Determination of Lactose in Lactose-Free and Low-Lactose Milk, Milk Products, and Products Containing Dairy Ingredients by the LactoSens^®R Amperometry Method: Final Action 2020.01

Abstract

Background

The LactoSens^®R method was previously shown to have acceptable accuracy and repeatability precision as required by AOAC Standard Method Performance Requirements (SMPR^®) 2018.009 for determination of lactose in low-lactose or lactose-free milk, milk products, and products containing dairy ingredients and was awarded Official Method of Analysis^SM (OMA) First Action status in 2020.

Objective

The method was subjected to a multilaboratory validation (MLV) study to evaluate the reproducibility precision of the method.

Methods

Fourteen validation materials were provided to 15 laboratories in seven countries as blind duplicates. The materials ranged from 0 to 173 mg/100 g lactose. Each laboratory analyzed the blind duplicates according to OMA 2020.01. The data were analyzed for repeatability and reproducibility precision.

Results

RSD_r values varied from 2.81 to 8.76%, and RSD_R values varied from 4.25 to 12.5%. When sorted by category and concentration range, these results met the repeatability and reproducibility criteria required by SMPR 2018.009.

Conclusions

The data generated in the MLV support the adoption of OMA 2020.01 as Final Action status.

Highlights

The LactoSensR method, as described by OMA 2020.01, provides an accurate and precise determination of lactose in a variety of low-lactose and lactose-free milk, milk products, and products containing dairy ingredients in minutes.

Lactose intolerance can cause bloating, diarrhea, gas, nausea, and abdominal pain following ingestion of foods containing lactose, such as dairy products or products containing dairy ingredients (1). The condition is found in individuals whose small intestine produces insufficient lactase enzyme and, therefore, the lactose consumed is not fully digested. Many of the symptoms can be attributed to fermentation of the excess lactose by the gut microbiome. Symptoms can be managed by the limitation or total avoidance of dairy products in the diet, which could in turn reduce calcium and vitamin D intake, causing other health effects. Dairy producers have developed low-lactose and lactose-free dairy products to accommodate the dietary needs of lactose-intolerant individuals. These dairy products are treated with lactase enzyme during production to reduce or eliminate lactose, thus allowing lactose intolerant individuals to manage their symptoms without eliminating dairy products completely from their diets.

To support and verify label claims for lactose-free and low-lactose milk and dairy products, new methods are needed that are accurate, precise, and rapid. The LactoSens^®R method is a biosensor assay kit intended for the determination of lactose levels in a variety of lactose-free or low-lactose milk and dairy products. Enzyme immobilized on a disposable test strip oxidizes the lactose in the test portion and the resulting electrons are detected amperometrically by the LactoSensR Reader. Proprietary software converts the electrical signal into a lactose concentration according to the factory-set calibration function. Each test strip is labeled with a QR code for sample tracking and batch-specific information and a ready-to-use positive control is provided for quality assurance.

The LactoSensR method was certified by NordVal in 2018 (2). After successful demonstration of method performance in a single laboratory study compared to the performance requirements of AOAC SMPR 2018.009, Standard Method Performance Requirements for Lactose in Low-Lactose or Lactose-Free Milk, Milk Products, and Products Containing Dairy Ingredients (3), the method was adopted as AOAC OMA 2020.01 First Action in 2020 (4). This report provides the results of a collaborative study undertaken to evaluate the reproducibility of the method.

Multilaboratory Validation Study

Study Design

The multilaboratory validation (MLV) study included 15 laboratories from 7 countries. The laboratories listed in Table 1 included representatives of academia, public health authority, analytical service laboratory, enzyme producers, and dairies. Each laboratory received blind duplicate test portions of 14 validation materials.

Table 1.

