Sex and gender differences in the molecular etiology of Parkinson’s diseaseconsiderations for study design and data analysis

Samantha L. Schaffner, Kira Tosefsky, Amy M. Inskter, Silke Appel‐Cresswell, Julia M. Schulze‐Hentrich•View original

ComprehensiveExpertshimmer voice

Let’s get straight to the substance. The big picture here is not that men and women carry different genomes for Parkinson’s disease. It’s that the same broad genetic liability is routed through sex-linked biology that shapes when, where, and how vulnerability shows up. Schaffner and colleagues frame it as three interlocking levers: chromosomal and early developmental programming, gonadal hormones with age-dependent effects, and sex-biased transcriptomic and epigenomic states, especially in the substantia nigra. Those levers engage at specific life-course inflection points, such as conception, adolescence, and midlife, so the same risk can play out differently across time in a male versus a female brain. That’s the through-line. And an important caveat rides alongside it: females are under-represented in a lot of the underlying datasets, which limits how fine-grained we can be about mechanisms. Start with the earliest programming. Weeks after conception, the Y chromosome’s SRY gene switches on in embryos destined to be male, steering testes formation and testosterone production. Then a twist you already know from developmental neuroendocrinology: testosterone can be locally converted to estradiol in the limbic system. That aromatization shapes astrocyte morphology, enhances glutamate release, and biases cell survival programs. In other words, even before birth, the circuitry that will one day carry dopamine is experiencing sex-dependent trophic and synaptic signals. Those organizational effects lay down the responsiveness to the hormone surges of adolescence. They don’t decide Parkinson’s outcomes, but they set the slope of the hill. Now, layer in X-dose biology. In females, one X chromosome is inactivated in each cell, but that inactivation is random, so you get mosaicism—one patch of cells with one X active, another patch with the other. Some genes escape that inactivation and are expressed from both X chromosomes, so their dosage is genuinely higher in females. Add genomic imprinting, which is parent-of-origin specific expression, and you’ve created a heterogeneous, sex-linked regulatory landscape before any environmental exposure comes into play. This is not abstract. In the substantia nigra, studies consistently report sex-specific expression of Parkinson’s-relevant genes. The ARMCX2 gene, an X-linked regulator of mitochondrial trafficking, tends to be upregulated in females but downregulated in males. The IFITM2 gene goes the other way—down in females, up in males. The DYNC1LI1 and REEP1 genes skew down in males. These are some of the bricks; the wall is a sex-biased transcriptome in dopaminergic territory that’s already set up to respond differently to stress. There is a Y-linked piece, too. SRY itself regulates enzymes in dopamine biosynthesis in models, and its expression increases after six-hydroxydopamine injury in male animals. That’s a hint—not proof—that Y-driven regulation could shift how male dopaminergic neurons handle insults. Pair that with robust observations that males show higher messenger RNA for the SNCA gene and the PINK-1 gene in substantia nigra tissue and isolated dopaminergic neurons. Females, in contrast, show more expression of neuronal maturation programs. You can hear the implication: the baseline homeostasis of synuclein handling and mitochondrial quality control may be tuned differently by sex from the outset. And then hormones arrive like a periodic forcing function. The most reproducible biology here is estrogen’s neuroprotection. Women with later menopause tend to have later Parkinson’s onset, and that roughly ten-fold drop in estradiol during the menopausal transition tracks with increased susceptibility. In epidemiology, postmenopausal hormone therapy is associated with reduced Parkinson’s risk. In clinical cohorts, estrogen therapy introduced early in the disease course associates with lower symptom severity on the Unified Parkinson’s Disease Rating Scale. Mechanistically, the cellular pathways line up: estrogen activates the mitogen-activated protein kinase and extracellular signal-regulated kinase cascade and the phosphoinositide three-kinase and Akt pathway, both of which push dopaminergic neurons away from apoptosis and protect against glutamate and oxidative hits in culture. In rodents, females and estrogen-treated males lose fewer dopaminergic neurons to toxins, with shifts toward anti-apoptotic protein profiles. The direction