Dietary restrictions, including elimination of foods containing artificial flavors and colorings, "should not be universally applied to the treatment of childhood hyperactivity at this time," an expert panel concluded yesterday.

But the 13-member committee convened by the National Institutes of Health emphasized that an "initial trial" of dietary treatment "may be warranted" in patients whose families and physicians feel it may be helpful.

The group said that the controversial dietary approach should be started only after "full consideration of all therapeutic options," including psychotherapy, behavior modification, educational approaches and treatment with drugs.

The NIH committee, in a carefully worded draft statement and news conference, steered clear of endorsing any one approach and stressed that treatment must be tailored to the individual child.

The group consisted of doctors, other scientists, a lawyer and consumer representatives who heard two days of presentations on the value of dietary treatment for hyperactivity.

The group said the conference was stimulated by discrepancies between favorable case reports and "much of the available scientific evidence," including research studies that have found the dietary approach "only occasionally" effective.

"Clinical observations indicate frequent dramatic improvements in many hyperactive children during uncontrolled trials of defined diets. However, such dramatic improvements have not been established in the controlled trials presented during the course of this conference," the panel said.

California physician Ben Feingold, who championed the dietary approach in the early 1970s, said that as many as 50 percent of hyperactive children can be helped with careful management. The panel said, however, that the evidence thus far suggests that only a "small group" of hyperactive children may respond positively to the dietary changes and that there is "insufficient" evidence to identify these children.

The group noted that there seems to be no harm in trying the dietary approach, but said studies are needed of the "possible adverse effects" as well as other areas in which information is lacking, including the definition of hyperactivity.

The condition is generally characterized by increased motor activity, lack of attention and impulsiveness inappropriate for a child's age. Estimates on the prevalence of hyperactivity range from 1 percent to as high as 20 percent of the school-age population, with the problem far more common in boys.

Dietary treatment includes a variety of individualized menus, but the basic Feingold approach includes the elimination of foods with artificial flavors, coloring and many preservatives as well as foods naturally containing chemicals known as salicylates, including cucumbers, tomatoes, berries, apples, oranges and other fruits.

Members of the Feingold Association, a group advocating the dietary approach, expressed satisfaction yesterday that the NIH panel, while not endorsing the diet, did at least consider it an option.

But the panel's statement is not likely to put an end to the emotional debate, which includes charges by Feingold proponents that some researchers who have conducted studies with negative findings were biased by industry connections, and charges by scientists who accuse the restricted-diet side of conducting a crusade without proven benefits.

The NIH panel was chaired by Dr. Floyd Denny, a pediatrician at the University of North Carolina. It was sponsored by the National Institute of Allergy and Infectious Diseases and the National Institute of Child Health and Human Development.