【作者】 陈波；

【导师】 刘建新；

【作者基本信息】 浙江大学 ， 动物营养与饲料科学， 2011， 博士

【摘要】 Meta分析是针对相同研究内容的多个独立研究结果进行的数据再分析,是一种量化的文献综述。利用适当的统计方法对按照某种特定标准收集到的研究资料进行分析概括,以获得更加普遍的结论,其特点是增大样本含量,增加结论可信度,其结论比一般综述更加科学可靠,能最大限度地减少主观影响和最大可能地确保结论的科学性和真实性。Meta分析可以用于多种试验结果的综合,本研究主要以多水平试验以及对照试验为主进行研究和探讨。生物素作为一种动物生存所必须的水溶性B族维生素,可以通过自身肠道微生物合成。但随着动物生产性能提高,动物的日粮组成和饲养条件都有很大变化,动物靠自身肠道微生物合成以及饲料摄取已经不能够满足对其需要,动物极易出现生物素缺乏症。但是添加生物素的饲喂效果却存在很大差异,这些研究结果的不同可能是由于单个研究的特定试验环境和试验动物数目过少造成的。很有必要对这些发表的文献利用Meta分析方法进行研究以获得综合性的结论。本研究利用首先以DCAD(Dietary Cation-Anion Difference)对奶牛生产性能影响为例进行多水平试验研究的Meta分析,为生物素对奶牛生产性能的影响的Meta分析建立方法,但通过对生物素的相关文献检索后发现,两者在实验设计上有很大差异,因此将医学领域的对照试验Meta分析方法引入,进行生物素对奶牛生产性能的影响的Meta分析；同时还全面研究了生物素对奶牛生产性能,瘤胃发酵,血液生化指标,蹄质改善的影响。主要研究结果如下：1 Meta分析在多水平试验中的应用研究：DCAD对奶牛生产性能影响的Meta分析结果表明,Meta分析可以很好地应用于多水平试验分析,在去除研究间效应影响后,研究多水平添加对动物的不同影响情况,找出最佳添加量。在本研究中,校正后的DMI与DCAD二次线性相关,峰值在31.8meq/100g of DM.校正后的产奶量和4%FCM同样与DCAD二次线性相关,峰值在31.6 meq/100 gof DM和38.0 meq/100 g of DM.校正后的乳脂率,血液HC03-浓度,血液pH值与DCAD正一次线性相关。DCAD通过改变体内酸碱平衡状态而影响到泌乳早期奶牛的生产性能。2 Meta分析在对照试验中的应用研究：Meta分析可以较好的应用于对照试验,通过综合多个研究结果,获得相同添加量下,添加剂对动物的确切效果。在本研究中日粮添加生物素,DMI(P=0.012,n=178),产奶量(P=0.002,n=238),乳脂产量(P=0.041,n=222),以及乳蛋白产量(P=0.001,n=222)分别增加0.87,1.66,0.05和0.05 kg/d,而生物素添加对乳脂率作用很小(P=0.531,n=238)乳蛋白率作用也很小(P=0.546,n=238)。通过计算I2值计算发现所有变量的异质性均不显著(最大I2=12%,P=0.328；最小I2=0,P=0.575).利用Begg’stest进行发表偏倚检测表明所有参数均不存在发表偏倚。敏感性分析表明去掉任何一个单独研究都不会改变整体趋势,对于全部研究结果影响不大,即不存在敏感性研究。通过对9个研究的Meta分析表明泌乳牛日粮添加生物素有利于DMI,产奶量,乳脂,乳蛋白产量,但是对乳脂率乳蛋白率没有显著影响。3不同生物素添加水平对奶牛泌乳性能,血液生化指标,瘤胃发酵,蹄质改善的影响。试验分为3个处理(0,20 mg/d,40 mg/d)，生物素添加对干物质采食量没有明显效果(P=0.29)。各处理的产奶量均值分别为32.7,35.3,35.3kg/d。与对照组比较,试验组的产奶量有显著性提高(P<0.05)。但是,在乳脂含量产量、乳糖含量、总固体含量上没有差异(P>0.05)。与对照组相比,20和40 mg/d处理组乳中蛋白含量和产量有显著提高(P<0.05)。40 mg/d组乳中尿素氮含量高于对照组(P<0.05),但20 mg/d与40 mg/d差异不显著(P=0.11)。添加生物素提高了血液中生物素浓度水平,但是20和40 mg/d添加组差异不大。水分含量和蹄硬度含量没有受到生物素添加量影响。但是通过蹄样切片观察,添加生物素的蹄质比对照组更加紧密。试验结果表明,添加20和40 mg/d生物素对奶牛生产性能及体质改善效果相近。生物素的添加对组间VFA的产量没有影响说明瘤胃微生物数量组间没有差异。生物素添加提高了瘤胃pH值,但对瘤胃发酵的其它参数没有影响。生物素对奶牛生产性能提高的原因还需要进一步研究。综上所述,在动物营养研究领域,可以利用Meta分析对多水平试验以及对照试验进行数据二次分析而更加广泛、可靠的结论,本研究通过Meta分析以及动物试验的研究,认为生物素添加明显提高奶牛生产性能,同时可以提高其蹄质状况,在生产中生物素的最佳添加量为20mg/d。更多还原

