This episode reviews ADDitude reader surveys on rejection sensitive dysphoria and hormonal shifts in ADHD, then turns to The Transmitter’s coverage of an autism–Parkinson’s registry study, efforts to identify autism subtypes, and nonhuman primate models. Across topics, the conversation distinguishes self-report from clinical evidence, association from causation, and exploratory findings from ready-for-practice guidance.
RSD survey results: “Rejection immobilizes me”
ADDitude’s reader survey of nearly 3,500 adults with ADHD focuses on rejection sensitive dysphoria, or RSD. RSD is a term coined by Dr. William Dodson to describe extreme emotional pain triggered by real or perceived criticism or rejection. The article states that RSD is not an official DSM-5 diagnosis and is not considered a core ADHD symptom, though it frames RSD as part of ADHD emotional dysregulation for many people. The survey is a reader survey, not a clinical sample, so it reflects experiences among respondents rather than prevalence in the broader ADHD population.
The article, updated October 8, 2026, reports that 63 percent called RSD somewhat or very debilitating, and two-thirds said it has a severe impact on emotional health. For 64 percent, the pain begins within one minute of a trigger; for most, it lasts hours or days. Triggers described are ordinary: an unanswered text, a reminder to do a forgotten task, a facial expression, or silence during a conversation. Reader descriptions include feeling as if a punch landed in the chest or restless legs inside the whole body. One 34-year-old woman described RSD as almost feeling like she was dying; another described leaving a meeting with a boss, going to the car, and being so overwhelmed they could not breathe.
Seventy-nine percent said they remember experiencing RSD before middle school. The article suggests RSD screeners could be used to help detect early signs of ADHD and emotional dysregulation in children with less noticeable symptoms, but presents this as an implication, not an established screening recommendation. Early memories are meaningful, but retrospective reports can be shaped by later experiences; a screener would need validation.
Reported helpful strategies included exercise (48 percent), mindfulness (46 percent), stimulant medication (44 percent), talk therapy (43 percent), and cognitive behavioral therapy (33 percent). These are self-reported benefits among people who tried each strategy, not controlled or comparative evidence. Medication decisions belong with a clinician. The broader point is that emotional dysregulation deserves attention even if RSD is not a formal diagnosis.
When Hormones Push ADHD Symptoms Over the Edge
This reader-driven ADDitude piece, updated October 12, 2026, discusses how hormonal shifts across puberty, pregnancy, postpartum, perimenopause, and menopause can intensify ADHD symptoms. A 2022 ADDitude survey found that 70 percent of women said ADHD had life-altering impacts in their 40s and 50s. One reader said perimenopause “completely unleashed and undone all of my coping mechanisms.”
The article summarizes research generally: estrogen fluctuations can affect mood, executive function, and emotional regulation. In the follicular phase, rising estrogen may help mood and focus; after ovulation, when estrogen drops, some women report more emotional dysregulation and executive dysfunction, and the article says that phase is associated with decreased effectiveness of ADHD medication. It does not name the studies or samples, and the reader accounts are self-reports. The topic remains understudied.
Practical takeaways focus on tracking and adjusting expectations, not guessing at hormones. Readers used calendars, phone apps, Oura rings, and Garmin watches. One reader said they take 20 mg in the first half of their cycle and 30 mg of lisdexamfetamine in the second half, and can feel when it is time to increase. Another said hormone replacement therapy and other medications helped. These are individual reports, not universal guidance. Tracking can reveal patterns and support a conversation with a clinician; changing stimulant doses or adding hormone replacement therapy is individual medical care. The article also includes one man who thinks hormonal fluctuations affect his ADHD, but it does not explain male hormonal mechanisms, so the conversation cautions against overgeneralizing. A pattern such as RSD worsening at certain cycle points can be brought to a professional rather than treated as a self-diagnosis.