Participating laboratories for MLV study

Country	City, Address	Institution	Collaborators
Austria	Vienna, Spargelfeldstraße 191	Austrian Agency for Health and Food Safety	Rupert Hochegger and Manuela Prochaska
Austria	Baden, Vöslauerstrasse 109	NÖM AG	Harald Dengg and Mrs Bogenreiter
Austria	Knittelfeld, Hautzenbichlstrasse 1	Obersteirische Molkerei eGen	Daniel Bara and Maria Rosenberger
Austria	Vienna, Muthgasse 18	University of Natural Resources and Life Sciences—Institute for Food Technology	Roland Ludwig and Cindy Lorenz
Austria	Vienna, Muthgasse 18	University of Natural Resources and Life Sciences—Core Facility Food & Bio Processing	Henry Jäger and Ragnar Witlaczil
Austria	Stainach-Pürgg, Bahnhofstraße 182	Ennstal Milch KG	Ulrich Griesbacher
Austria	Vienna, Muthgasse 11	DirectSens GmbH	Jasmin Korp and Zahera Zahir
Denmark	Hørsholm, Bøge Alle 10-12	Chr Hansen A/S	Hans Raj
Denmark	Esbjerg, Kvaglundvej 84	Arla Foods	Lene Loth Csikos
Germany	Oy-Mittelberg, Oberzollhauser Steige 4	Optiferm GmbH	Alexander Siegl
Germany	Piding, Hockerfeld 5-8	Milchwerke Berchtesgadener Land Chiemgau eG	Sara Wurzinger and Katharina Aicher
Italy	Bolzano, Via Galvani 38	Sennereiverband Südtirol—Federazione Latterie Alto Adige	Andreas Österreicher and Gianluca Martinelli
Spain	Asturias, Sierra de Granda S/N	CAPSA—Corporacion Alimentaria Peñasanta	Yoana Presedo, Eva Álvarez Segurola and Claudia Pérez Fernández
Sweden	Linköping, Roxtorpsgatan 3	Arla Foods	Kim Vang Nielsen, Per Almberg and Zandra Dahlqvist
Switzerland	Ostermundigen, Milchstrasse 9	Emmi Schweiz AG	Adrian Krebs

Preparation of Test Portions

Milk, dairy, and dairy-containing products were procured and prescreened for lactose content.

The 14 validation materials included seven liquid milk samples, two yogurt samples, three dairy-containing drinks, and two infant formula powders.

Commercially available samples covering the various analytical ranges of the method and SMPR 2018.009 were difficult to find as they are usually lactose free. Thus, besides commercially available samples (UHT milk 3 with 0.5% fat, UHT milk 4 with 1.5% fat, UHT milk 5 with 3.5% fat, UHT milk 6 with 1.8% fat, UHT 7 with 1.5% fat, Chocolate iced coffee with 0.5% fat), we also included samples that have been hydrolyzed by the study authors or collaboration partners using a yeast neutral lactase. For those samples it was important to inactivate the lactase enzyme when the desired lactose level was reached. In case of liquid samples, the lactase inactivation was done by heat treatment (UHT milk 1 with 3.5% fat, UHT milk 2 with 3.5% fat, vanilla milk with 1.4% fat, banana protein drink <0.5% fat) and for yogurt samples (yogurt 1 and yogurt 2 with 3.8% fat each) by acidification during fermentation.

Aliquots of powdered infant formula 1 and 2 (3.5% fat in ready-to-feed liquid) were individually spiked with lactose solution and allowed to dry. Due to clumping of the infant formula upon spiking, it was not possible to spike a bulk sample and obtain a homogeneous distribution of analyte, so individual aliquots were spiked. Duplicate aliquots (approx. 0.5 mL for liquid milk and drinks, approx. 1.0 g for yogurt, and 3.07 g for infant formula) of each material were packaged and blind coded for each participating laboratory. For the MLV, some test portion sizes were smaller than for the SLV due to limited quantities of appropriate homogeneous validation materials to be split among 15 laboratories in duplicate.

Test Portion Distribution

Validation material aliquots were shipped with blue ice packs in one shipment by overnight delivery and collaborators were instructed to store the materials frozen until the day of analysis. Each laboratory analyzed the blind coded materials according to their laboratory schedule within a 2-month timeframe. Most laboratories carried out all analyses on one day, but one laboratory performed analysis over 2 days.

Test Portion Analysis

Test portions were treated and analyzed according to OMA 2020.01. Briefly, for liquids, 0.4 mL liquid test portions were transferred to a clean tube, diluted with 0.4 mL LactoSensR buffer and mixed by shaking or vortexing. From the yogurt aliquots, test portions of approx. 0.8 g were accurately weighed into a clean tube, mixed with an equivalent volume of LactoSensR buffer (e.g., 780 µL buffer added to a 0.78 g test portion) and mixed by shaking or vortexing. Powdered infant formula aliquots (3.07 g) were reconstituted with 20 mL tap water pre-warmed to 45°C as per the ratio described by the infant formula package instructions. A 1 mL test portion of the reconstituted formula was mixed with 1 mL LactoSensR buffer. A 100 µL aliquot of each prepared test solution was then applied to a sensor in the reader and results were automatically reported.

