An important class of genetic conditions are caused by myriad rare genetic mutations that may emerge for the first time when the person who has it is born, or may be limited to small extended lineages or fairly small populations, rather than to all of humanity.
A small number of rare genetic mutations are produced in each new generation. Advanced parental age and environmental exposures and stresses can increase their frequency, the absence of these factors can reduce their frequency.
The most harmful of these rare genetic mutations are never passed on in the gene pool because they produce miscarriages or cognenital defects that cause someone to die before reproducing or to be infertile or both. Others appear where they are mostly harmless, for example in an inactivated gene or one triggered by environmental conditions that are absent (e.g. a gene that weakens ability to handle thin oxygen at high altitude in someone who lives on a flat Pacific island), and are passed on from generation to generation as little more than evolutionarily neutral ancestry informative markers.
Certain genes are more sensitive to disruption than others. The most sensitive genes code intricate biological processes that call upon many genes to work together to function properly which can be interrupted by defects in that biological system at a great many different points. Thus, interruptions to the same system at any of these different points is going to appear phenotypically as a similar syndrome associated with the overall complex biological system's failure to work properly due to some rare genetic variant.
Autism and schizophrenia are two of many conditions that seem to fit the profile of having overwhelmingly genetic causes that act on the same or interrelated groups of biological systems in a personn causing phenomenologically similar symptoms, but in which the specific mutations that cause the syndrome vary almost as widely as the number of extended families that have cases of it.
These genetic syndromes (as noted by my source linked above) adhere to "the “Anna Karenina principle”, based on Tolstoy’s famous opening line:
“Happy families are all alike; every unhappy family is unhappy in its own way”, people without these syndromes are are alike in the relevant genes (or least exhibit a handful of common variations in those genes); people with these syndromes all differ from this fixed standard in their own way.
These conditions can be among the most severely impairing of genetic conditions, because they rely for a substantial share of their prevalance on new mutations at each generation rather than Mendelin inheritance from parents. They don't have to be fitness enhancing to endure.
In contrast, genetic conditions whose prevalance relies on the precisely same mutation being passed on from parent to child need to either be genetic fitness neutral, or genetic fitness enhancing (at least at a long run multi-generational, extended familiy level of analysis), or the genetic fitness disadvantage nature of the mutation will cause its frequency in the population to decline unless other unrelated genetic fitness boosts cause the individual's population to expand notwithstanding its other faults.
The prevalance of these conditions is a product of mutation rates per generation and the proportion of the genome that constitutes a target area that would induce these conditions if a random mutation of some type arose there. The larger the group of genes required to code a biological function, the more vulnerable it is to this kind of condition. Given the extremely complexity of the brain relative to that of the brains of most other animals, and relative to many other parts of our bodies, it isn't too surprising that many of these rare gene variant disruptions of complex systems conditions give rise to mental health conditions.
Even if every known sufferer from autism and schizophrenia and everyone who was a potential carrier of those conditions was removed from a sufficiently control population, that control population would develop new autistic and schizophrenic individuals in the very next generation and would in not many generations after that have the same prevalance of autism and schizophrenia as the source populations if environmental factors that impact mutation rates were held constant between the source and control populations.
Spontaneous v. Familial Inheritance As A Matter of Degree
Of course, Mendelin inheritance models and sponanteous rare genetic variant models are merely ideal types. Every Mendelin inherited genetic variant started at some point as a spontaneously arising rare genetic mutation. Every rare genetic variant that is passed on to at least one child has a Mendelin inheritance element.
Rare genetic variants that are dominant in expression get noticed and classified by the conditions they give rise to, and have negative fitness effects, are going to have predominantly sporadic rather than familial prevalance patterns even though they are genetic.
Recessive rare genetic variants considered
Rare genetic variants that are recessive in expression are invisible until (1) inbreeding causes someone to have two copies of the recessive gene, probably many generations later, (2) someone who has already inherited a recessive gene from a parent spontaneously has a germline mutation in the same gene from another parent (either the same, or a functionally equivalent version), or (3) two independent lineages each have mutations in the same gene (possibly many generations apart from each other) in a manner that are recessive in effect and admixture of those lineages gives rise to an individual who expresses the recessive gene variant's effect.
Given the fairly modest number of mutations per generation (IIRC about 2500) out of 3 billion genetic bases in an individual's genome, method (1), or method (3) with different versions of the recessive mutation that have the same effect, are likely to be much more common than the alternatives and from the point of view of an individual born with a recessive genetic condition it is almost always going to look like a predominantly Mendelin inheritance based genetic condition.
The odds of one novel germ line mutation from a father and one novel germline mutation from a mother of the same recessive mutation in the same gene arising at the same time in the same individual is almost nil unless it is the product of intentional manipulation or a highly targeted environmental effect (e.g. lateral gene transfer as a result of a shared germ line retroviral infection, or an environmental effect that only impacts the very end point of a chromosome for chemical geometry reasons of some kind). There is some circumstantial evidence that a retrovirus carried by a seasonal germ vector like lice may account for the seasons variation in schizophrenia (and perhaps also autism) prevalence.
Also, recessive conditions, as a result, are likely to start somewhere and see their prevalance slowly grow in a particular reproductive population due to population expansion and founder effects, or due to major selective advantages associated with them, rather than having the more random, generally global prevalance one expects in syndroms caused by rare genetic variants that influence complex biological systems that have themselves reached fixation or near fixation on a species or population.
Tell Tale Symptoms Of New Mutation Rare Gene Variant Syndromes
Strongly hereditary conditions that show similar prevalance in all global populations, like schizophrenia, are more likely to be dominant genetic conditions caused to a great extent by new rare gene variant mutations in large complex biological systems, while strongly hereditary conditions that show major frequency variations between different populations are more likely to be associated with Mendelin recessive genes.
Rare gene variants that are fitness enhancing in almost all circumstances, in contrast, whether dominant additive, or recessive, such as new mutations that enhance intelligence, should have strongly familial inheritance patterns. Every time one of these arises, it should tend to stay in the genome and should tend to stall on the road to fixation only if it interfers with some other fitness enhancing gene (perhaps either gene individually changes brain chemistry in a positive way but the presence of both genes at the same time has no effect or a negative effect) or has other negative side effects (e.g. an enhancement of connective tissue development that fosters brain connectivity but also increases the risk at any given point in life of developing connective tissue cancer, a scenario that has been proposed as a hypothetical basis for the fairly rapid spread of one of the breast cancer risk genes).