【Abstract】 Meta-analysis is a statistical technique that summaries the results of several studies that address a set of related research hypotheses. Studies chosen for inclusion in a meta-analysis must be sufficiently similar in a number of characteristics in order to accurately combine their results. Meta-analysis can enlarge the numbers of sample, and get more credible conclusions than general systematic reviews.Biotin, as a water-soluble B-vitamin, is an essential nutrient for both rumen bacteria and the dairy cow. Biotin can be synthesized in the rumen for the requirement by low producing cows. As to the high-yielding cows, a high amount of concentrate was fed to meet the nutrition requirement. With the increasing proportion of dietary concentrate, biotin synthesis by ruminal microorganisms will be reduced. Therefore, it is necessary to add biotin in diet for dairy cattle. However the studies on effect of biotin did not get consistent conclusion. The difference between these results may be caused by the small number of animals used in one study and the specific conditions of the experiment. A thoroughly systematic and quantitative assessment of published findings is necessary to arrive at a comprehensive conclusion.The objective of this study was to introduce the application of meta-analysis in two kinds of experiments:the different dosage study with the effect of DC AD on dairy cattle as an example, and the comparative study with the effect of biotin on dairy cattle as an example, And then the animal experiment was used to evaluate the effect of different additional levels of biotin supplementation on milk production, blood biochemical parameters and quality of the hoof in high-yielding dairy cows. Results obtained are as below:1. Application of Meta-analysis in dose studies:it is indicated that Meta analysis can be used to adjust the study effect between studies, and get the optimal additional levels of feed additives. The database was established from 7 studies published between 1994 and 2010 that included 23 dietary treatments and 261 cows. Milk yield, 4% FCM and feed intake were highest when DCAD was 31.6,38.0 and 31.8 meq/ 100 g of dry matter (DM), respectively. Highest blood pH and maximum HCO3" concentrations were reached when DCAD was 45.0 and 46.3 meq/100 g of DM. Milk fat percent increased with the increasing value of DCAD. It is indicated that the DCAD affect the voluntary intake and milk performance, and improve the buffering ability of acid-base balance in the cows.2. Application of Meta-analysis in comparative studies. It is indicated that Meta analysis can be used to get an comprehensive conclusion by combination of servel studies. A total of 11 studies from 9 papers, with information on the milk production and composition data from a total number of 238 cows were extracted and analyzed using meta-analysis software in Stata. Over all the trials analyzed, after the supplementation of biotin, DMI (P= 0.012, n= 178), milk yield (P= 0.002, n= 238), fat yield (P= 0.041, n= 222), and protein yield (P= 0.001, n= 222) increased by 0.87, 1.66,0.05, and 0.05 kg/d, respectively, while there was little effect on the percentage of milk fat (P= 0.531, n= 238) or milk protein (P= 0.546, n= 238). The I2 statistic indicated that heterogeneity for all the variables was not significant (the highest I2= 12%, P= 0.328; the lowest I2= 0, P= 0.575). The result of the Begg’s test for publication bias was not significant for all of the variables. Sensitivity analysis revealed that there were no influential studies. The removal of any single study did not change the direction of effect in any analysis, and had little or no change in the overall effect size estimates. It can be concluded that the use of biotin supplements increases dry matter intake and milk yield in lactating dairy cows.3. Effect of biotin on milk performance and hoof quality The objective of this part was to determine the effects of supplemental biotin (0,20,40 mg/d) on lactation performance and hoof quality in dairy cows (n= 45). Treatments started at 40 d and lasted to 110 days in milk. Milk samples were measured weekly; blood samples were taken from coccygeal vein biweekly. Dry matter intake during lactation was not different across treatments (21.4 kg/d). Milk production increased with biotin supplementation (32.7,35.3, and 35.3 kg/d for 0,20, and 40 mg/d of supplemental biotin, respectively). Milk fat percentage and yield, lactose, total solid in milk and concentration of glucose, triglyceride, NEFA in plasma were not affected by biotin supplementation. Similar to 20 mg/d group, protein concentration and yield, and concentration of urea nitrogen in milk and biotin in plasma of 40 mg/d group, were higher than control group. Biotin supplementation did not have effect on moisture content and hardness of hoof in a short period, but made hoof more compact than control group. Supplemental biotin increased milk and protein yields, however, there was no significant difference in milk performance and hoof quality between supplemental levels of 20 and 40 mg biotin per day. Supplemental biotin has no effect on vitro rumen fermentation parameters. The mechanism of biotin on dairy cattle needs further investigations.In summary, Meta-analysis can be used in the research of animal science, including the dose studies and comparative studies, to get more comprehensive conclusions. In this study, the result of meta-analysis shows that biotin improved the milk performance. The experiment shows that additional level 20mg/d is the best for dairy cattle and biotin can reduce microcrack in the hoof of dairy cows.更多还原