Large study links autism to Parkinson's disease
The Transmitter covered a JAMA Neurology study published May 27, 2025, linking autism to a more than fourfold increased risk of early-onset Parkinson’s disease. The study used Swedish national records for more than 2.2 million people born between 1974 and 1999, followed through 2022. By the end, 24 of 51,954 autistic participants (0.05 percent) had a Parkinson’s diagnosis, versus 438 of 2,226,611 non-autistic participants (0.02 percent).
These are small absolute numbers because everyone in the study was under 50, with a median age of 34. This is early-onset Parkinson’s, which may have different risk factors from later-onset disease. The association translated to a 4.4 times greater risk and held after adjustment for sex, socioeconomic status, and family history of mental illness or Parkinson’s. It is an association, not proof that autism causes Parkinson’s.
After adjusting for depression and antipsychotic drug use, autism was still associated with a twofold increased risk. Study investigator Weiyao Yin said depression and antipsychotic use are both more common among autistic people and associated with Parkinson’s or Parkinson’s-like symptoms, and the conditions appeared to influence risk independently based on comparisons between groups. She also suggested risk may be significantly more elevated for people with autism plus depression, antipsychotic use, or both. That is a subgroup interpretation, not a new clinical rule.
Limitations include registry gaps: Ezra Susser, who was not involved, noted that the Swedish registry did not include outpatient data until about 2000, so it may have missed autism cases before then. Sergio Starkstein said diagnosis quality cannot be fully assured, but the study’s size compensates. Registry studies trade precision for scale; they can miss people, and diagnostic practices change over time. The study is described as the first rigorous longitudinal study of this potential link. Principal investigator Sven Sandin said health services should keep autistic people under long-term observation for neurological conditions like Parkinson’s. That is a call for clinical awareness, not a reason for every autistic person to fear Parkinson’s. The absolute risk remains low, and the pattern needs larger studies and longer follow-up. The conversation also notes that neurodevelopmental and neurological conditions can overlap across the lifespan.
The spectrum goes multidimensional in search of autism subtypes
This Transmitter feature from August 14, 2025, examines researchers looking for multiple “autisms” or multidimensional subtypes. A SPARK analysis sorted 5,392 autistic children into four subgroups. The largest, about 37 percent, had social challenges and repetitive behaviors but tended to reach milestones around the same time as neurotypical peers, with many having ADHD, anxiety, depression, or OCD. The smallest, about 10 percent, had more profound features and delays and, compared with children in the other groups, were more likely to have rare gene variants not inherited from a parent.
The rationale is that “spectrum” captures variability but does not say much about biology or prognosis. Subgroups could make clinical trials more homogeneous and eventually help match supports. The results are not ready to use as labels. Fewer than 30 percent of autism diagnoses are linked to specific gene disruptions, so genetic subgrouping leaves many people out, and the SPARK groups may be refined with more data.
An MRI study applied two different machine-learning algorithms to structural MRI data from 4,115 people, including 1,305 with autism, 987 with ADHD, and 1,823 neurotypical controls. Both algorithms found two autism subgroups based on overall brain size, but depending on the algorithm, some people changed groups. When the team looked at specific brain areas, new groups emerged. Clara Pecci Terroba said researchers should spend more time understanding what the algorithms are doing and what they are picking up. Matthew Siegel compared describing brain surface area and grey matter volume to explain autism to answering “What makes this car run?” by describing the car’s size and metal thickness: those are parameters, not mechanism.
A preprint using EEG, or brain-wave recordings, clustered autistic people into two groups, one with more excitation than neurotypical controls and one with less; researchers recapitulated those signatures in mice. That work is exploratory. Alessandro Gozzi argued for a middle ground: not one monolithic autism, not pure individual differences, but robust, reproducible groups that can be decoded into testable mechanisms. The article does not say subtypes have been found; it says the field is searching for them with mixed success.
For families, the conversation notes that a label is not destiny and the same supports will not fit everyone, but new subtypes should not become new stereotypes. Siegel says categorical boundaries are porous and inaccurate, but categorization is still needed for progress. The useful question is which group a person is in for a particular purpose, and whether that changes what is offered—not which autism type someone is.
