5
Critique of the Pilot Program

Uniformity of Blended Commodities

Four of the six U.S. plants producing corn-soy blend (CSB) were assessed to determine plant performance and conformity to good manufacturing practices (GMPs). Attainment of target fortification levels for vitamin C in CSB requires understanding the inherent capability of both the dosing-metering unit operation and the blending equipment. CSB is made by a continuous process in which very low percentages of minor ingredients are metered into a large bulk stream. In this type of mixing, there is the distinct possibility that minor ingredients may remain segregated or unmixed. Failure to achieve uniform dispersion in a continuous process is a common food processing concern.

The sampling analysis submitted by SUSTAIN (Sharing United States Technology to Aid in the Improvement of Nutrition) provides ample evidence of a problem in the uniformity of finished CSB product. Analytical sampling for vitamin C content showed significant failure to achieve consistent enrichment levels of 40 mg or 90 mg per 100 g of commodity. Extensive plant-to-plant variability was observed, supporting the notion that plants encountered technical challenges in metering and blending the vitamin premix. SUSTAIN data indicated that two plants (C and E) were more capable than plants A and D in achieving acceptable compliance levels. It should be noted that none of the four plants consistently achieved target fortification levels of vitamin C at either doseage.

Wheat-soy blend (WSB) is prepared in a batch system whereby all ingredients are mixed simultaneously. It is easier to obtain homogeneous mixing with a batch process than with a continuous process, but batch processing is suitable for only relatively small amounts of product. WSB production on a batch system was sampled at one of the two U.S. manufacturing sites: plant B was capable of consistently achieving the conventional target of 40 mg/100 g



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 47
--> 5 Critique of the Pilot Program Uniformity of Blended Commodities Four of the six U.S. plants producing corn-soy blend (CSB) were assessed to determine plant performance and conformity to good manufacturing practices (GMPs). Attainment of target fortification levels for vitamin C in CSB requires understanding the inherent capability of both the dosing-metering unit operation and the blending equipment. CSB is made by a continuous process in which very low percentages of minor ingredients are metered into a large bulk stream. In this type of mixing, there is the distinct possibility that minor ingredients may remain segregated or unmixed. Failure to achieve uniform dispersion in a continuous process is a common food processing concern. The sampling analysis submitted by SUSTAIN (Sharing United States Technology to Aid in the Improvement of Nutrition) provides ample evidence of a problem in the uniformity of finished CSB product. Analytical sampling for vitamin C content showed significant failure to achieve consistent enrichment levels of 40 mg or 90 mg per 100 g of commodity. Extensive plant-to-plant variability was observed, supporting the notion that plants encountered technical challenges in metering and blending the vitamin premix. SUSTAIN data indicated that two plants (C and E) were more capable than plants A and D in achieving acceptable compliance levels. It should be noted that none of the four plants consistently achieved target fortification levels of vitamin C at either doseage. Wheat-soy blend (WSB) is prepared in a batch system whereby all ingredients are mixed simultaneously. It is easier to obtain homogeneous mixing with a batch process than with a continuous process, but batch processing is suitable for only relatively small amounts of product. WSB production on a batch system was sampled at one of the two U.S. manufacturing sites: plant B was capable of consistently achieving the conventional target of 40 mg/100 g

OCR for page 47
--> but operated below target for the higher vitamin C level. Even though the batch process delivered the best performance, it is recognized that WSB represents only a small fraction of total blended products distributed by the Public Law (P.L.) 480 Title II program. Capability of the Production Process to Meet Product Specifications The target vitamin C values for conventional and higher levels of fortification are 40 and 90 mg per 100 g, respectively. Many factors combine to produce variation in the measured values of vitamin C in samples from production runs. When processes are in control, vitamin C is approximately normally distributed. Consequently, 99.7 percent of the values will be within 3 S.D. (standard deviations) of the mean. When making control charts, extremes such as the mean ±3 S.D., are called the upper and lower control limits (UCL and LCL). Specifications (SPECS) refer to the limits that define acceptable product. Generally these are determined by the customer, and any product outside these limits is reprocessed or discarded. Appropriate SPECS depend on the product and the property being measured. For example, a 12-ounce soft drink product might have a lower SPEC only for volume (if the can contains less than 12 ounces the customer has not received what has been paid for, but excess product is acceptable). On the other hand, the sugar content would have upper and lower SPECS; a drink that is either too sweet or not sweet enough is unacceptable. For some products an 80 percent rule is used: a product that contains at least 80 percent of the quantity advertised is acceptable. It is important to note that SPECS refer to the acceptability of a product, not to the mean and standard deviation of the process that yields the product. Very often, a target value is used, which is in the middle of the upper and lower SPECS. It is also important to note that under current USDA-CCC procurement procedures for P.L. 480 Title II blended commodities that specifications are written for the manufacturing process (e.g. how much of the nutrient must be added during manufacture), not on the nutrient content of the final product. Process capability refers to measures that express how well the production process satisfies the SPECS. These measures take several forms but generally can be translated into the proportion of product that is unacceptable. There are two ways to ensure that this proportion is small: (1) the process mean should be close to the target value, and (2) the standard deviation of the process should be small. In many processes, the first is relatively easy to achieve by adjusting a component. For example, if the process mean for vitamin C in conventional CSB is 27 mg/100 g, the equipment should be adjusted to add approximately 50 percent more premix in order to achieve the desired level of 40 mg/100 g.

