|The ruling said:
The district court held that the composition claims were inherently misleading because 'they imply a compositional difference between those products that are produced with rb[ST] and those that are not,' in contravention of the FDA's finding that there is no measurable compositional difference between the two. This conclusion is belied by the record, however, which shows that, contrary to the district court's assertion, a compositional difference does exist between milk from untreated cows and conventional milk ("conventional milk," as used throughout this opinion, refers to milk from cows treated with rbST). As detailed by the amici parties seeking to strike down the Rule, the use of rbST in milk production has been shown to elevate the levels of insulin-like growth factor 1 (IGF-1), a naturally-occurring hormone that in high levels is linked to several types of cancers, among other things. The amici also point to certain studies indicating that rbST use induces an unnatural period of milk production during a cow's "negative energy phase." According to these studies, milk produced during this stage is considered to be low quality due to its increased fat content and its decreased level of proteins. The amici further note that milk from treated cows contains higher somatic cell counts, which makes the milk turn sour more quickly and is another indicator of poor milk quality. This evidence precludes us from agreeing with the district court's conclusion that there is no compositional difference between the two types of milk. In addition, and more salient to the regulation of composition claims like "rbST free," the failure to discover rbST in conventional milk is not necessarily because the artificial hormone is absent in such milk, but rather because scientists have been unable to perfect a test to detect it. [emphasis added]
Here's what Hansen said to VMAC on March 31, 1993, explaining how Monsanto only gave the FDA some (but not all) of its data in order to hide the somatic cell count issue, and how they bullied scientists who tried to expose the truth:
Second, the Committee is not being allowed to review all available data on incidence of mastitis, something we think seriously undermines any conclusions it may reach. At the December, 1990 National Institutes of Health Technical Assessment Meeting on rbST we voiced our concern on the issue of the rbGH-mastitis-antibiotic connection (Hansen, 1990). We also made public copies of a letter sent by the FDA to Monsanto which, in part, outlined serious health impacts for rbGH-treated cows (Lehmann, 1988). With respect to mastitis, the letter stated: "[d]ata presented indicate that there is an increase in mastitis at the levels at which you wish to market bovine somatotropin" (Lehmann, 1988: pg. 6). Further, "[y]ou have not established a margin of safety, nor have you established a no effect level for some of the parameters in your submission. Based on available data, this is particularly true of major clinical entities such as mastitis and reproduction" (Lehmann, 1988: pg. 7). Since this letter sharply contrasted with the then published studies on animal health (such as the OTA report) which claimed little or no adverse health effects, we called on the companies to release all their animal health data for independent scientists to review.
In fact, Monsanto did release some data for independent review. Monsanto conducted a large-scale multi-center trial, involving 8 separate experiments in 4 U.S. states and 4 European countries and some 620 cows. (It is possible that 2 (or 3) of these studies are among the 8 (or 7) discussed by FDA today; however, the other six are from different locations.) The experiments all shared a common randomized controlled design so that the animal health data could potentially be pooled (or lumped together) for a more powerful statistical test. In addition to data on the incidence of mastitis, data on somatic cell counts (SCC)--which are indicative of subclinical cases of mastitis and are basically a measure of dispersed puss cells--were also gathered. Monsanto contractees published their analysis of the data, including SCC data, in 1990 (Peel et al., 1990).
In late 1989 Monsanto sent all the raw SCC data to independent researchers in England who had written a previous paper on the subject for England's Veterinary Products Committee (Millstone et al., 1989). This represented a truly independent look at all the SCC data from the large multi-center trial. Their reanalysis (Brunner et al., 199?) showed that the SCC data were worse than what the previously published data revealed (Peel et al., 1989). Although the paper was accepted on scientific grounds for publication in a British journal, Monsanto has refused to allow it to be published, claiming that it is confidential information and that it would impair the ability of Monsanto scientists to publish data on these trials (see series of letters). The authors have given me permission to release their study and the letters to this Committee.
