrition labels.
The study reveals the NLEA did not lead to an increase in the number of people who used labels. The survey placed approximately 69% of individuals in the label-user category consistently from 1991 to 1998, but label usage is of little concern as the main goal was to observe any effect among those who do use labels.
Measuring the effect of the NLEA is difficult due to the many other factors influencing individuals weight gain or weight loss. It is difficult under any circumstance in a study such as this to measure one of those factors impacts accurately. For example the group who doesn’t use labels may have certain characteristics leading to weight that those who use labels don’t. Just as those who use labels may use them simply because they are more likely to be obese, those who don’t use them may show less concern for what they eat and disregard any nutritional advice or information. If this is true then the true effect of the NLEA on weight gain is masked by unaccounted-for outside determinants. In order to properly analyze any effects of the NLEA research needs to examine that changes in both groups weight gain over the following year (1995) to account for the possible lagged weight gain in non-label users.
Also there is no way to be certain that the impact on non-Hispanic white women wasn’t caused by possible other benefits of the NLEA, such as healthier food options offered by companies now that they must account for their nutrition facts. For example those who claimed they use labels may be more careful in choosing foods, and as a result of the NLEA, more healthy snacks or foods were offered.
A study on the different race/gender awareness levels of the NLEA is another area left to further research. Maybe the white non-Hispanic female demographic was more exposed to relevant information on nutrition awareness, than other demographics.