OCR for page 47
--> Reducing the standard deviation is often a more complex job, involving the identification and control of sources of process variation. The process capability index (Cp) compares the UCL and LCL to the SPECS. Specifically, Cp is the ratio of the difference between the upper and lower SPECs divided by the difference between the UCL and the LCL. Implicit in the use of this measure is the assumption that the process mean is either on target or can easily be moved there. If the Cp is 1.0, the two sets of limits coincide and the proportion of unacceptable product is the proportion falling above or below 3 S.D. from the mean (i.e., 100 percent–99.7 percent, or 0.3 percent). In some industries, a Cp of 1.33 or better is required to become a qualified supplier, whereas 2.00 is required to be a preferred or long-term supplier. Samples collected by SUSTAIN of production runs at four or five CSB manufacturing plants and one of two WSB manufacturing plants were analyzed for both reduced and oxidized forms of vitamin C. A commercial statistical software program was used to analyze this data on production samples. This program, in addition to calculating upper and lower control limits and a number of descriptive statistics, also calculates the Cp (Ranum and Chomé, 1997). For CSB the Cp values ranged from 0.26 to 0.90 when the SPECs were set at 29–55 mg/100 g for conventional fortification and 65–120 mg/100 g for the high level of vitamin C. The Cp's were 0.20 and 0.85 for the two WSB productions in the same plant. Although these SPECs are somewhat arbitrary, the resulting calculations indicate that improvements are needed. For conventional CSB with a mean vitamin C content of 27.0 mg/100 g, 55 percent of the product is below the lower SPEC and 4 percent is above. It is also important to note that in three of the seven runs examined, an outlier was discarded. For a sample size of approximately 50 samples per run, this amounts to 2 percent, which indicates that these processes are not in control. This lack of control will simultaneously affect the content of other vitamins and minerals added to the CSB since both the vitamin and the mineral premixes are metered in at very low quantities. Stability of Vitamin C During Transport and Storage The SUSTAIN report (Ranum and Chomé, 1997) notes that there were minimal losses of vitamin C in shipping, transport, and storage of CSB shipped to India and WSB to Haiti. The WSB (conventional vitamin C) loss was 13 percent (p < .01). However, the high vitamin C WSB shipped to Haiti and the conventional vitamin C CSB shipped to India had no losses (retention ~100 percent). Shipping, transport, and storage time for shipments to Haiti was nine months (see Box 4-2) and for India, five months. No data are available on transport and storage losses for shipments to Tanzania because of problems in uniformity encountered in the production of fortified CSB sent there. However, the time from production to distribution was seven months (see Box 4-1). The

OCR for page 47
--> results of the pilot study, therefore, indicate that vitamin C is stable under the conditions encountered in shipping, transport, and storage. Vitamin C Cooking Losses Limited numbers of samples were collected for vitamin C retention studies in prepared foods made with CSB and WSB. Haiti—WSB The most complete set of information on vitamin C retention during food preparation was provided for WSB foods in Haiti. The numbers of samples tested for the high and the conventional fortification levels were not the same (Table 5-1): for the high level, five gruel and five dumpling samples were collected; for the conventional level, three gruel and four dumpling samples. This is a very small sample for testing the hypothesis that vitamin C is retained in cooked foods. TABLE 5-1. Summary of Food Preparation Samples Collected in Selected Countries WSB, Haiti No. of Samples CSB, Tanzania No. of Samples High vitamin C   High vitamin C   gruel 5 gruel 7 dumplings 5 ugali 4 Conventional vitamin C   Conventional vitamin C   gruel 3 gruel 9 dumplings 4 ugali 1     fried cake 1   SOURCE: Ranum and Chomé, 1997. Means and standard deviations (S. D.) of vitamin C retention in WSB foods shipped to Haiti are fairly consistent (high C level: 32 ± 7 percent was retained in gruel and 33 ± 18 percent in dumplings; conventional C level: 27 ± 4 percent was retained in gruel and 18 ± 8 percent in dumplings, see Table 4-3). These results suggest that a smaller percentage of total vitamin C is retained at the conventional level of fortification. This is expected because vitamin C is less stable in dilute solutions than in more concentrated ones, particularly at the neutral pH of these foods (Hornig and Moser, 1981; Erdman and Klein, 1982). However, the overall average of ~27.5 percent retention in both gruel and dumplings is lower than expected or desirable. The average amount of vitamin