is consistent across models. The magnitude depends on timing and context. And what about androgens? That picture is muddy. Some studies report lower testosterone in men with Parkinson’s disease, but testosterone replacement yields inconsistent motor and non-motor outcomes. In animal models, testosterone’s effects are mixed and can flip sign depending on conversion to estradiol. The signal is clearly weaker and less coherent than the estrogen story, which is why Schaffner and colleagues keep coming back to estradiol decline at midlife as an inflection point for risk. Adolescence matters too. The organization-activation hypothesis fits what’s seen here: organizational effects laid down prenatally and in early life meet the activating surges of puberty and peripuberty to install durable sex differences in circuitry and transcription. That’s a setup for the midlife fall in gonadal hormones to land differently across sexes. Remember, many prodromal Parkinson’s features—things like constipation, rapid eye movement sleep behavior disorder, hyposmia—show up years before diagnosis, often straddling midlife. The timing aligns with a hormone-gene conversation that’s gone from whisper to shout. Now, zoom into the substantia nigra’s molecular readouts. If there’s a hub for sex-biased regulation in Parkinson’s biology, this is it. Across region-spanning meta-analyses, two hundred thirty-seven genes in substantia nigra are differentially expressed in sex- and disease-specific patterns, with seventy-five relatively increased in males and one hundred sixty-two increased in females when you compare Parkinson’s to control. Fifteen genes keep their sex- and disease-specific expression across substantia nigra, striatum, and frontal cortex. That’s a small but instructive set that probably includes nodal regulators. Within the substantia nigra, nine chromosomal segments show differential expression when you compare male controls to male Parkinson’s and female controls to female Parkinson’s. Two segments—Yq11.1 and 5q34—are common to both comparisons, a reminder that some chromosomal architecture shifts cut across sex while others don’t. Single-cell and cell-type-enriched analyses sharpen the contrast. In substantia nigra tissue, thirty-six genes are flagged as female-specific and five hundred thirty-nine as male-specific at a false discovery rate under five percent, with thirty-seven described as sex-dimorphic. When you collect dopaminergic neurons specifically and compare Parkinson’s to control within sex, eighty-six genes come up as differentially expressed across male and female analyses, and five of those also differ between male and female controls. On an X-chromosome-enriched array—the X3P platform—one hundred twenty genes show sex-specific expression regardless of Parkinson’s status. Stratify by sex and disease, and the differential expression balloons: two hundred eighty-eight genes in female Parkinson’s versus female control, two hundred ninety-two in male Parkinson’s versus male control. The pattern is not a sprinkling; it’s a wholesale redrawing of transcriptional state space by sex within a vulnerable circuit. Epigenetics tracks with that. In cortex neurons from Parkinson’s brains, sex- and disease-specific CpG methylation shifts involve targets you’d expect—the PARK7, SLC17A6, PTPRN2, and NR4A2 genes among others. Meta-analyses suggest females show more pronounced changes in histone and chromatin-modifying gene expression, which could amplify transcriptomic divergence over time. In blood, there are Parkinson’s- and sex-specific methylation regions too, and the variance explained in female analyses increases when genotype is included, hinting at gene-epigenome coupling that is sex contingent. Not every tissue cooperates: platelet assays of mitochondrial DNA methylation at the MT-TL1, MT-CO2, and MT-CO3 loci show no Parkinson’s- or sex-associated differences at those loci. That negative control is useful; it tells you where not to look first. Peripheral immune readouts echo a sex bias. In blood, about one hundred fifteen genes distinguish female Parkinson’s patients from female controls, enriched for immune functions and B cell signaling. You could take that two ways: immune signatures mirror what’s going on in the brain, or immune cells themselves live in sex-biased regulatory states that interact with the disease milieu. Both are plausible. What the data do say is that immune and glial programs are not operating against a neutral background; they’re riding sex-specific baselines sculpted by hormones and chromatin. Let’s pull genetics back into the frame. At the level of common variant risk, men and women look remarkably similar. Across large genome-wide