Autism, an Epidemic?

In recent years, we’ve seen a considerable increase in autism prevalence worldwide. Some interpret this as a sign of environmental degradation, which may increase the frequency of pervasive developmental disorders (PDDs), and even go so far as to call it a “health catastrophe” and an “epidemic.” But what is the truth? This question is put to Eric Fombonne, a former Inserm researcher and current Chair of Research in Child and Adolescent Psychiatry at McGill University in Montreal, after a tenure at King’s College London. He is internationally recognized for his expertise in child psychiatry epidemiology, especially in autism.

With the rise in autism numbers, some suggest it’s an epidemic. But can we really call it that?

Not at all! All we know comes from prevalence studies, which indeed have shown increasing numbers over time. In the 1970s, autism cases were estimated at 4 to 5 per 10,000 people. However, back then, researchers didn’t include all pervasive developmental disorders, such as atypical or incomplete forms. The 1980s brought a major shift in understanding PDDs, expanding the concept of autism to include Asperger’s syndrome, atypical forms, and nonspecific PDDs between 1992 and 1994. The definition broadened, raising the diagnosis rate to about 20/10,000 by 1994.

But how did we arrive at a rate three times higher today?

It all started with our work at King’s College London. In 1998-1999, Suniti Chakrabarti and I conducted a study in Staffordshire (JAMA. 2001 Jun 27;285(24):3093-9) that was far more proactive than previous studies, which only counted already diagnosed or treated cases.

This time, we reached out widely, enlisting all frontline professionals in identifying suspected cases. Nurses, speech therapists, and even general practitioners received training and tools to spot signs of PDD, referring children with even a single diagnostic item to a specialized center where autism specialists completed the final diagnosis and patient profile. This led to an overall PDD prevalence of 62-63/10,000, with autism specifically at 16.8/10,000 and other PDDs at 45.8/10,000.

Hasn’t this prevalence tended to increase over time?

Not as far as we know. In an article to be published in Pediatrics in July 2006, we show that in Montreal, the prevalence is 65/10,000—0.6% of the population, or 1 in 165 children. But we can’t compare this to past rates. To determine if there’s an epidemic or significant prevalence increase, we’d need incidence studies repeated yearly, but such data isn’t available anywhere. While we can’t completely rule out a rise in incidence, there’s no closed hypothesis here, and certainly no cause for alarm.

What reasons might explain a potential increase in incidence, if there is one? The environment?

Indeed, that’s the main argument made by some researchers. Two environmental hypotheses have been proposed.

The first involves the MMR (measles-mumps-rubella) vaccine, a live attenuated vaccine without thiomersal or mercury. London gastroenterologist Andrew Wakefield stirred up concern by claiming to find a persistent viral infection from the vaccine’s measles strain in the intestines of young patients (15-24 months) with gastrointestinal symptoms and regressive neurological syndrome. He suggested these digestive issues altered intestinal permeability, allowing neurotoxic substances to enter the bloodstream and causing neurological damage. But no research team has been able to replicate his findings, and all epidemiological evidence refutes this hypothesis.

However, Wakefield still has followers in the UK, which is unfortunate. Due to his theory, measles vaccination coverage fell from 94% in 1998 to 82% in 2004, insufficient to protect the population. Measles is a real danger, not just in developing countries: in 1999-2000, outbreaks in the Netherlands and Ireland killed five children and hospitalized several others in critical condition. In 1990-1991, a measles epidemic in the U.S. killed 150 people. In each case, vaccination rates had dropped below the critical threshold of 90%.

The second hypothesis also involves vaccination...

Yes, the other hypothesis concerns mercury in the form of ethylmercury in thiomersal, used as a preservative in some non-live vaccines to prevent bacterial infections. Mercury is known for its neurotoxicity, and mercury poisoning mimics some autism symptoms. The alarm was raised by a study that simply totaled the ethylmercury doses children received between birth and two years if the vaccination schedule was strictly followed, concluding that this amount far exceeded WHO and EPA (U.S. Environmental Protection Agency) limits.

First, it’s essential to know that these standards protect the general population, representing not a toxicity threshold but a maximum protection level. The authors also failed to consider excretion: ethylmercury in young children is excreted within 14 days. Their calculations are flawed because, with sporadic exposures like vaccinations, this threshold is never exceeded. Finally, no correlation has been found between PDD occurrence and measured mercury levels in hair or blood, nor in PDD incidence trends with changing mercury exposure when thiomersal was removed from vaccines (Fombonne et al. Pediatrics, July 2006, in press).

Another hypothesis that doesn’t hold up, though it’s possible that certain environmental factors might play a role in autism’s onset. No doubt new hypotheses will emerge in the future…