2023 FDA Science Forum
Comparison of Sample Size and Methods of Homogenization for the Enumeration of Bacillus cereus in Artifically Preserved Cosmetic Wipes
- Authors:
- Center:
-
Contributing OfficeCenter for Food Safety and Applied Nutrition
Abstract
Microbial contamination in cosmetic wipes can cause health issues and economical loss due to product recalls. The microbiological analysis of cosmetic wipes using the procedures in FDA’s BAM Chapter 23 for solid products consists of diluting a 1 g sample size 1:10 with 1 ml Tween® 80 and 8 ml Modified Letheen Broth (MLB), using a vortex mixer to recover microbes for enumeration. This sample size may not be optimal for the analysis of wipes. Therefore, in this study, we tested an entire wipe using a vortex mixer (EV), or a stomacher lab blender (ES) for recovery of B. cereus and compared the recovered count with those obtained using 1g sample size with a vortex mixer (1gV) or 1g sample with a stomacher lab blender (1gS). Two types of dry wipes were humidified with 0.45% Sodium Benzoate (SB) or 0.002 % Benzalkonium chloride (BAK) solutions inoculated with Bacillus cereus 3A spore suspensions at three different levels of contamination, low (~4.7 log CFU/ml), medium (~5.7 log CFU/ml), and high (~6.7 log CFU/ml), and at 5 replicates per level. After 14 days at room temperature, the wipes were either cut to a 1 g sample size in the center or left untouched. The samples were diluted 1:10 with the same proportion of Tween® 80 and MLB as above, similar to the BAM protocol. All the samples were enumerated (CFU/g) on nonselective MLA (Modified Letheen Agar) plates, and on selective BACARA™ agar plates, after homogenization with a vortex mixer for 30 seconds, or with a stomacher for a minute. Uninoculated wipes were used as negative controls. No significant difference in B. cereus counts was observed between the vortex mixer samples and the stomacher samples, regardless of sample size (p=0.668). However, differences were observed between agar plates (p<0.001), the type of wipes (p<0.001), and the sample size (p<0.001). Higher microbial counts were seen on non-selective MLA compared to counts on BACARA™ agar. Microbial recovery was also higher from 1 g sample than from the entire wipe, which might be due to possible death of cells in dried edge areas of a sheet, and will be further investigated.