Autism research involving monkeys contends with uphill battle
The Transmitter piece covers nonhuman primate models of autism, including genetically modified macaques with a CHD8 variant and genetically modified macaques and marmosets with SHANK3 variants. It is a story about the science and about funding and policy pressures, including the planned downsizing of the Oregon National Primate Research Center.
Some researchers say monkey models can capture features mouse models cannot, such as certain social and sleep patterns, but the work is expensive and technically demanding. Alternatives like organoids and human recordings are growing. Guoping Feng said no model is perfect, not even a human is a perfect model for another human. Monkey models are not required for clinical trials, though they can increase confidence. The conversation treats this as a policy conversation as much as a scientific one: animal models are one tool among many, with real limits and real costs.
The episode closes by noting that none of this replaces individual clinical care, and that listeners who recognize something in these topics can bring it to a professional who knows their history.
Podcast Transcript
Aaron: Welcome to Living on the Spectrum. I'm Aaron, and with me is Jamie. Today I want to start with a line from an ADDitude survey that stopped me: “Rejection immobilizes me.” The survey went to nearly 3,500 adults with ADHD. One 34-year-old woman described rejection sensitive dysphoria, or RSD, as almost feeling like she was dying. Another described leaving a meeting with their boss, going to the car and being so overwhelmed they couldn't breathe. Jamie, before we get into numbers, what is RSD?
Jamie: RSD is a term coined by Dr. William Dodson. It describes extreme emotional pain triggered by real or perceived criticism or rejection. The article is careful to say it is not an official DSM-5 diagnosis and not considered a core ADHD symptom, even though it frames RSD as part of ADHD emotional dysregulation for many people. The survey is a reader survey, not a clinical sample, so it tells us about experiences among people who answered, not prevalence in the whole ADHD population.
Aaron: The ADDitude article, updated October 8, 2026, reports that 63 percent called RSD somewhat or very debilitating. Two-thirds said it has a severe impact on emotional health. For 64 percent, the pain begins within one minute of a trigger, and for most it lasts hours or days. The triggers are ordinary: an unanswered text, a reminder to do a forgotten task, a facial expression, silence during a conversation. That ordinariness is what gets me.
Jamie: Yes. It is not just “being sensitive.” The article includes physical descriptions like a punch to the chest or internal discomfort like restless legs inside the whole body. But because RSD is not formal, we should be careful not to turn it into a diagnosis or use it to explain every strong reaction. The survey can tell us people recognize this pattern; it cannot tell us cause or prevalence.
Aaron: Seventy-nine percent said they remember experiencing RSD before middle school. The article says that suggests RSD screeners could be used to help detect early signs of ADHD and emotional dysregulation in children with less noticeable symptoms. Is that ready for practice?
Jamie: The article presents that as an implication, not an established screening recommendation. Early memories are meaningful, but retrospective reports can be shaped by later experiences. A screener would need validation. I would treat it as a research question, not something to start labeling children with.
Aaron: Fair. What did respondents say helped? Exercise 48 percent, mindfulness 46, stimulant medication 44, talk therapy 43, and cognitive behavioral therapy, or CBT, 33. But those are people who tried each strategy and found benefit, not controlled evidence.
Jamie: Right. Self-reported benefit. It is useful for generating options and reducing isolation, but not comparative effectiveness. Medication decisions belong with a clinician. The broader point is that emotional dysregulation deserves attention, even if RSD is not a formal diagnosis.
Aaron: That connects to another ADDitude piece about hormones. It is reader-driven, updated October 12, 2026, about how hormonal shifts across puberty, pregnancy, postpartum, perimenopause and menopause can intensify ADHD symptoms. A 2022 ADDitude survey found 70 percent of women said ADHD had life-altering impacts in their 40s and 50s. One reader said perimenopause “completely unleashed and undone all of my coping mechanisms.” That is vivid.