LactoSensR is factory calibrated and labeled with a QR code, linking that lot with the appropriate calibration curve. Data above the calibration ranges are reported as >200 mg/100 g and the method instructions direct the user to dilute the test solution and repeat the analysis. When results are below the calibration range (below the LOQ), the software returns a result of <5 mg/100 g when the infant formula setting is selected or <8 mg/100 g when the standard setting is selected. For the purposes of the MLV, participants were provided with a software option to override the “<5” or “<8” result reporting and instead report a value when the result is below the LOQ. This allows for averaging of results near the LOQ, where some values may fall below.

Statistical Analysis

Data were analyzed by Cochran and Grubbs tests for statistical outliers. Statistical outliers were not removed without justifiable cause. For each validation material, the mean, standard deviation of repeatability (s_r), standard deviation of reproducibility (s_R), relative standard deviation of repeatability (RSD_r), relative standard deviation of reproducibility (RSD_R), and HorRat value were determined. The AOAC INTERNATIONAL Interlaboratory Study Workbook for Blind (unmatched) Replicates, version 4.8, was used for statistical analyses in the C metamer (5).

AOAC Official Method 2020.01

Determination of Lactose in Low-Lactose and Lactose-Free Milk, Milk Products, and Products Containing Dairy Ingredients

LactoSens^®R Amperometry Method

First Action 2020

Final Action 2023

[Applicable for determination of lactose in a variety of liquid, semi-solid, solid, and powdered milk and dairy products in the range 8–200 mg/100 g and powdered infant formula in the range 5–200 mg/100 g produced with yeast-neutral lactases.]

SeeTables 2020.01A and 2020.01B for the results of the multilaboratory study supporting acceptance of this method.

Caution: Use appropriate personal protective equipment such as a laboratory coat, safety glasses and gloves. Dispose of all materials according to federal, state, and local regulations.

A. Principle

The LactoSensR Method is a biosensor assay kit intended for the determination of lactose levels in a variety of lactose-free or low-lactose milk and dairy products. An enzyme immobilized on a disposable test strip oxidizes the lactose in the test portion and the resulting electrons are detected amperometrically by the LactoSensR Reader. Proprietary software converts the electrical signal into a lactose concentration according to the factory-set calibration function. Each test strip is labelled with a QR code for sample tracking and lot-specific information and a ready-to-use positive control is provided for quality assurance and calibration check.

B. Apparatus and Reagents

• LactoSensRReader Kit.—LR10 (DirectSens), including software version 2.0 or higher.

• LactoSensRBiosensor Assay Kit.—LK1225 (DirectSens). Includes 25 LactoSensR Biosensors; LactoSensR Buffer (two bottles, 125 mL each, pH 5.6, ready-to-use, LB10, DirectSens); LactoSensR Positive Control (one vial, 2 mL, ready-to-use, LC12 POC, DirectSens); User Manual; and Quality Assurance Certificate.

• Laptop or PC.—For connection with LactoSensR Reader.

• Vortex mixer.

• Sample vials.—Polypropylene tubes with caps.

• Pipets and tips.—1 mL and 100 µL.

• Centrifuge.—With rotor capable of 10 000 × g, not refrigerated.

• Paddle blender.—Stomacher^® 400 (available from major laboratory suppliers) or equivalent, with sample bags.

• Mortar and pestle.

C. General Preparation

• Bring sample(s), LactoSensR Buffer, and required number of biosensors to room temperature (22 ± 2°C).

• Connect LactoSensR Reader to laptop or PC.

• Turn on LactoSensR Reader.

• Start LactoSens software. Green icon in lower left corner of software window indicates a connected reader.

D. Sample Preparation

Note: Fat content is based on label claims. When fat content is unknown, determine using 989.05 (AOAC-ISO/IDF Method).