association studies, heritability estimates are closely matched—about zero point two one in males versus zero point one nine in females—and the genetic correlation between male and female summary statistics is high, roughly zero point eight seven seven. That means the bulk of shared polygenic signal is the same. But “similar” doesn’t mean “identical.” In a Korean dataset built on the Korean Chip, five loci cross the association threshold in female-only analysis—one near the LRRK2 gene, four near the SNCA gene—with a p-value below one point zero three times ten to the minus seven. Male-only analyses in that cohort yield no hits. There are also reports of sex-by-locus interactions with reproductive timing, such as age at menarche and age at menopause, nudging Parkinson’s associations at several sites. The takeaway is pragmatic: don’t expect an entirely different architecture by sex, but do expect pockets where sex and reproductive history shape penetrance or expression. You can already see how this meets mitochondria and proteostasis. If males run higher PINK-1 transcription in substantia nigra, and if females emphasize maturation gene programs, you’re looking at sex-skewed set points for mitophagy and differentiation. No one has shown sex-differential mitophagy flux in human dopaminergic neurons in vivo. That measurement is still out of reach. But the messenger RNA differences, the estrogen-tilted redox protection via mitogen-activated protein kinase and extracellular signal-regulated kinase and phosphoinositide three-kinase and Akt, and the catalog of sex-biased autophagy and lysosomal genes together argue for regulatory, not hard-coded, dimorphism in mitochondrial maintenance and protein handling. That matches the clinical pattern: broadly similar genetic risk, different trajectories and modifiability across the life course. The immune and glial axis slots in here too. Estrogen doesn’t just touch neurons; it shapes astrocyte and microglial states. Early-life aromatization to estradiol molds astrocyte morphology and glutamatergic tone. Later in life, estradiol supports anti-inflammatory phenotypes and antioxidant defenses. Schaffner and colleagues synthesize this into a convergent view: sex-biased transcriptional and epigenetic set points, modulated by hormone status, tilt inflammatory cascades, antigen presentation capacity, and oxidative stress handling in substantia nigra. That tilt is not necessarily protection for females across the board—recall the more extensive methylation remodeling reported in females—but it is a different starting line and a different trajectory under load. Now, a reality check about scale and context. Parkinson’s remains more prevalent in men—roughly one point four times as many men as women carried the diagnosis between nineteen ninety and twenty sixteen—although more recent analyses suggest the gap might be narrowing. Globally, prevalence rose dramatically from nineteen ninety to twenty nineteen, and projections point to tens of millions affected by twenty fifty. Those numbers set the public health stage, but they don’t overturn the molecular message: sex is not a covariate to be neutralized; it’s a biological system organizing risk. And this biology is hierarchical—chromosomes to hormones to chromatin to transcripts—and contingent on timing. You’ll notice how consistently the findings converge on a few methodological imperatives. Balance the cohorts. Don’t treat sex as a nuisance variable to be adjusted away. Use covariate adjustment when that’s all the data allow, but, where powered, run sex-stratified models and explicit sex-by-genotype or sex-by-hormone interaction terms. Midlife hormonal states matter—menopausal status, hormone replacement therapy, contraceptive use, parity—so collect them. In transcriptomic and epigenomic designs, put sex in the primary design matrix, not in the footnotes. And because stratification chews up power, build multi-center collaborations and lean on public data to get the sample sizes needed to avoid chasing ghosts. There are limitations worth pausing on. Under-representation of females is not a trivial footnote; it’s a recurrent bias that can dampen signal detection or inflate apparent male-specificity. The androgen literature is noisy and inconsistent, making it hard to draw conclusions about testosterone beyond “context matters.” Animal toxin models are invaluable for mechanism but imperfect proxies for human disease, so the impressive estrogen effects in six-hydroxydopamine or MPTP paradigms are suggestive, not dispositive. And the most tempting mechanistic links—say, from higher PINK-1 messenger RNA in males to actual differences in mitophagy capacity—sit a couple of experimental