The welfare of farm animals is strongly influenced by the man-made environment. Welfare problems also arise from reduced homeostatic capacities in animals. Feed additives, used to promote growth or to prevent diseases can alter the animals' self-regulating capacities thus affecting their welfare. The EU regulates the use of these additives within specified groups of Directive 70/524/EEC. Although these feed additives can be regarded as prescription-free veterinary drugs, critical remarks on their desired and adverse effects have received little attention.

A survey of the available literature shows that about one-third of licensed feed additives alter adrenal function in vitro. Reports of the adverse effects of anticoccidial additives in vivo suggest they can be classified under three headings: (i) substances with a very narrow safety margin (the difference between the permitted dose and the dose with adverse effects) and often irreversible effects on growth and feed conversion; (ii) substances with a narrow safety margin and largely reversible effects; (iii) substances with an adequate safety margin. The growth promoters (including antibiotic growth promoters) can - on the basis of their adverse effects - be classified into two groups: (i) substances with a very narrow safety margin; and (ii) substances with an adequate safety margin.

On the one hand, animal welfare considerations require use of disease-preventing additives, but on the other hand, they also demand discontinuation of current practices. Judicious use of additives can add to animal welfare. However, their unlimited use to obscure defects in husbandry is detrimental to animal welfare. A major obstacle to the judicious use of feed additives, is the lack of published, unbiased information on their efficacy and safety for farm animals.

Selective breeding and improved nutritional management over the past 20–30 years has resulted in dramatic improvements in growth efficiency for pigs and poultry, particularly lean tissue growth. However, this has been achieved using high-quality feed ingredients, such as wheat and soya that are also used for human consumption and more recently biofuels production. Ruminants on the other hand are less efficient, but are normally fed poorer quality ingredients that cannot be digested by human subjects, such as grass or silage. The challenges therefore are to: (i) maintain the current efficiency of growth of pigs and poultry, but using more ingredients not needed to feed the increasing human population or for the production of biofuels; (ii) improve the efficiency of growth in ruminants; (iii) at the same time produce animal products (meat, milk and eggs) of equal or improved quality. This review will describe the use of: (a) enzyme additives for animal feeds, to improve feed digestibility; (b) known growth promoting agents, such as growth hormone, β-agonists and anabolic steroids, currently banned in the European Union but used in other parts of the world; (c) recent transcriptomic studies into molecular mechanisms for improved growth efficiency via low residual feed intake. In doing so, the use of genetic manipulation in animals will also be discussed.

Avilamycin (AGP) and zinc oxide (ZnO) are both frequently included in the post-weaning piglet diet to enhance growth performance and prevent diarrhoea. This study investigated what effect removing these compounds from the post-weaning diet would have on the growth performance and faecal microbiota of weaned piglets. Fifty-two crossbred piglets (JSR Healthbred) were allocated, at weaning, to one of two dietary treatments on the basis of weight, litter origin and gender. The diets were (i) control (no supplemented ZnO or AGP); (ii) ZnO + AGP (supplemented with 3100 mg ZnO per kg food and 40 mg avilamycin per kg food). These diets were offered ad libitum for 20 days post weaning. Thereafter, the pigs received the same non-supplemented grower and finisher diets ad libitum. All piglets were individually weighed, and faecal samples were obtained from pre-selected piglets, at various time points throughout the trial period. Ten-fold serial dilutions of faecal material were cultured on specific media to enumerate aerobes, anaerobes, Lactobacillus spp. and Escherichia coli. ZnO + AGP supplementation enhanced weaned piglet average daily food intake (ADFI) (P < 0·001), average daily live-weight gain (ADG) (P < 0·001) and food conversion ratio (FCR) (P < 0·01) during the initial 20 days post weaning. Piglets previously supplemented with ZnO + AGP gained more weight per day during the non-supplemented grower phase (days 21 to 60) than their control counterparts (741·5 v. 672·5 g per pig per day) (P < 0·01). The bacteriological data showed that ZnO + AGP piglets had lower counts of anaerobic bacteria in their faeces than control piglets (P < 0·01). These findings indicate that dietary AGP + ZnO may enhance growth by reducing gastro-intestinal bacterial populations, and that their removal from the post-weaning diet will increase days to slaughter.

This trial investigated the effects of virginiamycin on sow and litter performance. Virginiamycin was added to the sow diet during the period of gestation and lactation at the levels of0, 40 and 60 mg/kg, throughout two breeding cycles. Virginiamycin did not significantly affect sow weight change during gestation or lactation, or litter size and weight; there were no significant effects on food consumption (P > 0·05). However, when added at a level of 40 mgl kg virginiamycin tended to produce a better condition of sows which resulted in a significant reduction of replacement rate (P < 0·05), even though the proportion of gilts culled in the treated and untreated groups was high. The most evident result was a significantly shorter weaning-to-conception interval for the virginiamycin group at the end of the second reproductive cycle (first period, 5·9 days less than the control group, 2·5 days less than the 60 mg/kg group; second period, -9·5 days v. control group, P < 0·05, and -7·1 days v. 60 mg/kg group, P < 0·05), allowing a numerical improvement in productivity. Piglet performance was not significantly improved by the use of virginiamycin even though there was an evident trend in favour of the 40 mg/kg group which presented a higher proportion of heavier piglets (P < 0·05) and a reduction of pre-weaning mortality. A significant quadratic regression was found for weaning-to-conception interval and numeric productivity, indicating that the dosage more efficient to improve these parameters was 32 mg/kg. The results obtained in the virginiamycin 60 group were generally inferior to those of the 40 mg/kg group and also to the control group and seem to demonstrate that the 60 mg/kg dosage was too high.

Responses of lambs to cimaterol (CIM), diethylstylbestrol (DES) and ovine growth hormone (GH) were examined using a mechanistic model of growing lamb metabolism. All three compounds increase growth of lean tissue (protein) and decrease fat gain, although the magnitudes of these responses vary. Our working hypothesis was that observed changes in nutrient partition between lean and fat gain were caused by alteration of rate constants for turn-over of muscle protein and fat. Individual experiments were simulated whilst varying values of the protein degradation constant (Kprolein) and Vmax for lipolysis (Kfat). Optimal parameter values were found by minimizing residual errors, calculated as the deviations of model predictions from experimental values for carcass protein and fat. Fitted values of Kfat and Kprotein (expressed as proportions of controls) for each simulation were: CIM (grazing), 1·20 (s.d. 0·05) and 0·86 (s.d. 0·025); CIM (pellet-fed), 1·11 (s.d. 0·115) and 0·87 (s.d. 0·032); DES, 1·33 (s.d. 0·111) and 0·94 (s.d. 0·024); GH, 1·77 (s.d. 0·139) and 0·97 (s.d. 0·025) respectively. These results demonstrate that different mechanisms may be responsible for the changes in carcass composition due to 3-adrenergic agonists, anabolic steroids and growth hormone. CIM probably exerts its effects via changes in protein and fat metabolism, whereas DES and GH appear to act mainly through changes in adipose tissue, with little or no effect on the rate constant for protein turn-over. Carcass composition is less sensitive to manipulation of adipose tissue metabolism than to changes in muscle protein metabolism.