OCR for page 47
--> C in the cooked foods ranges from 2 to 6 mg per 100 g of gruel, and 3 to 12 mg per 100 g of dumplings. The intake of vitamin C from a serving of gruel (estimated as 150 g) would be greater than from a serving of dumplings (estimated at 40 g). Differences in retention between dumplings and gruel may be attributed to the fact that gruel absorbs all of the cooking water, including any vitamin C that is dissolved during the soaking process. When dumplings are cooked, some vitamin C may be leached into the cooking liquid, but would ultimately be consumed. In this case, the total amounts of vitamin C retained in the two types of cooked foods may be similar, although product analysis indicates different amounts present. Tanzania—CSB For Tanzania, the sampling was more limited than Haiti. There were seven gruel and four ugali samples for the high C level and nine gruel, one ugali, and one fried cake for the conventional C level. Complete preparation information is not available for all the samples because of a problem with the scales when data were collected. Recipes and cooking times were not as well documented in Tanzania as in Haiti. Data for CSB products from Tanzania are complicated by the inconsistent vitamin C content of the CSB produced by plant A. Because of this, CSB with the high fortification level of vitamin C was stratified into four groups (68, 93, 140, and 160 mg per 100 g of CSB). This further complicates interpretation of the data. Data are presented only for products prepared with 93 or 160 mg of product (see Table 4-2). CSB gruel in Tanzania had higher moisture content (90 percent) than WSB gruel in Haiti (80 percent). Thus, losses would be expected to be higher in the more dilute gruel. This was true for the conventional-C CSB gruel (initial content 22 ± 5 mg/100 g), where retention was below detectable amounts of vitamin C for almost all nine gruel samples. Therefore, the data for retention, which were estimated as 17–32 percent, are questionable. For the high-C CSB gruel, the three samples initially containing ~100 mg/100 g of CSB retained 48 percent of the vitamin C. In samples with an initial content of 177 mg/100 g, ~67 percent of the vitamin C was retained. On average, gruel contained 4–16 mg/100 g when made with any of the higher levels of vitamin C. Less than 1 mg/ 100 g remained in the conventional C gruel. This is in line with the theory that vitamin C is more stable in more concentrated solutions.

OCR for page 47
--> General Comments Sample handling after cooking may have been less than optimal. The samples were placed in 4-ounce containers, which were put into insulated containers with frozen ice packs and moved to frozen storage approximately eight hours later. The cooler and freezer temperatures were monitored. However, it is both possible and likely that some vitamin C was lost in the initial cooling phase, particularly in the lower vitamin C samples. Thus, the vitamin C content of some or all of the analyzed samples of prepared foods may actually have been higher than reported. This suggests the need for follow-up of changes in vitamin C content during the immediate postcooking period. Cooking procedures used in the field (in both Tanzania and Haiti) included soaking steps followed by heating. Cooking times were ~20–30 minutes, including soaking. Both CSB and WSB are processed to allow briefer cooking periods, and in situations of limited fuel this would be very important. Shorter heating periods could result in greater vitamin C retention because the primary cause of loss is heating. In theory, 30 g of the WSB or CSB fortified at the conventional level of 40 mg/100 g or the high level of 90 mg/100 g would provide 12 or 27 mg of ascorbic acid, respectively if there were no cooking losses. If ~30 percent retention is assumed for WSB gruel, the intake would be 3.6 or 11.4 mg. Similarly, for conventional-level CSB gruel, retention was estimated at 30 percent, so intake would be 3.6 mg. At the high levels (starting material >>90 mg/ 100 g), between 13.5 and 18 mg could be provided. If the intake of vitamin C was actually above 10 mg per day, this would be sufficient to prevent scurvy.