The data from these trials raises disturbing implications. First, Millstone and Brunner found that Monsanto consultants had not published all the relevant data. The experiment began 7 weeks after calves were born. The Monsanto-sanctioned article (Peel et al., 1989) had used the last 2 weeks before the experiment began as their baseline for SCC, and had included data from 28 weeks of treatment with rbGH. However, the raw data sent to Dr. Millstone consisted of data for all 7 weeks before the experiment began and for 43 weeks of treatment with rbGH. Dr. Millstone and Brunner used all the data in their analysis. Since Millstone et al. found that the data were highly positively skewed, they did a log-transformation which greatly reduced the skewness. Using all the data and log-transforming it allowed for greater statistical power in their analysis.
Briefly, Brunner et al. found: i) at 7 of the 8 sites, use of rbGH increased SCC; the effect was statistically significant at 3 sites; ii) the pooled analysis showed that SCC counts were, on average, 19% higher in rbGH-treated cows compared to controls, a highly statistically significant effect (95% confidence interval [C.I.]: 5.8 - 33.9, p = .004); iii) when you control for the covariates (baseline SCC, lactational age and trial site), you get the more accurate estimate for the effect of rbGH of 22.3% (95% C.I.: 12.6 - 32.9, p < .001) over the whole lactation; iv) if we just include the data during the later part of the lactation that were not included in the Monsanto-sponsored publication, i.e. weeks 28-43, the increase in SCCs due to rbGH use is 44.6% (95% C.I.: 27.1 - 64.5, p < .001).
The Brunner et al. paper clearly demonstrates that previous analyses of SCC data from the multi-center trial had been presented in a way that downplayed the supposed effect of rbGH on somatic cell counts.
Data from a controversial Monsanto-sponsored study done in Vermont, published in December, 1992 (Pell et al., 1992), are also disturbing. This study attracted much attention in Vermont because in October, 1991 the University, under orders from Monsanto, refused to release the animal health data to Vermont's Senate and House Committees on Agriculture. In a December, 1991 letter from the FDA to a Congressional committee, FDA revealed that over 40% of the rbGH-treated cows had to be treated for mastitis, compared to less than 10% of the control cows (9 of 21 rbGH-treated cows vs. 2 of 21 control cows (Holcombe, 1991). In 1992, Monsanto refused to release the data to the General Accounting Office (GAO).
The results, published in December, 1992, show that during the experiment: i) four times as many rbGH-treated cows (8 vs 2) had to be treated with antibiotics for mastitis compared to untreated cows, ii) more than seven times as many cases of mastitis occurred in rbGH-treated cows compared to untreated cows (29 vs. 4), iii) the average length of treatment for a case of mastitis was almost six times longer in the rbGH-treated cows compared to untreated cows (8.9 days vs. 1.5 days), and iv) more than seven times as much milk, on average, had to be discarded due to mastitis in rbGH-treated cows compared to untreated cows (73 kg vs. 10 kg).
Although the Vermont trial only included 46 cows, the facts that the experiment was completed in 1988, yet not published for four years; that Monsanto refused to give the data to both the Vermont legislature and the GAO, the official watchdog agency for the Congress; that the mastitis related results were quite disturbing when finally revealed; and that this study was not made available to the Committee; lends credence to the view that the company and the agency are publicly trying to downplay the size and significance of the mastitis problem.
The behavior of Monsanto, vis-a-vis the Vermont study as well as the UK reanalysis of their SCC data, is quite troubling. Clearly, this issue will not be fully resolved until all the mastitis data, both clinical and subclinical, are made publicly available for independent scientific analysis. This committee should demand to see all the mastitis data that FDA has before it renders an opinion on a question as important as that FDA has put before it today."