• Low-fat liquids (<20% fat).—Mix product by shaking or vortexing. Dilute a test portion with an equal volume LactoSensR Buffer in sample vial and cap. The minimum test portion size is 0.4 mL. Mix by shaking or vortexing.

• High-fat liquids (≥20% fat).—Centrifuge 10 min at 10 000 × g, and transfer aqueous phase to new container. Dilute a test portion of aqueous phase with an equal volume LactoSensR buffer in sample vial and cap. The minimum test portion size is 0.4 mL. Mix by shaking or vortexing.

• Low-fat semi-solids (<20% fat).—Homogenize representative sample using mortar and pestle or paddle blender. Accurately weigh a test portion into sample vial, note exact weight, and add equivalent volume of LactoSensR Buffer. For example, at a minimum weigh a 0.75 g test potion and add 0.75 mL buffer. Cap the vial. Mix by shaking or vortexing.

• High-fat semi-solids (≥20% fat).—Homogenize representative sample using mortar and pestle or paddle blender. Centrifuge homogenized product 10 min at 10 000 × g, and transfer aqueous phase to new container. Dilute 1 mL test portion of aqueous phase with 1 mL LactoSensR Buffer in sample vial and cap. Mix by shaking or vortexing.

• Solids.—Finely chop with knife and macerate product with mortar and pestle. Accurately weigh a test portion into clean mortar or paddle blender bag, note exact weight, and add equivalent volume of LactoSensR Buffer. For example, at a minimum weigh a 2 g test portion and add 2 mL buffer. Homogenize for 1 min.

• Infant formula and powdered milk.—Pre-warm tap water to 45°C in a water bath. Measure appropriate volume of pre-warmed water in graduated cylinder, and then add appropriate amount of powder and dissolve completely by mixing. Reconstitute powder according to package instructions for water volume and powder mass, such as 25 g infant formula powder dissolved in 200 g pre-warmed water. Dilute 1 mL test portion with 1 mL LactoSensR Buffer in sample vial and cap. Mix by shaking or vortexing.

E. Determination

• In the software window, type the sample name or code and click on “Scan sensor QR code.” Hold scanner 10 to 20 cm from the QR code on sensor package with red light pointing directly on QR code. Acoustic signal indicates successful scanning of QR code.

• In “Settings” menu, choose either “infant formula” setting with a quantitation range of 5–200 mg/100g lactose or “standard” setting with a quantitation range of 8–200 mg/100 g lactose depending on the samples being analyzed.

• Open sensor package, being careful not to touch the sensitive sample application area.

• With LactoSens logo facing upward, insert end with silver contacts into reader as far as it will go. Confirm by pressing “Next.”

• For viscous samples, cut pipet tip prior to using to avoid clogging. Transfer 100 µL aliquot of prepared sample onto the sensitive sample application area. Use pipet tip to distribute aliquot and ensure the sensitive area is completely covered. Correct application of aliquot is crucial for accurate measurement. Aliquot application must be completed within 15 s.

• LactoSensR Reader will automatically detect sample, and measurement will start after 15 s.

• Select desired measurement units (g/L or %).

• Choose output file and click “Save and Next.”

• If result is above upper limit of method range, dilute sample (e.g., 10-fold) with LactoSensR Buffer and repeat analysis. Multiply result by dilution factor (e.g., 10).

• Analyze next prepared sample.

F. Correction Factors

High fat.—Samples with fat content ≥20% must be centrifuged for 10 min at 10 000 × g prior to analysis. The aqueous phase is used for dilution with dilution buffer and subsequent analysis. Obtained results must be corrected with the respective fat concentration. Measured values of samples with fat content ≥20% are multiplied by a factor of 100 - × % fat content.

G. System Suitability

• LactoSensRpositive control.—Positive control is provided with kit in ready-to-use format and can be used to verify calibration status and correct handling of sensor. Test 100 µL of positive control solution and verify that the result is 20 ± 5 mg/100 g.

