By bobb |
doctor? in lab. coat with stethoscope to young girl's chest

Caitlin E. Cox

What’s driving the increased risk of diabetes, dyslipidemia, and CVD in autistic people—especially children—isn’t clear-cut.

People with autism are at increased risk for developing cardiometabolic disease, including diabetes, dyslipidemia, and CVD, a meta-analysis confirms. The associations are particularly strong for children, suggesting a need to address risk factors at an early age.

Autism, with a global prevalence of 23 per 1,000 people, “is associated with multiple medical, neurologic, and psychiatric comorbidities,” particularly obesity, that “often exacerbate disparities in the quality of life and life expectancy” of affected individuals, Chathurika S. Dhanasekara, MD, PhD (Texas Tech University Health Sciences Center, Lubbock), and colleagues note in their paper, recently published in JAMA Pediatrics.

Senior author Chanaka N. Kahathuduwa, MD, PhD (Texas Tech University Health Sciences Center), said this work was inspired by a prior meta-analysis they’d done on autism and obesity in children. Their latest report contains data from 34 studies, said Kahathuduwa, but ”the problem is we can only do observational studies—it’s very difficult to prove causality in associations like this.”

There are numerous possible drivers for the associations—from lifestyle to genetics and shared origins—that could explain the connection, but none have yet been pinned down. “I wish I could give you an answer” on mechanism, he commented to TCTMD.

Regardless of the exact cause, “I think we have sufficient evidence [that] beyond a reasonable doubt, there is an association,” Kahathuduwa said, such that clinicians caring for patients with autism should be asking about risk factors for cardiometabolic disease and guideline-writing committees should take note.

I think we have sufficient evidence [that] beyond a reasonable doubt, there is an association. Chanaka N. Kahathuduwa

Carissa M. Baker-Smith, MD (Nemours Children's Hospital, Wilmington, DE), a member of the American College of Cardiology’s Adult Congenital and Pediatric Cardiology Leadership Council, told TCTMD that as a preventive cardiologist she isn’t surprised by the findings.

“I can’t speak for all cases of children with autism,” she stressed, but in her own practice she certainly sees the influence of challenges like strict “dietary preferences and some limitations of physical activity [as well] as social concerns that can exist as stressors in their lives.” Additionally, some antipsychotic medications used in autism can cause appetite changes that lead to weight gain.

That said, “any child who consumes a diet that’s restrictive and in many cases very carbohydrate rich is going to be at risk for some of the things that the authors saw,” Baker-Smith explained. Hyperlipidemia and obesity are a concern, and once a patient with autism develops diabetes it can be hard for them to take their needed medications and make lifestyle changes.

Young patients whose risk factors get detected at their annual checkups, irrespective of the cause, are often referred by their pediatricians to specialists like herself, she noted.

“What I’ve been advocating for all pediatricians to do is, one, pay attention to the growth curve for every patient. If you see that the child’s growth curve percentiles are accelerating across two lines, interventions need to be put in place,” Baker-Smith advised. That’s because even if it’s not apparent yet at that visit, once body mass index begins to rise at that pace, the patient will begin to show changes like higher triglycerides, lower HDL cholesterol, nonalcoholic fatty liver disease, and elevated values for blood pressure and HbA1c within the next 6 months or few years.

Much of the risk for cardiometabolic disease “is driven by obesity, . . . and so if you can address the obesity, sometimes you can prevent some of these other things,” she said, adding, “These aren’t isolated conditions—they’re linked in many, many cases. I would say in probably most cases.”

Most importantly, “we need to think outside the box and not be afraid to have the conversations about what this really means for the child long term,” said Baker-Smith.

Dyslipidemia Risk 69% Higher

The researchers identified 34 studies with a total of around 8 million patients, including 276,173 (3.4%) with an autism diagnosis. Mean age was 31.2 years, and 47% of participants were female. Diabetes was an endpoint in 20 studies (type 1 in six studies and type 2 in three studies), while seven studies looked at dyslipidemia and three at atherosclerotic heart disease. Hypertension was evaluated in 12 studies and stroke in four.

Individuals diagnosed with autism—at any point, under the assumption that autism is exposure that’s present since birth—had greater risks of diabetes (both overall and types 1 and 2), dyslipidemia, and heart disease, but not of hypertension or stroke.

Risks With vs Without Autism Diagnosis



95% CI


    Type 1

    Type 2










Heart Disease









In studies where lab data were available, people with autism tended to have higher triglyceride and lower HDL cholesterol levels compared with those without the diagnosis, though there were no significant differences in LDL and total cholesterol.

Risk of diabetes in conjunction with autism rose more sharply when looking at children versus adults, as did hypertension. These findings are “worth further exploration,” the investigators note. “One possible explanation could be that as age progresses, the prevalence of these [cardiometabolic] diseases increases in the entire population, decreasing the group differences between those with and without autism.”

The age-related disparity in hypertension that came “when we specifically looked at kids” was a surprise, Kahathuduwa told TCTMD. “What we think may have driven this . . . may be that the studies that include adults [aren’t] quite representative samples,” perhaps because people with autism have, on average, shorter life spans, he explained. “Patients who have hypertension and autism might not be living long enough to be captured in these databases.”

What’s Driving the Differences?

As for mechanisms, the researchers suggest type 1 diabetes might be the result of autoimmune disease. For type 2 diabetes, they suggest shared risk factors with diabetes, among them genetic variants, prenatal infections, medications, toxins, maternal obesity and diabetes, prematurity, and intrauterine growth restriction.

“Furthermore, behaviors and characteristics common among people with autism (eg, food selectivity, physical limitations, sedentary behavior, and sleep disturbances) and medications intended to improve challenging behaviors (eg, atypical antipsychotics) could be mediating or moderating the association between autism and cardiometabolic disease,” Dhanasekara et al note, pointing out genetic variants associated with lipid metabolism in patients with autism may be contributors to dyslipidemia.

Elizabeth M. Weir, PhD (University of Cambridge, England), writing in accompanying editorial, points out that all clinicians will see autistic patients in their practice. “Despite recognition of autism, autistic people still have poor long-term outcomes regarding their health and healthcare,” with data suggesting they are dying 12 to 30 years younger, she notes.

This latest study adds by broadening the types of health concerns experienced by people with autism, which she calls a “frameshift” that’s key to prevention and quality care.

We need to think outside the box and not be afraid to have the conversations about what this really means for the child long term. Carissa M. Baker-Smith

“Clinicians need to think carefully,” Weir stresses, “about how to intervene with health education to empower autistic people, while also recognizing the real challenges that trauma and mental health conditions can pose for maintaining healthy lifestyle behaviors.”

She, too, paints a complex picture when it comes to mechanism. “Autism is a highly heritable, polygenic condition, meaning that it is oftentimes hundreds of small changes across both rare and common variants in the genome that contribute to the likelihood that someone is autistic. Similarly, cardiovascular and metabolic conditions have strong genetic components,” though few studies have addressed a potential genetic overlap, Weir writes.

Baker-Smith, for her part, said she’s skeptical that the meta-analysis’ data are granular enough to delineate between type 1 and type 2 diabetes. “Because autism is multifactorial, I don’t think that it’s fair to say that this is some sort of genetic predisposition in most cases” for type 1, she said, though for type 2 the increase is likely driven by diet.

Kahathuduwa said next steps for research on cardiometabolic disease and autism should use statistical methods to come closer to showing causality, adding, “Then, we’re interested in seeing whether obesity and overweight may be mediating these associations—what comes first?”