Jamie: It is. The article summarizes research generally: estrogen fluctuations can affect mood, executive function and emotional regulation. In the follicular phase, rising estrogen may help mood and focus; after ovulation, when estrogen drops, some women report more emotional dysregulation and executive dysfunction, and the article says that phase is associated with decreased effectiveness of ADHD medication. But it does not name the studies or samples, and the reader accounts are self-reports. The topic remains understudied.
Aaron: So practical takeaways are tracking and adjusting expectations, not guessing at hormones. Readers used calendars, phone apps, Oura rings and Garmin watches. One reader said they take 20 mg in the first half of their cycle and 30 mg of lisdexamfetamine in the second half, and can feel when it is time to increase. Another said hormone replacement therapy, or HRT, and other medications helped. We should not present those as universal.
Jamie: No. Tracking can reveal patterns and support a conversation with a clinician. Changing stimulant doses or adding HRT is individual medical care. The article also includes one man who thinks hormonal fluctuations affect his ADHD, but it does not explain male hormonal mechanisms. So we should not overgeneralize.
Aaron: Still, if someone notices RSD is worse at certain points in their cycle, that is a reasonable thing to note. It is a pattern to bring to a professional, not a self-diagnosis.
Jamie: Agreed. And it keeps us from treating emotional regulation as separate from the rest of the body.
Aaron: Let's move to a bigger research story. The Transmitter covered a JAMA Neurology study published May 27, 2025, linking autism to a more than fourfold increased risk of early-onset Parkinson's disease. It used Swedish national records for more than 2.2 million people born between 1974 and 1999, followed through 2022. By the end, 24 of 51,954 autistic participants, or 0.05 percent, had a Parkinson's diagnosis, versus 438 of 2,226,611 non-autistic participants, or 0.02 percent. Jamie, help us read that.
Jamie: The first thing is that these are small absolute numbers because everyone was under 50, with a median age of 34. So this is early-onset Parkinson's, which may have different risk factors from later-onset disease. The association translated to a 4.4 times greater risk, and it held after adjustment for sex, socioeconomic status and family history of mental illness or Parkinson's. That is a real signal, but it is an association, not proof that autism causes Parkinson's.
Aaron: Then there is another number. The article says that after adjusting for depression and antipsychotic drug use, autism was still associated with a twofold increased risk. What does that tell us?
Jamie: It tells us the association remained after accounting for those factors. The study investigator, Weiyao Yin, said depression and antipsychotic use are both more common among autistic people and associated with Parkinson's or Parkinson's-like symptoms, and the conditions appeared to influence risk independently, based on comparisons between groups. She also suggested risk may be significantly more elevated for people with autism plus depression, antipsychotic use, or both. But that is a subgroup interpretation, not a new clinical rule.
Aaron: What about the limitations? Ezra Susser, who was not involved, noted the Swedish registry did not include outpatient data until about 2000, so it may have missed autism cases before then. Sergio Starkstein said you cannot be sure of diagnosis quality, but the size compensates.
Jamie: Yes. Registry studies trade precision for scale. They can miss people, and diagnostic practices change over time. The study is still described as the first rigorous longitudinal study of this potential link, which is notable. But the absolute risk is low. The principal investigator, Sven Sandin, said health services should keep autistic people under long-term observation for neurological conditions like Parkinson's. That is a call for clinical awareness, not a reason for every autistic person to fear Parkinson's.
Aaron: So if a family is reading the headline, what is the grounded takeaway?
Jamie: Keep the association in perspective. It does not mean Parkinson's is common in autistic adults under 50. It means researchers are finding a pattern worth following with larger studies and longer follow-up. And it reminds clinicians that neurodevelopmental and neurological conditions can overlap across the lifespan.
Aaron: That is measured. Now, another Transmitter piece challenges the idea of a single autism spectrum. It is a feature from August 14, 2025, about researchers looking for multiple “autisms” or multidimensional subtypes. The SPARK analysis sorted 5,392 autistic children into four subgroups. The largest, about 37 percent, had social challenges and repetitive behaviors but tended to reach milestones around the same time as neurotypical peers, with many having ADHD, anxiety, depression or OCD. The smallest, about 10 percent, had more profound features and delays and, compared with children in the other groups, were more likely to have rare gene variants not inherited from a parent.