Results and Discussion

The raw data by laboratory and test portion are shown in Tables 2 and 3. Laboratories 10 and 13 did not select the software mode that allows reporting of values below the LOQ, so duplicate test portions where one or both yielded results “<8” were excluded from statistical analysis. In addition, Laboratory 10 had a result of >200 mg/100 g for one test portion of yogurt 2, but did not dilute the test solution and reanalyze, so these duplicates were excluded from the statistical analysis. There were several statistical outliers identified by the Cochran and/or Grubbs tests, including yogurt 2 and banana protein drink test portions analyzed by Laboratory 5, chocolate iced coffee analyzed by Laboratory 9, and infant formula 1 test portions analyzed by Laboratory 14. These are indicated in Table 2, but none were excluded from the statistical analysis since no justifiable cause was identified for any of these cases. Finally, Laboratory 12 followed the wrong sample preparation procedure for infant formula 1, so no data were reported for these test portions.

Table 2.

Raw data for lactose determination by LactoSensR from blind duplicates for seven UHT milk materials

Laboratory	UHT milk 1 mg/100 g		UHT milk 2 mg/100 g		UHT milk 3 mg/100 g		UHT milk 4 mg/100 g		UHT milk 5 mg/100 g		UHT milk 6 mg/100 g		UHT milk 7 mg/100 g
	UJPX5GS3	R6ULUUZV	BYMGH0OG	4U4RBRJR	LJW7XMND	B98RMMON	P911EIV2	AIZH4HO7	12F5PNL7	QZMYFGSU	MZZ7BO81	54RH0KJ9	UY8SR74O	A0UIDN9C
1	102	94	17	19	166	157	0	0	11	11	8	9	7	8
2	104	100	21	21	182	187	0	0	12	11	10	10	9	8
3	91	96	20	19	173	159	0	0	13	12	9	9	8	8
4	103	106	20	19	172	177	0	0	12	12	9	9	8	7
5	94	102	18	22	169	170	0	1	12	13	10	9	8	8
6	106	105	20	20	161	172	0	1	11	14	10	10	9	11
7	101	95	20	20	174	176	0	0	12	12	9	8	8	9
8	102	106	21	20	184	179	0	0	13	14	11	12	9	11
9	107	102	21	21	181	174	0	0	15	13	10	10	10	11
10	102	103	20	23	173	166	<8a	<8a	14a	<8a	9	10	9	9
11	103	95	19	20	169	173	0	0	12	13	9	10	10	9
12	101	103	20	21	178	171	0	0	13	13	10	9	9	10
13	103	108	21	20	170	178	<8a	<8a	12	12	8	9	8a	<8a
14	98	96	18	21	180	179	0	0	14	13	11	11	9	8
15	93	95	19	19	165	165	0	0	12	12	9	9	10	9

a

Laboratory selected incorrect software mode; duplicate pairs removed from statistical analysis.

Table 3.

Raw data for lactose determination by LactoSensR from blind duplicates of yogurt, dairy-containing drinks, and infant formula materials

Laboratory	Yogurt 1 mg/100 g		Yogurt 2 mg/100 g		Vanilla milk mg/100 g		Banana protein drink mg/100 g		Chocolate iced coffee mg/100 g		Infant formula 1 mg/100 g		Infant formula 2 mg/100 g
	C7HZOTM4	OWKD4P46	BLV7D37G	X9L4FR9G	G0E6NYHL	8EWZWOAQ	SFLP7SHJ	IM8WG2ZK	S33UAEHF	3ZSRO44S	4KROXMQ1	T3NVSM4Q	2RMA43IA	6E15U565
1	33	34	131	133	89	92	47	44	0	0	5	5	15	16
2	34	34	129	132	95	99	46	45	0	0	5	5	15	16
3	30	30	125	126	89	95	45	44	0	0	5	5	14	13
4	34	34	139	147	97	96	49	48	0	0	4	4	14	15
5	33	34	135a	119a	97	94	36a	48a	0	0	5	6	15	11
6	35	35	143	143	94	99	50	48	0	0	6	6	15	18
7	32	34	135	136	96	98	47	45	0	0	5	5	15	16
8	41	38	137	136	96	95	47	53	0	0	6	6	17	18
9	39	38	140	143	97	101	50	51	0	3	4	6	16	15
10	33	35	143	>200b	98	102	43	40	<8c	<8c	<8c	<8c	16	16
11	31	32	127	129	97	93	50	45	0	0	4	4	15	15
12	33	33	143	135	104	97	48	50	0	0	NRd	NRd	13	12
13	36	38	142	141	96	82	49	46	<8	<8	<8c	<8c	14	17
14	35	38	143	136	93	91	48	49	0	0	2a	5a	13	13
15	36	33	148	147	92	95	50	47	0	0	5	5	15	16

a

Cochran test outlier; duplicate pair not excluded from statistical analysis.

b

Laboratory did not dilute and reanalyze according to method instructions; duplicate pair excluded from statistical analysis.

c

Laboratory selected incorrect software mode; duplicate pairs removed from statistical analysis.

d

Laboratory did not use correct sample preparation procedure; data not reported.