A few months later, on May 6, 1993, he testified before VMAC again, saying the following about
RbGH Use Causes Milk to Sour More Quickly
We also expect organoleptic changes (changes in taste, smell, texture or color) in milk from rbGH-treated cows. These changes relate to increased somatic cell counts (SCCs)--which are indicative of subclinical cases of mastitis and are basically a measure of dispersed pus cells. Milk with a high somatic cell count caused by mastitis turns sour more quickly, and is considered poorer quality milk, than milk with a low somatic cell count. Indeed, states have established a standard for SCCs of 1 million cells/ml, which is dropping to 750,000 in July, 1993. The European Community and the Milk Marketing Board (MMB) have recommended guidelines which call for a reduction in milk SCCs (CEC, 1990, MMB 1990) and which divide milk into lower, middle and upper SCC bands (0 - 400,000 cells/ml, between 400,000 and 1 million cells/ml, and > 1 million cells/ml, respectively); the premium milk quality standard refers to milk with less than 400,000 cells/ml.
Milk from rbGH-treated cows has higher somatic cell counts, which means that it contains more pus and bacteria, than milk from untreated cows. Monsanto conducted a large-scale multi-center trial, involving 8 separate experiments in 4 U.S. states and 4 European countries and some 620 cows. In addition to data on the incidence of mastitis, data on somatic cell counts were also gathered. The data show a greater than 50% increase in the percentage of cows giving milk whose peak SCC exceeds one million in any given week, in rbGH-treated cows compared to controls (37% vs. 24%, respectively). The data show a greater than 80% increase in cows whose peak SCC exceeds one million for at least two successive weeks (10.9% vs. 6%, respectively) (Brunner et al., 1991). Thus, rbGH use causes a statistically significant increase in SCCs in milk, thereby decreasing milk quality and shortening the shelf life.
Conversely, rbGH-treated cows are less likely to produce very good quality milk (SCC below 400,000). Data from the Monsanto trials show that only 27.9% of bGH-treated cows had a maximum SCC of under 400,000 in a one-week period, compared to 46.6% of control cows.
Consumers may want to avoid pusy milk which goes sour more quickly. However, since milk from rbGH-treated cows cannot be visually distinguished from milk from untreated cows, the lack of a label, we believe, would constitute consumer deception.
That same day, he also said this:
RbGH Use Affects Nutritional Value
Second, milk from rbGH-treated cows should be labeled because rbGH use can affect its nutritional value. In all dairy cows, there is a period of time at the start of the lactational cycle during which the cow is consuming less energy, in the form of food, than it is losing, in the form of milk. This condition is called negative energy balance, and it often lasts for the first 8 to 10 weeks of lactation. While cows are in negative energy balance, they tend to produce milk of lower quality (i.e. increased fats and decreased protein) than when they are in positive energy balance, or during the rest of the lactation. This makes perfect sense; if the cow cannot eat enough food to replace the energy she is losing to milk production, her milk will not be the best quality.
Use of rbGH invariably begins at some point after the cow has come out of negative energy balance, and induces a second such period. RbGH simulates (or mimics) the early stages of lactation. Thus, cows treated with rbGH go into negative energy balance for an additional period of time that frequently lasts for at least six weeks. Studies done during the early stages of rbGH use have found that rbGH can increase fat content and decrease casein proteins (which are crucial for cheese manufacture) (see references in Juskevich and Guyer, 1990). The studies that the industry relies on to show that there are no changes in the milk (see references in Juskevich and Guyer, 1990) invariably look at the entire lactational cycle after rbGH treatment, thereby lumping periods when rbGH-treated cows are in negative energy balance with those when they are in positive energy balance. Lumping of data like this hides the fact that the milk quality declines when the cows are in negative energy balance. Since rbGH treatment rarely begins before the tenth week of the lactational cycle, rbGH-treated cows have two periods of negative energy balance during each lactation, compared to one period for control cows. Thus, rbGH use increases the amount of time during which a cow gives milk of lower quality. As the number of cows treated with rbGH increases, so will the amount of milk of poorer quality. But consumers will not be able to detect such differences simply by looking at the milk. We think this constitutes deception of the consumer.