Jamie: The rationale is that “spectrum” captures variability but does not tell you much about biology or prognosis. Subgroups could make clinical trials more homogeneous and eventually help match supports. But the results are not ready to use as labels. Fewer than 30 percent of autism diagnoses are linked to specific gene disruptions, so genetic subgrouping leaves many people out. And the SPARK groups may be refined with more data.
Aaron: There is also an MRI study that made me pause. Researchers applied two different machine-learning algorithms to structural MRI data from 4,115 people, including 1,305 with autism, 987 with ADHD and 1,823 neurotypical controls. Both algorithms found two autism subgroups based on overall brain size, but depending on the algorithm, some people changed groups. When the team looked at specific brain areas, new groups emerged. That is a lot of instability.
Jamie: Yes. Clara Pecci Terroba said researchers should spend more time understanding what the algorithms are doing and what they are picking up. Matthew Siegel made a great analogy: describing brain surface area and grey matter volume to explain autism is like answering “What makes this car run?” by describing the car's size and metal thickness. Those are parameters, but they do not tell you the mechanism.
Aaron: So subgroups might be useful, but only if they are reproducible and explain something. The article also mentions a preprint using EEG—electroencephalography, or brain-wave recordings. That preprint clustered autistic people into two groups, one with more excitation than neurotypical controls and one with less, and researchers recapitulated those signatures in mice. That is exploratory.
Jamie: Right. Alessandro Gozzi argues for a middle ground: not one monolithic autism, not pure individual differences, but robust, reproducible groups that can be decoded into testable mechanisms. That is the key. The article is not saying we have found the subtypes. It is saying the field is searching for them, with mixed success.
Aaron: For families, that might actually be reassuring: a label is not destiny, and the same supports will not fit everyone. But we should not replace one stereotype with new subtypes.
Jamie: Exactly. Siegel says categorical boundaries are porous and inaccurate, but categorization is still needed for progress. The useful question is: which group is this person in for this purpose, and does it change what we offer? Not: which autism type are you?
Aaron: One more research note before we close. The Transmitter had a piece on nonhuman primate models of autism, including genetically modified macaques with a CHD8 variant and genetically modified macaques and marmosets with SHANK3 variants. It is a story about the science and about funding and policy pressures, including the planned downsizing of the Oregon National Primate Research Center. Some researchers say monkey models can capture features mouse models cannot, like certain social and sleep patterns, but the work is expensive and technically demanding. Alternatives like organoids and human recordings are growing.
Jamie: Yes. Guoping Feng said no model is perfect, not even a human is a perfect model for another human. That is a useful frame. The article also notes monkey models are not required for clinical trials, though they can increase confidence. It is a policy conversation as much as a scientific one. For our listeners, I would keep it as context: animal models are one tool among many, with real limits and real costs.
Aaron: That is a good place to land. To recap: RSD is a widely described but non-official experience that deserves attention without becoming a label; hormones can shape ADHD symptoms, but tracking is not the same as treating; the autism-Parkinson's link is a real association with small absolute risks and registry limits; and autism subgrouping is promising but still exploratory. We did not solve any of it, which feels honest.
Jamie: And none of this replaces individual clinical care. If something here resonates, bring it to a professional who knows your history.
Aaron: Thanks for listening to Living on the Spectrum. Links to the articles we discussed are on the episode page. Take care.
Jamie: Take care.
References
- Autism research involving monkeys contends with uphill battle
- Large study links autism to Parkinson's disease
- The spectrum goes multidimensional in search of autism subtypes
- RSD survey results: “Rejection immobilizes me”
- When Hormones Push ADHD Symptoms Over the Edge
- “What Should I Do with My Life?” 6 Revealing Questions for Adults with ADHD
