Autism, that's the name of a relatively common developmental and behavioural disorder, that is it is a condition in of which a child's mental and behavioural development is impaired, or stunted. According to Better Health Victoria, a Victorian Government website, the following are the symptoms:
Language – absent, delayed or abnormal patterns
Play – isolated, repetitive, a preference for predictable play, difficulty with flexible thinking, such as pretending that a box is a boat or a stick is a horse
Body movements – stereotypical behaviour, such as flapping and toe walking, and other behaviours that may cause self-injury, such as hand biting
Restricted or obsessive behaviour – with favourite topics, objects, places, people or activities
Rituals and routines – bring some order to chaos and confusion. A change to routine can result in the individual displaying high levels of stress, anxiety or acting out
Tantrums – can be a way to express extreme confusion, stress, anxiety, anger and frustration when unable to express their emotions in another way
Sensory sensitivities – to certain sounds, colours, tastes, smells and textures.
These symptoms usually become apparent in the child before the age of four. There are some children in of which the symptoms can begin later in life. These children are said to have regressive, or late-onset autism.
Additionally autism, isn't a single condition, it's a "spectrum" of conditions known collectively as autism spectrum disorders (ASD). Asperger syndrome, is an ASD that's usually described as comparatively mild, in that people with it are usually high functioning and capable of living independently.
The cause of autism isn't exactly clear, but the environment a child grows up in and their genetics appear to play pivotal roles. While, from what I could find, there have been no specific genetic risk factors (i.e. specific faulty genes that contribute to the disorder) identified so far there appears to be a strong and definite genetic component of autism.
A study was conducted a few years ago in of which patients with autism that had since died due to a wide range of unrelated causes and a few living patients were found to either have swelling in their brains (or neuroinflammation; these patients were the dead ones, such tests would be dangerous to perform on living subjects) or had markers of inflammation in their cerebrospinal fluid (CSF; a fluid, that is not blood that flows throughout your brain and spinal cord. This was taken from living subjects, these markers of inflammation were determined, directly, from the deceased's brain tissue in the case of the deceased subjects). None of these patients' causes of death or their pre-existing conditions, both in the living and deceased subjects, could have explained this, thus leaving their one major and apparent thing in common to be the cause: their autism. One study demonstrated that persons with late-onset autism have abnormal gastrointestinal (GI) flora (bacteria living in their stomachs, small and large intestines) compared to unaffected individuals. This would further support the aforementioned neuroinflammatory hypothesis of autism.
Another finding that supports the neuroinflammatory hypothesis of autism is that patients with autism often find some respite from their symptoms from antidepressant therapy. It so happens there's a theory that that depression is due to certain pro-inflammatory compounds that are released during an inflammatory reaction. This theory is well-substantiated by the fact that every antidepressant I'm aware of, at least, has been demonstrated to posses additional anti-inflammatory effects.
You may, of course, be asking me how could we maintain that there is a strong genetic component to autism while also proposing that it is due to an inflammation of the brain when inflammation is usually, as the abnormal GI flora finding supports, environmental in cause and not genetic. While there are some exceptions to this very broad view of inflammation genes do a play a key role in determining our susceptibility to different microbes and the inflammation they can cause. Faulty genes can also cause inflammatory responses to things that are meant to be in our body, a good example of multiple sclerosis which is when an inflammation arises in response to the myelin we all have insulating our brain, spinal cord and nerve cells.
I would like to clear up a few things some of you may have heard about the cause of autism, there is a large body of evidence to discredit the following theories:
- The immunisation-based theory, that vaccines cause autism 
- That a casein (milk protein)-free and gluten-free diet can prevent or reverse the symptoms of autism 
Although both these theories would be supported by an inflammatory framework of autism the vast body of evidence has severely discredited them. The casein and gluten theory could also work without the inflammatory theory of autism.
Casein and gluten are interesting proteins in that the body converts them to opioid peptides. The term opioid refers to the fact that these peptides bind to and activate the very receptors that the opioids, morphine, codeine, fentanyl and pethidine (meperidine; demerol) bind to and activate. The term peptide refers to the chemistry of these compounds. Chemically they are basically relatively (relative to the giant chains that make up proteins) short chains of amino acids (the building blocks of proteins) chemically bonded head-to-tail. Opioids, even when they are administered in low doses, if they are given to young lab animals has, in one study, been found to lead to behaviours and other symptoms consistent with autism in humans. This finding, however, makes no sense. See, if this were true for humans, guess what? Breast milk, or for that matter any milk would cause similar permanent deficits in children as they are good sources for casein. The amount and potency of opioid peptides produced after the consumption of casein and gluten is minuscule. The strongest of the casomorphins (casein-derived opioid peptides) is 1/10th the strength of codeine or less than 1/100th the strength of morphine. The other thing about this theory is that it proposes that removal casein and gluten will cause the disappearance of symptoms. The study in lab animals showed permanent damage, not reversible damage. Additionally, as animals and humans develop our blood-brain barrier, the cellular barrier that keeps large and water soluble compounds, both of which peptides like casomorphins and gluten opioid peptides are, away from the brain and spinal cord.
While I still stand by the evidence, recently there was a journal article published that proposed that maybe these anti-vaccine cranks were onto something. Thiomersal, a common mercury-based preservative used in vaccines, has been found, when administered to baby rats, to cause permanent changes in the rats' mu opioid receptors. The mu opioid receptor (MOR) is a receptor that opioids like morphine, fentanyl and pethidine bind to in order to elicit the bulk of their effects like pain relief, sedation and, in sufficient doses, memory and learning impairment. The changes found in these rats' brains included a decreased mu opioid receptor expression (or density) in the dentate gyrus, a part of the brain that appears to play a key role in memory formation. While this finding is unclear in the context of autism it is interesting and is causing, in many countries, the removal of thiomersal from vaccine formulations. This study also found that thiomersal causes increased MOR expression (or densities) in the periaqueductal grey, a part of the brain that appears to play a key role in pain perception. Some anecdotal (anecdotes are stories that are recalled, in this case, likely to the patients' health care professional. This means that as far as the strength of evidence goes this is very weak, hearsay evidence) evidence suggests that autistic patients have an increased sensitivity to pain, which is in accordance with this finding.