exercise) would disrupt an initial crucial event in acute inflammation and the accumulation of neutrophils. This would result in a diminished inflammatory response and a concomitant reduction in delayed onset muscle soreness and waste products. Generally, assessments are made before the exercise and at several intervals after the exercise. In many studies the effects are positive, but in other studies negative results occurred. The conflicting results have stimulated at least three reviews of that literature. As for many of the questions, the assessment tools that we have may simply not yet be sufficiently developed. Measurement technology has, as already mentioned, been one of the most limiting problems in conducting massage therapy research. The question of whether mechanical stimulation is as effective as stimulation provided by a massage therapist has never been addressed by directly comparing the two. However, an extensive literature on vibrator stimulation (most of which comes from Sweden) suggests significant therapeutic effects at least for pain reduction (Lundeberg, 1984; Lundeberg, Abrahamsson, Bondesson, & Haker, 1987, 1988; Ottoson, Ekblom, & Hansson, 1981). In these studies, vibration, typically at 100 Hz, is applied to various points in different locations; in one study to the facial region affected by dental pain and in another study to different areas of the body in a patient suffering chronic musculoskeletal pain. Typically patients have reported a pain intensity reduction of 75%-100%. The greatest pain reduction occurred either in the area of pain, the affected muscle or tendon, the antagonistic muscle, or a trigger point outside the painful area. In most patients the greatest pain-reduction effect occurred when the vibratory stimulation was applied with moderate pressure. To obtain a maximum duration of pain relief the stimulation had to be applied for about 25-45 minutes. After 12 months of treatment most patients reported a greater than 50% reduction in analgesic drug intake, and in one study the vibratory stimulation was a more efficient pain suppressor than aspirin. Double-blind studies where the vibratory stimulator was compared with a "placebo unit" (the vibrator turned on to make the sound but not vibrating) also revealed significant vibrator effects. A more direct assessment of this question is needed in which vibrator therapy is compared with manual therapy using the same participants with the same condition. Still another important question is whether self-massage can be as effective as being massaged by another individual. The third question on underlying mechanisms has rarely been addressed. The most common theory that is based on anecdotal data and a very mixed empirical literature is that of massage increasing circulation or blood flow. Much of the literature suggesting that massage enhances circulation is an old literature that has been reviewed by Wakim (Wakim, Martin, Terrier, Elkins, & Krusen, 1949). As early as 1900 one author reported that after massage cutaneous temperature increased three degrees. A subsequent study demonstrated an increase in the diameter and permeability of the capillaries following mechanical stimulation in frogs and mammals. In a later study investigators measured skeletal muscle blood flow before, during, and after different forms of massage using a more sophisticated method for determination of blood flow, called the Xenon washout method. During rigorous massage blood flow increased comparable with exercise hyperemia (Hovind & Nielsen, 1974). In contrast to the earlier studies, a very recent study (Shoemaker, Tidus, & Mader, 1997) reported a failure of manual massage to alter blood flow as measured by Doppler Ultrasound. The authors noted that the mean blood velocity and blood flows for the brachial and femoral arteries respectively were not altered by any of the massage treatments whether they were administered mild 1278 December 1998 • American Psychologist or deep treatments in either the forearm or the quadricep muscle groups. Mild voluntary hand grip and knee extension contractions, in contrast, resulted in peak blood flow for brachial and femoral arteries, respectively, which were significantly elevated from rest. Shoemaker et al. concluded that light exercise was more beneficial than massage in increasing blood flow. Although the Doppler Ultrasound methodology is the most sophisticated way to currently measure blood flow, the study suffered from several methodological problems, including the use of relatively little pressure (light Swedish massage was used) and a very small sample size (10 participants). Another controversial literature involves sports massage. Several studies investigating the impact of massage on the reduction of delayed onset muscle soreness have concluded that massage was ineffective. However, in these studies massage was administered either immediately after exercise or 24-48 hours after exercise. In contrast, data from the Netherlands suggest that massage should be administered between one and three hours after the termination of strenuous exercise. Unlike the previous studies, Rodenburg, Steenbeek, Schiereck, and Bar (1994) found significant effects from administering the massage after exercise. They noted that the combination of a warm-up, stretching, and