steps away from direct evidence in human nigral neurons. Let’s stitch the threads. Sex chromosomes and early hormonal organization install different baselines in the dopaminergic system—differences in astrocyte architecture, synaptic glutamate dynamics, and the transcriptional programs of neurons that will live in a high-oxidative, high-calcium environment. As adolescence and early adulthood pass, those programs settle, helped along by sex-specific mosaics from X-inactivation escape and imprinting and, in males, by Y-linked regulation such as SRY’s hand on dopamine biosynthesis. Midlife arrives, and with it a collapse in estradiol in women and a more gradual androgen decline in men. The same polygenic risk, which looks nearly identical across sexes on a Manhattan plot, now funnels through different gene-hormone-chromatin contexts. In the substantia nigra, that translates into distinct patterns of differential expression—hundreds of genes shifting in opposite directions by sex—and sex-skewed epigenetic remodeling. In blood and brain, immune and glial readouts echo those biases. Add in pockets of sex-specific variant associations, like the female-only hits near the LRRK2 gene and the SNCA gene in the Korean dataset, and interactions with reproductive timing, and you have a coherent, layered account of why Parkinson’s looks the way it does across sexes without invoking radically different genetic architectures. One more layer, because the field often asks about lifestyle. Gendered exposures—pesticides, smoking, head injury, even caffeine—do interact with this biology, but in Schaffner’s synthesis they are the frame, not the canvas. The caffeine story is a good example of a three-way conversation: the protective association of caffeine varies with estrogen receptor genotypes, the ESR1 and ESR2 genes, and with postmenopausal hormone status. That doesn’t negate the core hormone-gene mechanisms; it shows how they plug into lived environments. Useful, but peripheral to the central point that sex-linked regulation is the organizing principle here. So where does this leave practice and the next round of studies? First, timing is biology. If estradiol’s neuroprotection is real in humans—and the clinical and mechanistic data point that way—the window matters. The benefit signal shows up when hormone therapy is started early in the disease course, and age at menopause tracks with age at onset. Trials that lump perimenopausal, postmenopausal, and non-menopausal women together will blur effects. Second, the substantia nigra is not the only site of sex-biased regulation, but it is ground zero. Single-cell atlases that resolve dopaminergic subtypes, astrocytes, and microglia by sex, menopausal status, and genotype can turn the current transcriptional and epigenetic associations into causal pathways. Third, gene-hormone-lifestyle interactions need formal tests with prespecified hypotheses—interactions between the ESR1 and ESR2 genes and caffeine is a template, not an outlier. And if you’re building cohorts or pipelines, treat sex as you would a strong prior. Balance enrollment. Capture granular hormonal history—menarche, pregnancies, menopause, hormone therapy. Include sex-stratified analyses in the statistical plan, with realistic power calculations for interaction terms. Make single-nucleus RNA and assay for transposase-accessible chromatin sequencing designs explicitly sex-aware. And if resources force a choice, prioritize sampling at the life-course edges where the levers move—adolescence and midlife transitions—because that’s where gene regulation is likely to tip. The closing thought is simple. Similar heritability does not mean interchangeable biology. Men and women carry largely the same polygenic risk for Parkinson’s disease, but sex chromosomes, hormone dynamics, and sex-biased regulation recalibrate how that risk is expressed, especially in the substantia nigra. That recalibration shows up as higher SNCA and PINK-1 messenger RNA in male nigral neurons, as broader maturation programs in females, as hundreds of differentially expressed genes that flip direction by sex, as methylation patterns that remodel more extensively in women, and as a handful of sex-specific variant associations tightly linked to reproductive timing. It also shows up in the clinic—later onset with later menopause, lower symptom severity with early estrogen therapy—and in the immune system, where women with Parkinson’s show blood signatures tuned to B cell signaling. None of this requires inventing a sex-specific genetic architecture. It asks us to read the same genome through a sex-specific regulatory lens, and to design our experiments—and eventually our interventions—accordingly.

