Assessment of the microbiological quality of koozh, a fermented millet beverage

Koozh is a fermented beverage made with millet flour and rice, and consumed by ethnic communities in Tamil Nadu, India. Six street vended samples were assessed for the total bacterial count (TBC), lactic acid bacteria (LAB) count, yeast-mould count (YMC), coliforms at 35°C and pathogens. The koozh pH ranged from 4.3 to 4.9 with high titratable acidity. Although no Staphaylococcus sp. and Listeria sp. were found, high colony counts of Clostridia sp., Salmonella sp. and Shigella sp. were present in some samples. The LAB was dominant as compared to TBC, YMC and coliforms. Pathogens were detected, indicating contamination following processing in the traditional fermented food.


INTRODUCTION
Millets are important minor cereals in tropical and subtropical regions and India is the largest producer. Many traditional fermented products are made from millets both in African and Asian countries. Koozh (Tamil term for porridge) is a ready-to-eat (RTE) food/energy beverage either made from finger millet-Eleusine coracana (Kezhvaragu) or pearl millet-Pennisetum glaucum (Kampu), broken rice flour (noyee) (Kumar et al., 2010).
Koozh is prepared in an outdoor traditional kitchen common in rural India, using traditional methods that have been followed for generations. A flow chart of the typical process is given in Figure 1. The millet flour is made into slurry with water by hand-mixing on the first day and left to ferment overnight (12-15 h); on the second day, broken rice (20% by weight of millet) is cooked in excess water, into which the fermented (12-15 h) millet slurry is mixed, stirred and cooked to make a thick porridge called noyee.
The fermentation of this porridge overnight (24 h) results in kali, a semi solid porridge to which the required amount of potable water is added (1:6 w/v) and handmixed with salt to prepare koozh. The product may be further mixed with buttermilk (diluted curd a local preparation of yoghurt) to give a thin porridge consistency. The end product has a characteristic fermented flavour as a result of the microbial succession, which develops complex flavours. While kali has a shelf life of approximately one week at room temperature (25-30°C), koozh has low shelf life and is usually consumed within about 12 h of preparation. It is prepared in homes and offered in temples during special festivals. It is also sold in the streets or by mobile food vendors and consumed by daily-wage earners of Tamil Nadu, India.
Koozh is unique in that it is fermented twice-once before cooking and again after cooking, its preparation last for two days. Traditionally, koozh is considered nutritious and health promoting, but there is little scientific documentation on its nutritive or microbial composition, other than some studies on the nutritional advantages of fermented millets -finger millet (Antony and Chandra, 1998) and pearl millet (Kheterpaul and Chauhan, 1994). Geetha and Kalaichelvan (2013) had reported the microbial succession and biochemical changes in koozh made from finger millet-ragi, pearl millet-cumbu, sorghum and maize. The second fermentation after cooking of the millet in koozh preparation makes it an excellent source of live bacteria, while its storage and service may allow contamination. Hence, this study was undertaken to assess the microbiological safety of koozh sold as street food.

Sample collection and its microbial load
Finger and pearl millet koozh were collected in 250-mL autoclaved, wide-mouthed, screw-capped plastic containers from market places in Salem district (Sa) and Chennai district (Ch), Tamil Nadu, India and immediately transported to the laboratory in insulated food containers with ice packs and analyzed. The finger millet koozh samples were collected from street vendors in different Chennai suburbs and tested on the same day according to Cappuccino and

Determination of pH
The sample pH was determined using AP-1 plus pH meter (Susima Technologies, Chennai). The titratable acidity was estimated in koozh filtrate by titration with 0.1 N sodium hydroxide to the end point with phenolphthalein indicator (AOAC, 2000).

Statistical analysis
The microbial count values were tested using paired student's t test and the correlation between variables was also determined using GraphPad 6 (SanDiego,Ca, USA) software. The calculated r values are interpreted.

RESULTS AND DISCUSSION pH
The pH of koozh ranged from 4.3 to 4.9 (Table 1) with acidity ranging from 0.16 to 0.35%, in all samples. This may pose a problem because some samples with pH over 4.5 may allow spoilage or growth of pathogenic microbes. This observation is similar to other milletfermented products reported in literature: an alcoholic Himalayan beverage from finger millet called Kodo ko jaanr, with pH ranging from 3.7 to 4.5 (Thapa and Tamang, 2004); the Northern Ghana's Koko sour water from pearl millet with a pH of 3.9 ± 0.1 (Lei and Jakobsen, 2004); a fermented mixture of millet and sorghum flour, called bushera of Uganda with 3.7 ± 0.1 pH, and bushera made only with sorghum with a pH range of 4.0-4.5 (Muyanja et al., 2003).

