Good news and bad news for children: FDA’s 2014 to 2016 food testing for lead

Tom Neltner, Chemicals Policy Director and Maricel Maffini, Consultant

In Environmental Defense Fund’s June 2017 “Lead in food: A hidden health threat” report, we evaluated the Food and Drug Administration’s (FDA) publicly available data for 2003 to 2013 from its Total Diet Study (TDS). Since the 1970s, the TDS has tracked metals, pesticides, and nutrients in food. We found that 20% of baby food samples had detectable levels of lead compared to 14% for other foods. We also identified eight food types where more than 40% of samples had detectable lead. Finally, based on an analysis from EPA, we estimated that more than 1 million young children exceeded FDA’s limit for lead and that eliminating lead from food would save society an estimated $27 billion annually.

In November 2017, FDA publicly released TDS data for 2014 and 2015. And this December, the agency provided us with TDS data for 2016 in response to our Freedom of Information Act (FOIA) request. We evaluated the combined information[1] and have an update that is both good news and bad news. The good news is that the overall trends for detectable rates of lead in baby food and other food, especially in 2016, appear to be on the decline. There is a similar downward trend in apple and grape juice, especially the ones marketed as baby food. Unfortunately, there appeared to be stubbornly high rates of detectable lead in teething biscuits, arrowroot cookies, and baby food carrots and sweet potatoes.

FDA switched to new more sensitive analytical method in 2014

One of the challenges in making comparisons between the 2003 to 2013 data and the newer data is the fact that FDA switched to a new more sensitive analytical method in 2014 called Inductively Coupled Plasma Mass Spectrometry (ICPMS). The new method’s Limit of Detection (LOD) is significantly lower than the old method and varies with the type of food. For example, the LOD for bottled water dropped from 4 to 0.052 parts per billion (ppb). It was less dramatic for all other food with the new method LODs ranging from 1 to 7 ppb compared to 4 to 20 ppb under the old method. When considering the new method and lower LODs, it is important to remember that no safe level of lead in blood has been identified.

With the new, lower LOD, the percentage of foods with detectable lead levels increased in 2014-16. This increase makes sense because the new, more sensitive method detects lead at concentrations that would have been undetected with the old method. We are pleased that FDA updated its analytical process.

Good news in the overall trends of lead, especially in 2016

To make a fair comparison to the 2003-13 data presented in our June report, we evaluated the 2014-16 data using the old and new LODs (Figure 1). The average percent of samples with detectable levels using the old method’s LODs was 12.7% for baby foods and 7.7% for other (non-baby) foods – lower than in all of the prior years. Using the new method, the averages were 29.3% for baby food and 26.9% for other food, with a significant drop in the averages in 2016.

Good news for apple and grape juice, especially the baby food juices

As shown in Figure 2, we used the old LOD to enable comparisons across all the years. The news is also generally good for baby food (BF) fruit juices. From 2014-16, detectable lead levels dropped in all baby food fruit juices collected by FDA in the TDS. Rates for non-baby food apple juice stayed low, and grape juice reversed an upward trend. It may be relevant to note that in January 2012, Consumers Reports highlighted high levels of lead in grape and apple juice. We noted in our report that “Detectable rates of lead in apple and grape juice, especially baby food, appeared to be somewhat lower in 2013 compared to the earlier years, but there were too few composite samples to draw conclusions.”

Although the downward trends are encouraging, there is still progress to be made. There was a slight increase for non-baby food fruit juices other than apple and grape; these juices include orange, grapefruit, prune, pineapple, cranberry and tomato. Using the new LOD, 63% of grape, 36% of pear and 9% of apple baby food juices still have detectable lead levels.

Stubbornly high rates of detectable lead for some baby food products

The news for baby food desserts, infant formula and non-root vegetables was positive. From 2014-16, only 1 of 79 composite samples (1.2%) had detectable levels of lead using the old LOD. Using the new LOD, 15% of baby food desserts, 5% of infant formula, and 21% of baby food non-root vegetables had detectable levels.

The bad news for children, as shown in Figure 3, was for baby food carrots, sweet potatoes, teething biscuits and arrowroot cookies. Increases in lead detection in 2014-16 using the old LOD reversed a promising drop in 2011-13. Using the new LOD, 41 of 44 composite samples (93%) of baby food teething biscuits, sweet potatoes, arrowroot cookies, and carrots had detectable lead levels.

More progress needed

Further progress on reducing levels of lead in food is critical – particularly because no safe level of lead in blood has been identified. The good news in this update demonstrates that continued progress is being made in some categories but attention needs to be paid to a number of product categories.

In our June 2017 report, we provided a series of recommendations that food manufacturers and FDA can take to make progress on this issue – starting with FDA updating its decades-old standards for lead in food. We hope our report provides an incentive for both food manufacturers and FDA to do more on reducing this meaningful source of children’s exposure to lead.


[1] For a copy of the spreadsheet and the details of the analysis, please contact me at

This entry was posted in FDA, Food, Health policy, Health science, Lead, Public health, Regulation and tagged , , , , . Bookmark the permalink. Both comments and trackbacks are currently closed.