Cause of Autism: Global Research Challenges

In a world increasingly shaped by data, it's easy to mistake correlation for causation. A recent systematic review and meta-analysis by Wenrui Ye and colleagues published in The Lancet Diabetes & Endocrinology has made headlines in the New York Times, Reuters, and other established news outlets. The study reported findings from 202 observational studies investigating over 56 million pregnancies, identifying an association between maternal diabetes and neurodevelopmental outcomes in children, including autism spectrum disorder (ASD) and attention-deficit hyperactivity disorder (ADHD) (Ye et al., 2025).

But while the scale of this research is impressive, we still don't know what causes autism.

Ye et al. indicate that prior studies have shown patterns, not answers. Their work adds to a growing list of high-quality studies reporting links between various prenatal, perinatal, and environmental factors and autism diagnoses. These include maternal infections during pregnancy, advanced parental age, low birth weight, air pollution exposure, and now, maternal metabolic conditions like diabetes (Gardener et al., 2011; Becerra et al., 2013). Yet none of these factors can definitively explain autism. They are risk indicators, not causes of autism.

A 2017 study published in Nature Genetics identified dozens of genetic variants associated with autism, suggesting that in some cases, inherited or spontaneous genetic mutations may be a major contributor (Grove et al., 2019). However, even genetic research has only scratched the surface. Most individuals with autism don't have a clear genetic explanation. Moreover, epigenetics's interaction between genes and the environment remains unknown.

Another large study, published in Pediatrics in 2014, examined parental obesity and the risk of autism spectrum disorder and reported associations that an observational design cannot establish as causal (Surén et al., 2014). Observational studies, while informative, cannot control for all possible confounding variables.

Autism diagnosis is further complicated by geographic and cultural variables. While autism research is growing in the Global North, much less is known about the condition's prevalence and presentation in Africa. In Nigeria, for example, autism is often underreported, misdiagnosed, or misunderstood. A study by Lagunju, Bella-Awusah, and Omigbodun (2014) found that despite a rising number of clinical diagnoses of autistic disorder in a tertiary health center in southwestern Nigeria, many children remained undiagnosed until late childhood. Cultural beliefs, stigma, and lack of trained personnel were identified as significant barriers to early diagnosis and intervention.

Moreover, diagnostic tools commonly used in high-income countries are not always validated and culturally modified for African populations, complicating efforts to estimate prevalence accurately. According to the study, many of the children diagnosed with autism also presented with epilepsy, compounding the challenges of care (Lagunju et al., 2014).

This lack of data and infrastructure makes it difficult to determine how autism manifests across diverse African communities, including prevalence. Without educational support and public health systems, planning becomes nearly impossible, and families are left to navigate autism alone.

This is the uncomfortable but necessary truth: despite decades of research and millions of data, the exact causes of autism remain unknown. As autism diagnoses continue to rise globally, with some countries reporting a doubling of prevalence in just a decade we still have a lot of work to do to understand the condition (Maenner et al., 2023).

We need to identify causal mechanisms, we need more than data, we need coordinated, innovative, and interdisciplinary research from labs worldwide. This means investing in longitudinal studies, experimental research, and cross-cultural comparisons. It means prioritizing not just the quantity of studies, but their diversity and methodological rigor.

Autism is not a condition with a single cause. It is a spectrum, and to understand it, the world must act with the same urgency and unity that we've shown in facing other global challenges. Today, we have identified multiple correlations but no causal mechanisms for autism. Moving forward, international research collaboration and methodological diversity represent our best approach for transforming associations into actionable scientific understanding.