SPIFAN Nutrients ERP Book_9-29-15

GOS-02 SLV FOR ERP USE ONLY DO NOT DISTRIBUTE

Single Lab Validation Report for GOS in Infant Formula and Adult Nutritionals

Table 3: Method accuracy determined by spike-recovery experiments

Spike level (g / 100 g)

Recovery (%)

RSD (%)

Meets SMPR Requirements?

Matrix

GOS

Vivinal Clasado Vivinal Clasado Clasado Vivinal Clasado Vivinal Vivinal Clasado Vivinal Clasado Clasado Vivinal Clasado Vivinal

0.02

91 84 86 85 98 90 87 88 90 85 85 84 71 78 86 84

11.5

Yes No No No Yes Yes No No Yes No No No No No No No

0.1 0.4 0.6

9.0 3.4 5.0

7 (11750017V3)

0.02

10.3

0.1 0.4 0.6

5.2 4.3 6.6

10 (00394RF00)

0.02

18.5

0.1 0.4 0.6

3.3 8.5 2.8

11 (00412RF00)

0.02

10.3 10.7

0.1 0.4 0.6

12 (00403RF00)

4.1 9.3

Recoveries are generally in the range of 80 – 90%, which is lower than SMPR requirements (90 – 110%). It is likely that part of the GOS is lost in the preparative step, in which lactose is removed, or in the SPE purification. This, however, needs further investigation.

As stated before, because of confusion regarding sample preparation, the spiking levels were a factor 9 lower than anticipated. This could explain the higher RSD values for samples spiked at 0.02 g/ 100 g.

Because of the low spiking levels there was more background, which makes it more difficult to determine the cut off point for fractionation. With higher spiking levels (and less background) the window of the fraction, which is discarded in the preparative separation, can be narrowed. This may improve recoveries. If the LOQ is calculated based on blank measurements (following SMPR guidelines), the value is about 0.001 g / 100 g product. However, this does not seem “fair”, because this method uses a complex calculation with an internal standard and subtracted values of the hydrolyzed and non-hydrolyzed samples. The LOQ is estimated at 0.1 g / 100 g, which still meets the requirements of the SMPR (≤ 0.5 g /100 g).

11