Nutritional and alcoholic contents of cheka: A traditional fermented beverage in Southwestern Ethiopia

Abstract Cheka is a cereal and vegetable‐based beverage which is consumed in Southwestern parts of Ethiopia particularly in Dirashe and Konso. In this study, nine cheka samples were collected from vending houses in Konso and Dirashe districts for the laboratory analysis of the nutritional profile and chemical properties of cheka. The pH and titratable acidity of the samples ranged from 3.53–3.99 and 0.80%–1.11%, respectively. The total solids, crude protein, crude fat, crude fiber, total ash, carbohydrate, and gross energy contents of the samples ranged from 21.05%–26.87%, 3.12–4.44 g/100 g, 1.17–1.81 g/100 g, 0.94–1.27, 0.65–0.93 g/100 g, 14.16–19.03 g/100 g, and 82.04–107.17 Kcal, respectively. The dietary Ca, Fe, and Zn content of the samples were ranged from 8.31–19.60 mg/100 g, 13.94–27.59 mg/100 g and 0.82–1.07 mg/100 g, respectively. The methanol and ethanol contents of the cheka samples ranged from 163.1–2,380 ppm and 3.04%–8.96% v/v, respectively. The findings of this study indicated that cheka has low nutrient content and thus, suggests that people in Konso and Dirashe should not rely on it without eating solid foods as it is almost always diluted with a significant amount of water. In conclusion, the longer fermentation time of cheka resulted in high methanol levels that can present adverse health effects to consumers.

excess quantities due to their relatively low price compared to factory produced drinks. This will help in reducing health complications and in avoiding other social problems associated with alcohol abuse (Sarah & Mattew, 2012).
Cheka is a cereal and vegetable-based fermented beverages which is consumed in Southwestern parts of Ethiopia mainly in Dirashe and Konso. People of all ages including infants, pregnant, and lactating women drink cheka. As adults who eat solid foods are considered childish in the communities, it is cheka that is being consumed all day long and from observation an adult man on average drinks up to 8 L of cheka per day. Several works have been done on Ethiopian fermented beverages such as tella (Desta, 1977;Yohannes, Fekadu, & Khalid, 2013), tej (Yohannes et al., 2013), arake (Yohannes et al., 2013), shamita (Ashenafi & Mehari, 1995), borde (Abegaz, Beyene, Langsrud, & Judith, 2002a), and kerebo (Rashid, 2013a). However, no work has been done on any aspect of cheka until present. Therefore, this study was intended to (a) determine the nutritional content of cheka and (b) detect and quantify the alcoholic contents of cheka using a gas chromatographic technique.

| Sample collection
Nine cheka samples were collected using screw-cap plastic containers (1 L each) from vending houses at three localities for analysis of their chemical properties, nutritional, and alcohol contents. All the samples were collected purposively while considering the processing techniques and duration of fermentation. The samples were coded with the first letter of the site and numbers in the order they were collected and were transported to the laboratory of the Centre for Food Science and Nutrition at Addis Ababa University using an icebox. In addition, one sample of cheka was prepared by the investigator in the laboratory following the Konso preparation method using yellow maize, sorghum grains, and barley obtained from local markets of Addis Ababa.

| Sample preparation
After transportation of the samples, moisture analysis and determinations of pH and titratable acidity were immediately carried out in the laboratory. For convenience of proximate analysis, the samples were lyophilized and were packed in polyethylene plastic bags to be stored in dry place. But the methanol and ethanol contents of the cheka samples were determined in liquid form.

| pH and titratable acidity
The pH of the samples was measured by dipping the glass electrode of a digital pH meter into 10 ml of the sample after blending with distilled water at a 1:1 ratio into thick slurry as described in Abegaz, Beyene, Langsrud, and Judit (2002b). For the determination of total titratable acidity, about 10 ml of cheka samples were added into beakers (50 ml) and titrated against 0.1 N standard solution of NaOH after adding 3 drops of 1% phenolphthalein indicator (Byaruhanga, 1998;cited in Rashid, 2013b). The percent of lactic acid present in the sample was calculated using the following formula.
where; N = normality of titrant (mEq/ml), V NaOH = Volume of titrant (ml), Eq. wt = Equivalent weight of predominant acid (mg/mEq which is 90.08 for lactic acid), V s = Volume of sample (ml) and 1,000 = factor relating mg to grams.

