D ogs do more than shed on the rug, bark at prowlers and provide reliable companionship. Sometimes the presence of a canine comrade acts as a natural sedative for its human master, lowering blood pressure and other bodily responses to stress while fostering improved performance on a difficult task, according to results from a new study of women and their dogs. Having a human friend close at hand while attempting the same task heightens the body's stress reactions and undermines performance, assert psychologists Karen M. Allen and James J. Blascovich, both of the State University of New York at Buffalo. Dogs take a bite out of stress because they provide unconditional support without scrutinizing their owner's sometimes frantic attempts to meet a challenge, Allen and Blascovich argue. A two-legged friend, harboring the human propensity to evaluate the success or failure of others, tends to stoke anxiety and embarrassment in the same situation. The new findings reported in the October JOURNAL OF PERSONALITY AND SOCIAL PSYCHOLOGY add to other indications that pets often help to buffer their owners from stress and illness. For example, in a 1980 study of 96 people with heart disease released after treatment at a coronary-care unit, psychiatrist Erika Friedmann of the University of Pennsylvania in Philadelphia and her co-workers identified a higher survival rate one year after hospital discharge among pet owners, even after accounting for individual differences in the extent of heart damage and other medical problems. In fact, having a pet at home proved a stronger predictor of survival than having a spouse or extensive family support, Friedmann's team found. In another investigation, 345 elderly pet owners cited proportionately fewer physician visits over one year than did 593 same-aged counterparts with no pets, psychologist Judith M. Siegel of the University of California, Los Angeles, reported in the June 1990 JOURNAL OF PERSONALITY AND SOCIAL PSYCHOLOGY. Pet owners told Siegel that their animals offered substantial comfort in times of stress. However, only dog owners showed no tendency to visit physicians more often following stressful experiences, such as the death of a close friend or spouse. Dogs, more than other pets, provided their owners with companionship and an object of attachment, Siegel concludes. Other studies have charted blood presA llen and Blascovich probed the calming influence of canines among 45 women faced with a standard psychology-lab challenge performing mental arithmetic. All the women owned dogs and described themselves as dog lovers. In an initial laboratory session, the psychologists measured blood pressure, pulse rate and electrical skin conductance as each participant sat with a female researcher and rapidly counted backwards from a four-digit number by threes for two minutes. After a break, they repeated the exercise, counting backwards from a new number by sevens. The mental arithmetic produced consistent jumps in the physical measures of stress. Two weeks later, a more difficult arithmetic challenge-counting backwards by 13s and 17s took place at each participant's home. The experimenters randomly assigned women to perform the tasks in the company of only the researcher or with the addition of either their dog or their closest female friend. With Fido at foot, physical stress markers during mental calculations showed little change from their levels during a pre-test rest period the same day. With only an experimenter sitting nearby, stress measures displayed moderate increases, and the presence of a female friend dramatically boosted the same measures. In addition, participants accompanied by a friend performed mental arithmetic much more rapidly but less accurately than women in the other two groups and suffered a marked drop in accuracy from the previous laboratory session. The women apparently perceived their dogs as less threatening than friends or Researchers sniff out the calming effects of canines