Schaffner and colleagues frame it as three interlocking levers: chromosomal and early developmental programming, gonadal hormones with age-dependent effects, and sex-biased transcriptomic and epigenomic states, especially in the substantia nigra. Those levers engage at specific life-course inflection points, such as conception, adolescence, and midlife, so the same risk can play out differently across time in a male versus a female brain. That’s the through-line.

And an important caveat rides alongside it: females are under-represented in a lot of the underlying datasets, which limits how fine-grained we can be about mechanisms.

Start with the earliest programming. Weeks after conception, the Y chromosome’s SRY gene switches on in embryos destined to be male, steering testes formation and testosterone production. Then a twist you already know from developmental neuroendocrinology: testosterone can be locally converted to estradiol in the limbic system.

That aromatization shapes astrocyte morphology, enhances glutamate release, and biases cell survival programs. In other words, even before birth, the circuitry that will one day carry dopamine is experiencing sex-dependent trophic and synaptic signals. Those organizational effects lay down the responsiveness to the hormone surges of adolescence. They don’t decide Parkinson’s outcomes, but they set the slope of the hill.

Now, layer in X-dose biology. In females, one X chromosome is inactivated in each cell, but that inactivation is random, so you get mosaicism—one patch of cells with one X active, another patch with the other. Some genes escape that inactivation and are expressed from both X chromosomes, so their dosage is genuinely higher in females.

Add genomic imprinting, which is parent-of-origin specific expression, and you’ve created a heterogeneous, sex-linked regulatory landscape before any environmental exposure comes into play. This is not abstract. In the substantia nigra, studies consistently report sex-specific expression of Parkinson’s-relevant genes.

The ARMCX2 gene, an X-linked regulator of mitochondrial trafficking, tends to be upregulated in females but downregulated in males. The IFITM2 gene goes the other way—down in females, up in males. The DYNC1LI1 and REEP1 genes skew down in males.

These are some of the bricks; the wall is a sex-biased transcriptome in dopaminergic territory that’s already set up to respond differently to stress.

There is a Y-linked piece, too. SRY itself regulates enzymes in dopamine biosynthesis in models, and its expression increases after six-hydroxydopamine injury in male animals. That’s a hint—not proof—that Y-driven regulation could shift how male dopaminergic neurons handle insults.

Pair that with robust observations that males show higher messenger RNA for the SNCA gene and the PINK-1 gene in substantia nigra tissue and isolated dopaminergic neurons. Females, in contrast, show more expression of neuronal maturation programs. You can hear the implication: the baseline homeostasis of synuclein handling and mitochondrial quality control may be tuned differently by sex from the outset.

And then hormones arrive like a periodic forcing function. The most reproducible biology here is estrogen’s neuroprotection. Women with later menopause tend to have later Parkinson’s onset, and that roughly ten-fold drop in estradiol during the menopausal transition tracks with increased susceptibility.

In epidemiology, postmenopausal hormone therapy is associated with reduced Parkinson’s risk. In clinical cohorts, estrogen therapy introduced early in the disease course associates with lower symptom severity on the Unified Parkinson’s Disease Rating Scale. Mechanistically, the cellular pathways line up: estrogen activates the mitogen-activated protein kinase and extracellular signal-regulated kinase cascade and the phosphoinositide three-kinase and Akt pathway, both of which push dopaminergic neurons away from apoptosis and protect against glutamate and oxidative hits in culture.

In rodents, females and estrogen-treated males lose fewer dopaminergic neurons to toxins, with shifts toward anti-apoptotic protein profiles. The direction is consistent across models. The magnitude depends on timing and context.

And what about androgens? That picture is muddy. Some studies report lower testosterone in men with Parkinson’s disease, but testosterone replacement yields inconsistent motor and non-motor outcomes.

In animal models, testosterone’s effects are mixed and can flip sign depending on conversion to estradiol. The signal is clearly weaker and less coherent than the estrogen story, which is why Schaffner and colleagues keep coming back to estradiol decline at midlife as an inflection point for risk.

Adolescence matters too. The organization-activation hypothesis fits what’s seen here: organizational effects laid down prenatally and in early life meet the activating surges of puberty and peripuberty to install durable sex differences in circuitry and transcription. That’s a setup for the midlife fall in gonadal hormones to land differently across sexes.

