Evaluation and comparison of Sensory, physicochemical and rheological properties of labane formulation produced by wheyless process using principal component analysis (PCA) and partial least squares regression (PLS regression)

Labane or concentrated yoghurt is a semi-solid product which is being producedd by dehydration of yoghurt or separation of water and some water-soluble solids from yoghurt. High nutritional value, long shelf life, more desirable taste and texture and feasibility to produce other product are the reasons for high acceptance among customers. There are wide ranges of production methods from traditional methods which are tedious, such as separation of whey by fabric bags, to complicated time-consuming methods which also contaminate product and reduce its nutritional value. These cause demand to establish more suitable methods such as “wheyless process” by dried milk, concentrated milk protein or concentrated whey protein. On the other hand, one of the methods for enhancing firmness and textural properties of yoghurts, similar to other dairy products is utilization of hydrocolloids. These components are used for enhancing rheological an textural properties of food and commonly used as additives for increasing viscosity, gel forming ability, enhancing physical stability, film forming ability, controlling crystallization, postponing syneresis and textural improvement. Konjac gum (KG) is a neutral polysaccharide that derived from Amorphophallus konjac C. Koch tuber, which is well known in east countries during centuries. Ability to hold water and reduction of cholesterol and glucose are the reasons for grossing demand of this gum. So due to the importance of labane produced through wheyless process, the effect of different ingredients of formulation such as milk protein concentrate, cheese whey powder and konjac gum on physicochemical, rheological and sensory properties of labane was investigated and compared these properties by using Principal component analysis (PCA) and Partial least squares regression (PLS regression).

Cow milk was purchased from local market (Mashhad, Iran). Dried skim milk, milk protein concentrate, cheese whey powder   and commercial starter provided from Khorasan Razavi Pegah Company (ABY1, Christian Hansen, Horsholm, Denmark) and KG purchased from Food Chem. (China). In order to concentrate the milk, 3% dried skim milk was added to1 Kg milk. Then, according to the experimental design (table 2), MPC and CWP added to milk at 0, 4, 6 and 8% levels which reached the total solid of milk to 21.5%. KG with percentage of 0, 0.05, 0.1, 0.15, and 0.2% was added to the mixture at 40˚C. The mixtures were homogenized at 50˚C and pasteurized at 90˚C for 1 min. and then cooled to 43-45˚C. The mixtures were incubated at 43-45˚C for 3-4 hours after addition of starter. Then the samples were slowly stirred and packed in polyethylene bags after reaching to suitable pH. The samples were transfered to refrigerator for 24 hours then the tests were conducted. PH was measured according to AOAC 2005 official method NO. 935.42.25 gr of samples were centrifuged at 4˚C for 10 min. at 4500 rpm. Syneresis was evaluated as parentage of separated serum. Texture analyzer was used for combination of back extrusion and texture profile analysis (TPA) test. 50 mm in diameter cylinder with 10 mm in height and probe with 4 mm diameter and 100 mm height with the speed of 1 mm/s were used to conduct the test. Dynamic rheological parameters were evaluated by rheometer (Parphysica). The devise was equipped by 50mm diameter parallel plate with 2 mm gap. Eheo plus/32 version V3.40 software was used to measure elastic modulus (G’), loss modulus (G”) and η*.

Results showed that none of the linear, quadratic and interaction effect of MPC, CWP and KG was significant on pH of the samples at the first day. Results represented that the pH of the samples was varied from 0.36 to 0.94 after 5 days and only linear effect of MPC and CWP (at 99% level of confidence) and interaction effect of CWP-KG (at 95% level of confidence) were significant on that. The effect of MPC and CWP and interaction effect of MPC-KG, CWP-KG and CWP-MPC-KG were significant on syneresis at 99% of confidence. Results showed that hardness of the samples varied between 3.25-9.58 N and the interaction effects of MPC-KG, CWP-KG and CWP-MPC-KG were significant at 99% of confidence. None of the linear, quadratic and interaction effect of CWP, MPC and KG was not significant on springiness of the samples (p>0.05). In strain sweep test, two separated regions were distinguishable: linear viscoelastic (LVE) region that elastic and viscous modulus (G’ and G”) were constant and G’ was higher than G” which showed the solid like behavior, and non-LVE region that G’ and G” decreased by increasing strain and led to a crossover point (flowing point) which G” goes over the G’ and liquid-like behavior shows itself. In this test, strain corresponding to start of the non-LVE region and sharp reduction of G’ define as critical strain (γc) and the corresponding stress of this point defined as critical stress (τc).Results represented that G’LVE and G” LVE of the samples varied from 22.54 to 750.1 and 11.01 to 242.1 Pa, respectively and the selected model (cubic x quadratic) showed that the interaction effect of MPC-KG, CWP-KG and CWP-MPC-KG on G’LVE and G” LVE was significant at 99% of confidence while the interaction effect of CWP-MPC was only meaningful on G” LVE. Results also showed that the variation trend of G’LVE and G” LVE was the same by alteration MPC, CWP and KG in a way that both of these parameters were increased by increasing KG. The effect of MPC-KG and CWP-KG was significant on γc and τc at 95% confidence and also, in addition to mentioned effect, the interaction effect of CWP-MPC was significant only on γc. Increasing KG increased γc at higher amount of CWP and lower amount MPC which can be concluded that increasing KG amplified the gel strength. Also, the results of the correlation study between sensory and instrumental measures showed that there was a close relationship between the score of sour taste and the overall acceptance of samples, which showed that among the measured characteristics, sour taste score is more important than other parameters.