One out of every 100 live births in the United States is affected by fetal alcohol spectrum disorders - a range of physical, neurocognitive, emotional, and behavioral impairments caused by prenatal exposure to alcohol. Prenatal alcohol exposure has been associated with negative outcomes in diverse areas of functioning throughout the life span. Studies of negative outcomes associated with prenatal alcohol exposure, however, have largely focused on neurocognitive domains. Prospective studies of long-term psychosocial outcomes (particularly internalizing problems) are lacking. A few prospective studies on long-term psychosocial outcomes of prenatal alcohol exposure have focused on externalizing problems such as conduct problems, aggression and incarceration. Emerging literature indicates that prenatal alcohol exposure is associated with internalizing problems such as depression. However, it is not clear whether prenatal alcohol exposure effects on internalizing problems are independent from effects on externalizing problems and neurocognitive deficits. To better characterize prenatal alcohol exposure effects on internalizing problems, it is crucial to consider externalizing problems and neurocognitive deficits in a same model, because the effect of prenatal alcohol exposure on internalizing problems may be largely explained by its well-established effects on externalizing problems and neurocognitive functioning.

Long-term effects of prenatal alcohol exposure may differ depending on the degree of exposure to adverse environments later in life. First, individuals prenatally exposed to alcohol tend to initiate alcohol use at an early age and develop alcohol problems. Given that early onset of alcohol use has been shown as a risk factor for adverse mental health outcomes later in life, early initiation of alcohol use may exacerbate the effects of prenatal alcohol exposure on long-term psychosocial outcomes (including alcohol abuse and problems).

Second, disadvantages of social environments including family, neighborhood and school have been shown to affect mental health. Although a few studies have characterized adverse familial environments, neighborhood and school environmental effects rarely have been explored. Although the larger social environmental factors may have smaller effect sizes than familial environmental factors, the neighborhood and school disadvantages may exacerbate the effects of prenatal alcohol exposure on long-term psychosocial outcomes independent of familial disadvantages.

Thus, this proposed research has three specific aims:

(1) To examine the effect of prenatal alcohol exposure on psychosocial outcomes (both externalizing and internalizing problems) in adolescence, after controlling for neurocognitive functioning;

(2) To examine a moderating effect of alcohol use in childhood;

(3)To examine a moderating effect of adverse school and neighborhood environments, after controlling for familial disadvantages

In order to achieve the above study aims, I will estimate three structural equation models, each of which is designed to address each of the above study aims. The first model (Model 1 as shown below) will include a manifest variable of prenatal alcohol exposure. The model also will include two latent factors of internalizing problems and externalizing problems in adolescence, each of which will be estimated using manifest variables of diverse mental health measured available in ALSPAC data. The internalizing and externalizing problem factors will be allowed to covary. Thus, an estimate of the path from prenatal alcohol exposure on the each problem factor will represent an unique effect of prenatal alcohol exposure on one problem factor after accounting for its effect on the other problem factor.