The long-running debate over the merits of day care for children is getting a new jolt of evidence that may worry working parents.

Children who spend large amounts of time in child-care centers exhibit more minor behavior problems, such as aggression and disobedience, than other children, at least through sixth grade, according to a long-term study that followed 1,364 children from birth through age 12. The problems had surfaced intermittently in studies of the same children at younger ages, but some researchers had expected the behavior issues to disappear by age 11 or 12. Children who spent large amounts of time in other setups, such as nanny care or family child-care homes, weren’t affected.

The study, published in the March-April issue of the journal Child Development, includes some good news too: Children who experience high-quality child care  whether in child-care centers or family child-care homes, or with sitters or nannies  have better vocabulary skills through fifth grade than children who get lower-quality care. However, math and reading gains, seen at younger ages in children who had spent time in high-quality care setups, didn’t last past first grade. High-quality care is defined as care by an engaged, responsive adult or adults in a rich, nurturing setting.

The study comes at a time when many parents are choosing child-care centers over other kinds of care in the belief that they’re more reliable and offer educational benefits. The proportion of preschool children of employed mothers enrolled in child-care centers and preschools rose to 24% in 2002, the latest data available, from about 21% in the late 1990s, the Census Bureau says.

In any case, the researchers emphasized, parents have a far more powerful impact on children than any of the effects of day care. “The quality of parenting children receive is a far stronger and more consistent predictor of achievement and social functioning,” the researchers said.

The child-care study has tracked the same children at intervals since birth, evaluating their cognitive and social skills, behavior and academic achievement in the context of their child-care experiences. Participating families chose all kinds of care setups, allowing for comparisons. As preschoolers, the kids’ time in child care, defined as care by someone other than their mothers, ranged from zero to more than 40 hours a week.

The study was initiated in 1991 by the National Institute for Child Health and Human Development, in part to bring together on a single team warring camps of researchers. In smaller studies, some researchers had asserted that nonmaternal care of any kind, including nannies or family child-care homes, was potentially damaging. Others portrayed child care in group settings, such as child-care centers, as potentially positive, speeding cognitive and social development. Since its inception, the NICHD database, which has yielded several major sets of findings and dozens of smaller studies, has been seen as the gold standard in child-care research.

The 15 researchers who co-authored the latest paper are faculty members at major universities or are aligned with nonpartisan research organizations. Their individual viewpoints on nonmaternal child care have ranged widely, lending credibility to their consensus findings.

Both the good and bad child-care effects identified by the study wax and wane at various stages of children’s development. Some behavior problems identified at ages two, 4 and in kindergarten faded into insignificance in third grade only to reappear in sixth, for example. The behavior effects were reported by teachers in questionnaires and include such problems as hits others, is disobedient at school and argues a lot. While many children exhibit such behaviors, researchers found teachers reported more problem behaviors for children who had spent more of their preschool time in child-care centers. Researchers will revisit the children at age 15.

Nevertheless, in a shift, the NICHD researchers for the first time publicly agreed in the study that because so many children spend time in child-care centers, the behavior effects may have collective consequences across classrooms, communities and society at large, says Jay Belsky, the lead author and a professor of psychology at Birkbeck University of London. About four million U.S. children under five spend time in child-care centers, compared with about 690,000 in nanny care and 1.1 million in family child-care homes, where a caregiver tends several children in his or her home.

The study draws no conclusions about what those consequences may be, but it suggests that teachers may have to spend more time managing classroom behavior, or playground conduct may deteriorate.

There was no specific amount of time in care that suddenly gave rise to the problems. The behavior issues rose gradually, in line with increases in hours spent in nonmaternal care. The mechanism by which group care shapes behavior isn’t clear, but the study suggests it is probably linked to the way peers relate to and influence each other. The results are controlled for many variables.

For many parents, “there are just so few options,” says Glen Fuller, a Moraga, Calif., father of two. Both his children have attended a child-care center. “My wife was working, I was working and we had to do something,” he says. “It was just about finding the right program.” Told about the study results, he says his experience has been the opposite. His daughter, now six, benefited from child care, he says, emerging with social skills that were significantly advanced compared with those of her kindergarten classmates.

Parents have criticized the NICHD study in the past for defining child care as care by anyone other than a child’s mother. Thus routine care by fathers and grandparents was defined as ‘child care.” The latest findings address the problem by comparing care by relatives with care by nonrelatives.

Indeed, certain problems found in child-care kids in earlier stages of the research, such as poorer work habits, were limited exclusively to children in nonrelative care. However, by the time the children reached fifth and sixth grade, the differences in relative vs. nonrelative care had dwindled to insignificance. “It is center care that has unique and enduring impact of a seemingly adverse kind,” the study says.

Child-care providers say it’s important to keep the findings in perspective. The behavior effects are modest, involve a small number of children and are still within the normal range of conduct, says Jim Greenman, an author and speaker on child-care issues and a senior vice president at Bright Horizons Family Solutions, a Watertown, Mass., operator of child-care centers. Also, it’s extremely difficult to sort out links among the abundant factors that shape individual children, he says. “Effects will vary because children and families are all unique.”