Table 2020.01A.

Statistical summary of lactose determination results for LactoSensR

Parameter	UHT milk 1	UHT milk 2	UHT milk 3	UHT milk 4	UHT milk 5	UHT milk 6	UHT milk 7	Yogurt 1	Yogurt 2	Vanilla milk drink	Banana protein drink	Chocolate iced coffee	Infant formula 1	Infant formula 2
No. of laboratories	15	15	15	13	14	15	14	15	14	15	15	13	12	15
No. of replicates	30	30	30	26	28	30	28	30	28	30	30	26	24	30
Mean, mg/100 g	101	20.0	173	0.0769a	12.5	9.53	8.89	35	136	95.3	46.9	0.115a	4.92a	15.0
sr, mg/100 g	3.45	1.21	4.85	NA	0.824	0.516	0.779	1.20	4.07	3.68	2.93	NA	NA	1.20
sR, mg/100 g	4.71	1.24	7.35	NA	1.01	0.949	1.11	2.65	7.44	4.27	3.42	NA	NA	1.64
RSDr, %	3.44	6.06	2.81	NA	6.61	5.42	8.76	3.47	2.98	3.86	6.25	NA	NA	8.00
RSDR, %	4.69	6.16	4.25	NA	8.07	9.95	12.5	7.67	5.45	4.47	7.28	NA	NA	11.0
HorRat	0.829	0.855	0.817	NA	1.04	1.24	1.53	1.16	1.01	0.785	1.15	NA	NA	1.46

a

Mean value is below the LOQ; statistical analyses not reported.

Table 2020.01B.

Comparison of LactoSensR results to SMPR 2018.009 criteria

	Infant formula		Milk, milk products, and products containing dairy ingredients
Parameter	SMPR	LactoSensR	SMPR	LactoSensR	SMPR	LactoSensR
Analytical range, mg/100 g	5–100	5–200	10–100	8–100	>100–2000	>100–6655
Single Laboratory validationa
Recovery, %	85–115	91.1–97.0	85–115	85.0–110	90–110	85.6–110
RSDr, %	≤10	1.0–6.8	≤10	0.0–10.5	≤7	0.2–5.4
Multi-Laboratory validationb
RSDr, %	≤10	8.00	≤10	3.47–8.76	≤7	2.81–3.44
RSDR, %	≤15	11.0	≤15	4.47–12.5	≤10	4.25–5.45

a

Data from single-laboratory validation study (4).

b

Data from current study.

The summary of statistical results is found in Table 2020.01A. There were three validation materials for which the mean lactose concentration was below the LOQ, so further statistical analyses were not reported. For the remaining materials, RSD_r values varied from 2.81 to 8.76% and RSD_R values varied from 4.25 to 12.5% on materials ranging in mean lactose concentration from 8.89 to 173 mg/100 g. HorRat values, the ratio of observed RSD_R to predicted RSD_R, ranged from 0.785 to 1.53, which are within the acceptable limits of 0.5 to 2.0 (6). When sorted by category and concentration range, these results compared favorably to the method performance requirements set forth in SMPR 2018.009 (2) as shown in Table 2020.01B. The data from this collaborative study, when sorted by food category and analytical range, met the acceptance criteria for repeatability (RSD_r) and reproducibility (RSD_R).

Conclusions

The MLV study data supports the acceptance of the LactoSensR method, OMA 2020.01, for final action status.

Acknowledgments

The authors sincerely thank the study collaborators listed in Table 1 for their valuable contributions to this study.

Conflict of Interest

Elisabeth Halbmayr-Jech, Roman Kittl, Patrick Weinmann, Christopher Schulz, Anna Kowalik, Matthias König, Jasmin Korp and Christoph Sygmund work for DirectSens GmbH and received a salary for this work.

Sharon L. Brunelle received payment from AOAC INTERNATIONAL for preparation of the manuscript.