Troubling stories were circulating about the so-called crack babies. They had small heads and were easily agitated and prone to tremors and bad muscle tone, according to reports, many of which were anecdotal. Worse, the babies seemed aloof and avoided eye contact. Some social workers predicted a lost generation - kids with a host of learning and emotional deficits who would overwhelm school systems and not be able to hold a job or form meaningful relationships. The "crack baby" image became symbolic of bad mothering, and some cocaine-using mothers had their babies taken from them or, in a few cases, were arrested.

Poverty is a more powerful influence on the outcome of inner-city children than gestational exposure to cocaine

At age 4, for instance, the average IQ of the cocaine-exposed children was 79.0 and the average IQ for the nonexposed children was 81.9.

As the children grew, the researchers did many evaluations to tease out environmental factors that could be affecting their development. On the upside, they found that children being raised in a nurturing home - measured by such factors as caregiver warmth and affection and language stimulation - were doing better than kids in a less nurturing home. On the downside, they found that 81 percent of the children had seen someone arrested; 74 percent had heard gunshots; 35 percent had seen someone get shot; and 19 percent had seen a dead body outside - and the kids were only 7 years old at the time. Those children who reported a high exposure to violence were likelier to show signs of depression and anxiety and to have lower self-esteem.

More recently, the team did MRI scans on the participants' brains. Some research has suggested that gestational cocaine exposure can affect brain development, especially the dopamine system, which in turn can harm cognitive function. An area of concern is "executive functioning," a set of skills involved in planning, problem-solving, and working memory.

The investigators found one brain area linked to attention skills that differed between exposed and nonexposed children, but they could not find any clinically significant effect on behavioral tests of attention skills.