| Proximate composition analysis
The total solids, crude protein, crude fat, crude fiber, and total ash contents of the samples were analyzed according to the AOAC (2000) methods and ASEAN manual of food analysis (2011). The utilizable carbohydrate content was determined by subtracting the sum of the percentages moisture, crude protein, crude fat, crude fiber, and total ash from 100%. The calorific value of the cheka samples were determined by calculation from protein, fat, and carbohydrate contents using Atwater's conversion factors (Guyot, Rochette, & Treche, 2007).

| Alcohol analysis
The ethanol and methanol contents of the cheka samples were determined by a gas chromatography method (GC 2010 Plus, Shimadzu Scientific Instruments, Kyoto, Japan). The samples were prepared according to the method developed by Tangerman (1997). The gas chromatographic conditions were set following the method developed by Wang, Choong, Su, and Lee (2003).

| Sample preparation
About 10 ml of cheka was measured in a plastic test tube and ultracentrifuged for 2 hr at 4°C and 30,000 g. About 5 ml of the cheka supernatant was carefully removed and transferred into a conical polypropylene tube and centrifuged for 1 min at 10,000 g.
The clear supernatants were stored in a refrigerator and later used for the chromatographic analysis. For analysis, the supernatant % Lactic acid (wt∕v) = N × V NaOH × Eq. wt × 100 V s (ml) × 1,000 was filtered through micro filter (0.45 m PTFE membrane) into vial (1.8 ml).

| Preparation of standard solution
Series of solutions ranging from 0.1%-15% and 0.005%-2% were prepared to make a standard calibration curves for ethanol and methanol, respectively. The limit of detection (LOD) and limit of quantitation (LOQ) were determined by multiplying the standard deviation obtained from 10 repeated injections of solutions with low concentrations of 0.005% (methanol) and 0.1% (ethanol) by 3 and 10, respectively. The recovery test was done by spiking a known amount of methanol (0.5 and 5 ml of 1% solution which were equivalent to 5 and 50 μl, respectively) and ethanol (0.5 and 5 ml of 5% solutions which were equivalent to 25 and 250 μl) into 10 ml of cheka.

| Gas chromatography conditions
Ethanol and methanol analysis was performed using Shimadzu GC-2010 Plus gas chromatograph which was equipped with an FID detector, AOC-20i+S autosampler and with a GC solution software for data handling system. The length, inner diameter, and film thickness of the column were 30 m, 0.25 mm and 0.25 μm, respectively. The flow rates of H 2 and N 2 gas were set at 30 and 300 ml/min, respectively. The temperatures of the FID detector and the injection port were set at 300 and 225°C, respectively. The column temperature was set initially at 45°C for 2 min and then ramped at a rate of 45°C/ min to the final 245°C. The injection volume was limited to 1.0 μl using split injection mode.

| Data analysis
Determinations were done in duplicates for the need of statistical analysis. Data were computed using SPSS (version 20) statistical software packages. Data were expressed as mean ± standard deviations of the replicate determinations. One-way analysis of variance (ANOVA) was used to study the significant difference between the samples with respect to the studied parameters. Least significant difference (LSD) at p < 0.05 was used to determine which means were significantly different.

