The article by Venables and Raine in this issue of the Journal (1) is the latest report from the Mauritius Child Health Project, a remarkable ongoing three-generation cohort study (2). During the study, neuropsychological assessments were conducted on the middle, or “G2,” generation at multiple time points from age 3 to age 40. When those in this middle generation were children, malnourishment was still common in Mauritius. The article by Venables and Raine analyzes data from participants ranging from 3 to 23 years old in this middle generation.

The article raises three provocative issues. First, it attempts to delineate a pathway by which early childhood malnutrition in developing countries might affect adult cognition, behavior, or mental health. Although not unique in this effort—for instance, Galler and Barrett (3) and Stein et al. (4) have also put forward their ideas on the topic—the debate on causal pathways is usefully widened to include the outcome of schizotypal personality. Second, it raises the prospect of primary prevention of schizophrenia and related disorders; reducing childhood malnutrition now joins other potential candidates such as prenatal or postnatal micronutrient supplements and prevention of prenatal and postnatal infections (5). Third, the article highlights a central challenge for global mental health that tends to be overlooked. Recent success in reducing child mortality in developing countries has left hundreds of millions of child survivors in poor living conditions with neurodevelopmental delays and disabilities that receive little or no attention (6). The Mauritius Child Health Project and studies like it help us understand the long-term adverse consequences of childhood malnutrition and what can be done to prevent them. Looking beyond this particular report, the data collected on the three-generation Mauritius cohort on a range of childhood adversities, as well as on a broad-based early-intervention component, have the capacity to shed light on resilience as well as on poor outcomes.

In the article by Venables and Raine (1), childhood malnutrition was defined by anemia and stunting at 3 years old. Stunting generally reflects chronic malnutrition up to age 3, and malnutrition often continues after age 3. The authors examined whether IQ at age 11 mediated an association between childhood malnutrition and schizotypal personality as measured by scores on the Schizotypal Personality Questionnaire at age 23. Their results suggest that performance IQ but not verbal IQ could be such a mediator and that the mediation pertains only to the two dimensions of schizotypy (disorganized or interpersonal) that the authors had previously hypothesized to represent “neurodevelopmental” schizotypy.

Many legitimate questions can be raised about these results and their interpretation. Some readers will be puzzled by the lack of an overall relationship between childhood malnutrition and schizotypy in analyses that do not take IQ into account. When there is an association between exposure and mediator and between mediator and outcome, the association between exposure and outcome will vary according to the presence of other pathways between exposure and outcome, the strength of relationships, and other factors (7). These relationships will need clarification from additional analyses of data from this study. Other readers may question why performance IQ but not verbal IQ was related to schizotypy, as this has not been consistently found in previous studies. Still others may question whether what we have considered to be schizotypy may overlap with autism spectrum disorders (8). The participants and researchers were not blind to childhood malnutrition status during follow-ups, but it seems unlikely that this was an appreciable source of bias when considering the nature of the study as a whole (2). The authors have a rich array of previously collected and new data that could and should be used to address these questions in subsequent research.

It may seem obvious that childhood adversity will lead to poor cognitive development and increase the risk for adult mental disorders. Some readers may ask, “So what?” In fact, the evidence does not support such a simple assumption. The relationship between early adversity, cognitive development, and adult mental health is extremely complex, and the existing evidence poses many quandaries.

Consider two puzzles about IQ mediation that emerge from findings in high-income countries. First, a gradient of parental socioeconomic status at birth is strongly related to a gradient in child and adolescent IQ. A lower adolescent IQ has been associated with schizophrenia, although the degree to which it is a cause or consequence of early manifestations is still unclear (9). Most studies have not found a gradient of parental socioeconomic status at birth to be related to offspring risk of schizophrenia. Second, in the Dutch famine studies, periconceptional malnutrition was specifically related to a greater risk of neural tube defects, schizophrenia, and schizotypal personality disorder (10), but it was not related to IQ at age 18 (as evaluated in male participants) (11). Recent studies, however, including a large study from Norway (12), suggest that periconceptional supplements of folic acid (which is proven to prevent neural tube defects) may reduce the risk of neurodevelopmental delay and disorder in children. These periconceptional supplements may therefore reduce the risk of low childhood IQ and perhaps later schizophrenia. Thus, the current evidence does not support any simple overarching conclusion about IQ as a mediator between childhood adversity and schizophrenia.

Studies of low- and middle-income countries have found that childhood malnutrition in resource-poor settings often extends from conception well into childhood or beyond. Most reports suggest that chronic childhood malnutrition is related to poorer cognitive development (13), which may reflect the impact of cumulative exposure to malnutrition during a “sensitive period,” perhaps up to 3 years old (although a “sensitive period” has not yet been clearly established). We have much less understanding of the relationship between chronic childhood malnutrition and adolescent and adult mental health, although some evidence has been reported (3, 13). We know still less about the role of IQ as a mediator, which points to the need for this study.

Finally, for readers who are unfamiliar with this field, it is important to distinguish general malnutrition (studied in the Mauritius Child Health Project) from micronutrient deficiencies. Many children across the globe are exposed to prenatal iodine deficiencies, periconceptional folic acid deficiencies, and iron deficiencies as infants, and these deficiencies, as well as deficiencies in several other micronutrients, are related to adverse neurodevelopmental outcomes at birth or in childhood. Supplements that correct these deficiencies can improve maternal and child health even in the context of ongoing undernutrition, although this should not divert our attention from improving overall nutrition.

Like any single study, the study by Venables and Raine is not likely to be the final word on the link between malnutrition and mental health. Nonetheless, it is certainly an important contribution.