How exactly has the "refrigerator mother" theory of autism been discredited?
The origin of this question for me has been a natural derivation from the online invalidation of Gabor Mate, a Canadian physician who has written a number of best-selling books and invited a lot of controversy for heavily focusing on early childhood trauma as being the cause for a lot of conditions such as ADHD. While not autism, ADHD is still a form of neurodivergence and Mate has generally said that a lot of such conditions are an early adaptation to adverse conditions (such as a volatile home environment where the helpless child who cannot flee or exercise self agency turns 'inward' to tune out the stimuli). He has been rightfully criticized by research experts such as Russell Barkley for not being an authority in neurodivergence research, but I've always felt that lack of expertise should not in and of itself invalidate an otherwise plausible hypothesis (and Mate does have extensive trauma/addiction field work experience, where patients have a high comorbidity of psychiatric and neurodevelopmental conditions).
Some of the refutations of Mate's working model comes from evidence from twin studies showing a high etiological overlap in conditions such as ADHD as compared to siblings, as well as such conditions tending to run in family. However, the fact still remains that not 100% of identical siblings share etiological profiles, and that 'running in families = hardwired genetics' is an almost impossible dynamic to disentangle from environment since infants/children heavily mimic patterns from their environment, so certain neurodivergent profiles may be the result of genetics in the same way that being the child of a dysfunctional home can "genetically" wire someone to have anxious or disorganized attachments in relationships later in life. This is not to say that genetics don't factor strongly or even predominate; only that it's almost impossible to really separate nature from nurture in these kinds of studies due to possible multifactorial contributions to early cognitive development and limitations inherent to creating controlled scientific environments.
So in sum: I've always felt that while there's certainly strong evidence for heavy genetic contributions to neurodiversity, the environmental contribution, if any, has never been ruled out (to the extent that we can rule such things out, which is also questionable). I guess I just more broadly don't think all forms of neurodiversity are exactly a binary of 'having it' vs. 'not having it,' and I anecdotally have found it to be the case in my own family that there's essentially a direct correlation between the severity of ADHD and/or autistic symptoms and the level of childhood trauma we faced in early childhood (with that said, it could be that some basic core neurodivergence was genetically passed on to all of us, but the presentation was environmentally conditioned such that for some of us you'd barely notice any issues with daily life and others are lacking in key social skills like maintaining relationships, eye contact, touch aversion, poor social skills, etc.; I am hypothesizing that just like the spectrum itself can be socially influenced, the actual onset/emergence itself may be influenced by things like uterine environment, early childhood socialization, lack of adverse familial stimuli etc.).
Moreover, the few cases of actual feral child cases that appear in the historical record for the past few centuries seems to show essentially with no exception that such children lack the ability to learn basic things like language, normative mannerisms, intellectually advanced matters, etc. I understand the primarily-genetic-camp will argue that these children were perhaps abandoned to begin with due to being mentally disabled, but that itself is choosing an interpretation that is not demanded by the evidence since we don't really know those children's early background -- moreover, it's essentially universally accepted among child psychologists that there are critical windows of life for certain milestone developments that must be nurtured, and that missing these periods will essentially cause irreversible damage or at the least adversely impact later gains if any in those fields, so it sort of makes intuitive sense that [at least severe] neglect can cause outcomes that may be associated with autism: stunted language, sense of human touch, eye contact, etc.
In reading a lot of the online publications undermining Mate's hypotheses, some have (imo accurately) made the comparison to the "'since-discredited' refrigerator mother theory of autism." Which in fact made me curious, because absent evidence to the contrary, I would otherwise have a prima facie hunch that a mom having a chronically stressed pregnancy, early nurturing that avoids eye-to-eye contact and bodily touch, and not responding intuitively and empathetically to the infant's cues would be able to cause some form of a stunted neurological profile in an impressionable nascent brain. (After all, there's a reason experts recommend things like bonding, soothing, eye contact, hugging/kissing, 'smiling' with lighthearted facial impressions, etc. in child rearing.) And let's not forget, autism is a HUGE spectrum, so I'm not saying that the hunch I'd have in this regard applies to the severely autistic child in an otherwise well-adjusted NT family as much as that more 'mild' AuDHD person who has sensory and regulation issues but is otherwise 'normal.'
So I go do a preliminary search to see how the refrigerator mom theory was disproved, and I couldn't really find much on that front besides authors casually referring to it as discredited without exactly explaining WHY. If anything, lots of various subreddits have autistic/ND individuals anecdotally reporting having had refrigerator moms, although they may qualify this by saying that isn't at all the reason for their ND (or else they'd say that autistic people create autistic children, and because being autistic is essentially genetically hardwired, having a refrigerator mom is merely correlation instead of causation, since attributes of refrigerator mom profiles are essentially just autistic profiles -- and it's that hardwired autism which created both the parent's parenting style and the child's diagnosis in the first place). Now I'm not saying that none of these interpretations are possible, only that they're literally interpretations and that any other reading of such data is discredited as being out of line with professional consensus -- despite it being eminently reasonable (e.g., as mentioned, we know that infants and young children learn key skills during critical periods and that neglect of these leads to adverse consequences; and we also know that genetics doesn't neatly account for all cases of diagnoses, which proves that at least to some degree there are non-genetic / environmental / epigenetic factors at play... put two and two together, it seems a bit bizarre to flippantly dismiss early childhood neglect or disordered environment as essentially negligible to neurodevelopmental profiles).
TL;DR: What is the basis for the widespread claim that the refrigerator mom theory of autism has been entirely discredited?