| pH and titratable acidity
pH and titratable acidity are one of the most important parameters that determine the flavor and shelf-life of a food product. In this study, the pH and titratable acidity of the cheka samples ranged from 3.53-3.99 and 0.77%-1.11%, respectively (  (Yohannes et al., 2013), tej (Gizaw, 2006;Yohannes et al., 2013), borde (Abegaz et al., 2002b), shamita (Ashenafi & Mehari, 1995), kerebo (Rashid, 2013a), and arake (Yohannes et al., 2013). On the other hand, ready to consume cheka had titratable acidity comparable to borde (Abegaz et al., 2002b) but lower titratable acidity than kerebo (Rashid, 2013a). Since the pH of the samples were measured once immediately after they had been brought to the laboratory, the investigators believe that the pH of cheka can even be lower than the reported values in this study if determined over a period of time until it turns not safe to consume. Evidences point out that the low pH of beverages can rob calcium in skeletal systems and lead to dental caries and osteoporosis (Cairns, Watson, Creanor, & Foye, 2002;Mettler, Carmen, & Paolo, 2006). Therefore, if the pH of cheka drops below the reported values (especially after 2 days of consumption), it can cause harms to the skeletal system of the consumers.

| Proximate composition
In this study, the total solids content of cheka varied between 21.05% (G2) and 26.87% (S3). There were no significant differences It was found that the crude protein and fat contents of the cheka samples ranged from 3.02 to 4.44 g/100 g and 1.17 to 1.81 g/100 g, respectively ( Table 2). Two of the samples collected from Gidole town (G1 and G3) had significantly (p < 0.05) high protein content of 4.44 g/100 g and 4.40 g/100 g, in that order and were followed by the sample coded as S3 (4.19 g/100 g) which was collected from Shelele kebele. On the other hand, the cheka sample prepared in the laboratory had a significantly low crude protein content (3.02 g/100 g) when compared to the samples S3 (4.19 g/100 g), G1 (4.44 g/100 g), and G3 (4.40 g/100 g).
The cheka samples had high mean protein content (15.68 g/100 g on dry matter basis) when compared to borde and shamita which have about 9.55 g/100 g and 10.37 g/100 g protein on dry basis, respectively (Ashenafi & Mehari, 1995). Sample K1 (1.81 g/100 g) had the highest fat content than all other samples whereas samples G1 (1.35 g/100 g), G2 (1.17 g/100 g), and G3 (1.37 g/100 g) had the least fat contents. But statistical analysis of the result showed no significant (p > 0.05) variation between the samples coded as K2 (1.44 g/100 g), K3 (1.42 g/100 g), and S1 (1.42 g/100 g). The crude fat content of CL (1.49 g/100 g) was significantly lower than that of K1 (1.81 g/100 g) but higher than the rest of the sample. Cheka samples had a mean fat content of 6.11 g/100 g on dry matter basis which is comparable with the fat content of borde (6.88 g/100 g), but more fat content than shamita (3.46 g/100 g) (Ashenafi & Mehari, 1995 The crude fiber content of cheka prepared in the laboratory was comparable with that of the samples collected from the study areas.
On the other hand, the gross energy content of cheka samples varied from 82.04 Kcal/100 g for G2 to 107.17 Kcal/100 g for K3.
Significant variations were observed among the samples and generally samples from Gidole (G1, G2, and G3) had low gross energy.
The ash content of cheka samples ranged from 0.65-0.93 g/100 g.
The sample coded as S3 had the higher ash content (0.93 g/100 g) than samples K2 (0.65 g/100 g), S1 (0.71 g/100 g), S2 (0.68 g/100 g), G2 (0.73 g/100 g), and G3 (0.74 g/100 g), but it did not significantly (p > 0.05) differ from K1 (0.82), K3 (0.76 g/100 g), and G1 (0.76 g/100 g). No significant variation was observed between CL (0.67 g/100 g) and most of the samples collected from the study areas in their ash contents except S3 (0.93 g/100 g). The sample represented by K2 had the least ash content though it did not significantly differ from most of the samples except S3. The cheka samples analyzed in this research had lower average ash content on dry matter basis (3.15 g/100 g dry basis) than borde and shamita which have mean ash contents of 6.85 g/100 g and 3.66 g/100 g, respectively (Ashenafi & Mehari, 1995). The variations observed in the proximate composition of the samples greatly reflects the differences in some of the raw materials utilized for cheka preparation and the fermentation time as well.

