Study on the analytical capability and readiness of selected laboratories to analyze fortified food – a quick look into key learnings and recommendations
In the past few years I attended numerous food fortification meetings in dozens of countries. The clear health benefits and cost-effectiveness of this intervention to improve public health are never disputed. However the implementation of food fortification is frequently delayed and complicated by fairly difficult questions about how to monitor and control the process as well as its effectiveness.
A number of manuals have been issued on how to monitor food fortification* (e.g., WHO, ECSA). A few also cover how the fortified samples should be tested listing different analytical methods. Still, there is no clear data on how reliable these specific methods are when testing fortified foods, how accurate the results coming out of various national laboratories are and how to interpret these results to make the right decision.
Obtaining such data is no easy endeavor. Measuring levels of vitamins and minerals in food is a complex task of understanding if the levels and variation in levels are acceptable. It is a different approach than food safety testing, where results are usually either YES (present) or NO (none detected). The proficiency testing schemes, which are designed to generate this data, are quite expensive and not easily executed in developing countries.
Luckily Dr. Philip Randall, a senior technical expert in food quality and safety, did take on this endeavor in 2016-2017. He designed, executed and evaluated a first-of-its-kind study to assess the capability of national laboratories to test fortified foods and the accuracy of their results compared to select accredited reference laboratories.
The capacity of laboratories to measure fortified foods was assessed by how many signed in to participate in the study compared to how many managed to deliver the results in time. The accuracy of the results was assessed by calculating measurement uncertainty. The reference values for accuracy were generated by reference laboratories with extensive international accreditation status.
The samples used in this study included key staple foods such as wheat and maize flours, sugar**, edible oil and table salt. The micronutrients measured were vitamin A in oil, iodine in salt, vitamin A in sugar and iron with vitamin A in wheat and maize flours.
This study became known as a “ring-trial” amongst the participants, whilst the official name is “Study on the analytical capability and readiness of selected laboratories to analyze fortified food.”
The samples mentioned above were not available as Certified Reference Material and had to be prepared specifically for this study. The food vehicles (flour, salt, sugar, oil) were combined with micronutrient premixes containing specific amounts of vitamin A, iron or iodine, homogenized, aliquoted and sent out to all laboratories that signed up for the study. Laboratories were instructed to apply all available and functional testing methods.
Technology used by the reference laboratories were high performance liquid chromatography (HPLC) for vitamin A, Inductively Coupled Mass Spectroscopy (ICP-MS) for iron and iodine; iodine was also measured photometrically according to Sandell-Kolthoff. The other laboratories reported results with the following technologies: HPLC, spectrophotometry acc. to East Central Southern Africa (ECSA) manuals and iCheck rapid test kits. The accuracy was evaluated applying extended measurement uncertainty at 95% confidence level. It was presented using Z-scores that were categorized in the following manner:
- <2 – satisfactory accuracy
- >±2 ≤ ±3 – questionable accuracy and indicative of attention being necessary to equipment and/or procedures
- >3 – unsatisfactory accuracy and require urgent investigation
- Accuracy clearly depended on the sample type and concentration levels of the target analyte
- Preparation of reference samples with known concentration of vitamins and minerals was a challenge with no ISO 17043 accredited lab offering reasonably priced services
- 21% of labs willing to participate were not able to deliver any results due to various capacity issues
- 45% of all data generated had Z-scores below 2
- 12% of labs that have iCheck test kits did not deliver analytical data
- iCheck provided 58% of all delivered analytical data, and 53% of this data had Z-scores below 2
- 43% of labs that have conventional methods (i.e. ECSA, AOAC) did not deliver analytical data
- 30% of data generated by conventional methods had Z-scores below 2
- Target accuracy for the results generated when testing fortified foods must be set based on the food vehicle, target analyte and concentration range
- Reference samples with known concentration are recommended to be prepared for future cases with certified reference materials (not premix) and by ISO 17043 accredited lab
- iCheck test kit performance is good and the method delivered overall more accurate results when compared to other methods such as spectrophotometry (ref. ECSA manual)
- Over 55% of labs do require training on all methods and regular proficiency tests to maintain the required level of skill and accuracy
The results of each laboratory were communicated confidentially back to each lab, while the aggregated results have been put into a report and are in the process of being published.
Action has already been triggered by the key findings of this study, with GAIN and ECSA making plans regarding analytical methodology and proficiency schemes. Significant planning as well as budget allocation would also be required by the individual labs to improve their capacity.
This study provides important input to facilitate urgently needed realistic monitoring guidelines for fortified foods and improve overall clarity and compliance within food fortification programs.
If you want to more information about this study you are welcome to contact the lead investigator Dr. Philip Randall at firstname.lastname@example.org.
This study is supported by Global Alliance of Improved Nutrition (GAIN) and funded by The Government of The Netherlands. Acknowledgements: David Morgan (GAIN), Gerhard Rimkus (Intertek Food Services) and Phillip Makhumula (fortification consultant) all contributed to the experimental design and revision of the study report.
*Manuals and instructions are available online on the websites of:
- Global Alliance for Improved Nutrition – https://www.gainhealth.org/knowledge-centre/search/
- Food Fortification Initiative – http://www.ffinetwork.org/monitor/index.html
- Smarter Future – http://www.smarterfutures.net/
- World Health Organization – http://www.who.int/nutrition/publications/guidelines/en/
- Food and Agriculture Organization – http://www.fao.org/docrep/W2840E/w2840e0b.htm
**Vitamin A in sugar data is being evaluated