Bacteria, LAB and yeast-mould counts
In all koozh samples, LAB were found to be dominant and yeast-mould counts were comparatively lower. LAB counts on MRS showed significant differences (p ≤ 0.05) with TBC and counts on M17 and yeast counts. The LAB counts on MRS showed a very strong correlation with counts on M17 (r = 0.9396) as both are selective media used for LAB enumeration. Yeast counts were strongly correlated with LAB count on MRS (r = 0.7205) and moderately correlated with counts on M17 (r = 0.6261). Co-metabolism between yeast and LAB may exist, where the bacteria provide the acid environment, which selects the growth of yeast, that in turn; provide vitamins and other growth factors to the bacteria (Steinkraus, 1996). The pH of samples correlated strongly with LAB counts on MRS (r = 0.7250) apparently due to the acid production by LAB, and correlated moderately with counts on M17 (r = 0.5950), while with yeast the correlation was low (r = 0.2327). Moulds were absent in all samples tested. Thus, it can be concluded that LAB and yeasts mediate koozh fermentation.

Bacterial pathogen (hazard) classification
The isolated pathogens were Bacillus cereus, Clostridium sp., Enterobacteria sp., Salmonella sp. and Shigella sp. Coliforms at 35°C were found in all samples (Figure 2) and the count was positively correlated with TBC (r = 0.7216) and are probably from contamination of the water added and the preparation involving hand-mixing. Their presence is associated with poor hygiene, and therefore, indicates a potential health risk. The presence of coliforms has been reported in other traditional fermented products due to its survival in acidic environment (Sterinkraus, 1996).
Coliforms had moderate correlation with pH (r = 0.5521) and a negative correlation with yeast (r = -0.4992) indicating inhibitory action of the microbes on coliforms. Simango (1995) compared the contamination in a fermented cereal gruel (Mahewu) with that of a non fermented thick maize meal pap (Sadza) in Zimbabwe. After storing for 4 h, both foods were found to have 50% Escherichia coli contamination but after 24 h no E. coli was found in the Mahewu. Simango and Rukure (1991) stated that this is due to further fermentation that leads to the inhibition of these contaminants during longer storage. Traditionally fermented laboratory-prepared sorghum bread had coliform in the first fermentation, but the microbes were not detected in the second fermentation (Gassem, 1999). In bushera, coliforms population decreased due to high acidity resulting from the metabolism of LAB (Muyanja et al., 2003).
The hazard classification based on pathogen count according to HPA (2009) for RTE is represented in Figure  2. The presence of Clostridium sp., Salmonella sp. and Shigella sp. in high numbers made three samples unsatisfactory. In one of these koozh samples pH was exceeded 4.5.
Other pathogens like Listeria sp. and Staphylococcus sp. were not detected in any of the samples. Sample S 3 -Ch alone met the microbial stan-dards, indicating safe and hygienic handling of the pro-duct. It was also the product with the lowest pH. Two samples (S 5 -Ch and S 6 -Ch) with B. cereus and Clostridum sp. at low levels were acceptable.
Studies with other fermented foods show that pathogens are inhibited during the fermentation. Pathogens such as enterotoxigenic E. coli, Shigella flexneri, Salmonella typhimurium, B. cereus, and Campylobacter jejuni, when inoculated were adversely affected during the fermentation of sorghum (Kingamkono et al., 1994). The fermentation of finger millet provides antimicrobial activity against S. typhirmurium and E. coli even after 48 h of fermentation (Antony and Chandra, 1998). The inhibitory effect in fermentation is not only because of lactic acid but other factors like other organic acids, fermentation process, pH and temperature.
Presence of Clostridium in heat-treated foods would be due to inadequate cooking or post-processing contamination (HPA, 2009). B. cereus survives on processing equipments by germinating prior to sanitization. Handling, storage, or processing can be sources of contamination. Bacteria attaches to surfaces of food equipments, made in polystyrene, hydroxyapatite, glass, rubber and stainless steel might transmit pathogens to food (Mafu et al., 2011). The major contamination is through unclean water used for dilution and hand-mixing. Direct contaminations from unhygenic environment are predominant in market places and in tropical regions were these locations attract house-flies. Koozh and accompaniments were found exposed to dust and pathogens; serving utensils were not cleaned properly due to lack of safe running water. All these factors resulted in post process microbial contamination.

Conclusion
The microbiological quality of koozh traditionally considered healthy for its nutritional content, and sold as a street food varies widely. Lactic acid bacteria and yeasts were present in significant numbers. Hazardous levels of pathogens in 3 of 6 samples indicate unhygienic handling, despite the large numbers of LAB. Safety awareness programmes targeting producers are imperative to eliminate such contamination and ensure safe food.