| Mineral composition
In the present study, one of the most important mineral elements such as calcium, iron, and zinc were analyzed. As presented in Table 3 Samples collected from Gidole (G1, G2, and G3) and Shelele (S1, S2, and S3) had significantly higher calcium than those collected from Karat ( Figure 1). One of the samples from Karat, K1 (1.07 mg/100 g), had the highest zinc content followed by samples K3 and S1, and S3 all having about 0.95 mg/100 g. The cheka sample produced in the laboratory had comparatively low mineral content than most of the collected samples. The relatively low mineral contents of the cheka samples collected from Karat could be due to the fact that brewers in Konso are highly market oriented and utilize only grains because it helps them produce cheka frequently.

| Methanol content
The methanol content of the cheka samples ranged from 0.0163% to 0.2385% (v/v) ( Table 4). The sample S3 (2,384.4 ppm) had significantly (p < 0.05) high methanol content than the remaining samples and was followed by K1 (0.1630% v/v) and S2 (0.1361% v/v). The lowest methanol content was recorded for sample S1 (0.0163% v/v), but it did not significantly differ from other samples except K1, S2, and S3. Generally, samples from Shelele contained significantly high amount of methanol. The reason for high methanol content in Dirashe could be due to the longer fermentation time(more than a month) that allows more pectins in the product to be degraded by pectinase enzymes into methanol (Singkong, Rattanapun, & Kaweewong, 2012). As presented in Table 4, the high methanol content of some of the samples correspond to the high ethanol content.
The amount of methanol reported in this study for most samples is much higher than the methanol contents reported for tella (32.37 ppm) and tej (45.67 ppm), and arake (320.87 ppm) (Fite et al., 1991). The samples coded as K1, S2 and S3 had much higher methanol content than the specifications for maximum methanol and wine (ESA, 2013). All samples had methanol content higher than the maximum limit specified by East African Standards for gin (EAS, 2013) and also half of the samples had more methanol content than the limit set by EU regulation (cited in Paine & Davan, 2001). This shows that there might be the possibility of methanol toxicity in the study localities where the fermentation time is longer.

| Ethanol content
The alcohol content of the cheka samples analyzed varied between 3.05% v/v and 8.96% v/v ( F I G U R E 1 Chromatogram of methanol and ethanol detected from cheka sample of fresh flour once the fermentation started, but fresh flour must be added into Dirashe cheka 1 day in advance of cooking the dough balls. Based on this finding, the ethanol content of the cheka samples was comparable with that of tella (2.5%-14.52%) and tej (6.2%-14%) (Desta, 1977;Gizaw, 2006;Sahle & Gashe, 1991;Yohannes et al., 2013). However, the alcohol content of the cheka samples was much lower than that of arake which has alcohol content as high as 48% v/v (Gizaw, 2006).

| CON CLUS ION
As people in the study areas use cheka after diluting with sufficient water due to its thick consistency, high alcohol content and low acidity, the crude protein, fat, fiber, carbohydrate, ash, and energy content of the cheka can be lower than the values reported in this study. This necessarily indicates that adults who rely on cheka are still needy for the consumption of other solid foods in order to meet their daily nutrient and energy requirements. The cheka samples with longer fermentation time had higher methanol contents to levels that can pose adverse effects on the health of consumers.
In addition, cheka had high ethanol levels that might be not safe for some individuals including pregnant and lactating women, children and adolescents. From food safety point of view, investigations on the mechanism of cheka production and means to avoid unpleasant contents are necessary.

ACK N OWLED G M ENT
The authors would like to thank Addis Ababa University Centre for Food Science and Nutrition for allowing us use their laboratory and providing necessary facilities for the analysis of the samples.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

E TH I C A L S TATEM ENT
This study did not involve any human or animal testing.