Remember, many prodromal Parkinson’s features—things like constipation, rapid eye movement sleep behavior disorder, hyposmia—show up years before diagnosis, often straddling midlife. The timing aligns with a hormone-gene conversation that’s gone from whisper to shout.

Now, zoom into the substantia nigra’s molecular readouts. If there’s a hub for sex-biased regulation in Parkinson’s biology, this is it. Across region-spanning meta-analyses, two hundred thirty-seven genes in substantia nigra are differentially expressed in sex- and disease-specific patterns, with seventy-five relatively increased in males and one hundred sixty-two increased in females when you compare Parkinson’s to control.

Fifteen genes keep their sex- and disease-specific expression across substantia nigra, striatum, and frontal cortex. That’s a small but instructive set that probably includes nodal regulators. Within the substantia nigra, nine chromosomal segments show differential expression when you compare male controls to male Parkinson’s and female controls to female Parkinson’s.

Two segments—Yq11.1 and 5q34—are common to both comparisons, a reminder that some chromosomal architecture shifts cut across sex while others don’t.

Single-cell and cell-type-enriched analyses sharpen the contrast. In substantia nigra tissue, thirty-six genes are flagged as female-specific and five hundred thirty-nine as male-specific at a false discovery rate under five percent, with thirty-seven described as sex-dimorphic. When you collect dopaminergic neurons specifically and compare Parkinson’s to control within sex, eighty-six genes come up as differentially expressed across male and female analyses, and five of those also differ between male and female controls.

On an X-chromosome-enriched array—the X3P platform—one hundred twenty genes show sex-specific expression regardless of Parkinson’s status. Stratify by sex and disease, and the differential expression balloons: two hundred eighty-eight genes in female Parkinson’s versus female control, two hundred ninety-two in male Parkinson’s versus male control. The pattern is not a sprinkling; it’s a wholesale redrawing of transcriptional state space by sex within a vulnerable circuit.

Epigenetics tracks with that. In cortex neurons from Parkinson’s brains, sex- and disease-specific CpG methylation shifts involve targets you’d expect—the PARK7, SLC17A6, PTPRN2, and NR4A2 genes among others. Meta-analyses suggest females show more pronounced changes in histone and chromatin-modifying gene expression, which could amplify transcriptomic divergence over time.

In blood, there are Parkinson’s- and sex-specific methylation regions too, and the variance explained in female analyses increases when genotype is included, hinting at gene-epigenome coupling that is sex contingent. Not every tissue cooperates: platelet assays of mitochondrial DNA methylation at the MT-TL1, MT-CO2, and MT-CO3 loci show no Parkinson’s- or sex-associated differences at those loci. That negative control is useful; it tells you where not to look first.

Peripheral immune readouts echo a sex bias. In blood, about one hundred fifteen genes distinguish female Parkinson’s patients from female controls, enriched for immune functions and B cell signaling. You could take that two ways: immune signatures mirror what’s going on in the brain, or immune cells themselves live in sex-biased regulatory states that interact with the disease milieu.

Both are plausible. What the data do say is that immune and glial programs are not operating against a neutral background; they’re riding sex-specific baselines sculpted by hormones and chromatin.

Let’s pull genetics back into the frame. At the level of common variant risk, men and women look remarkably similar. Across large genome-wide association studies, heritability estimates are closely matched—about zero point two one in males versus zero point one nine in females—and the genetic correlation between male and female summary statistics is high, roughly zero point eight seven seven.

That means the bulk of shared polygenic signal is the same. But “similar” doesn’t mean “identical.” In a Korean dataset built on the Korean Chip, five loci cross the association threshold in female-only analysis—one near the LRRK2 gene, four near the SNCA gene—with a p-value below one point zero three times ten to the minus seven. Male-only analyses in that cohort yield no hits.

There are also reports of sex-by-locus interactions with reproductive timing, such as age at menarche and age at menopause, nudging Parkinson’s associations at several sites. The takeaway is pragmatic: don’t expect an entirely different architecture by sex, but do expect pockets where sex and reproductive history shape penetrance or expression.

You can already see how this meets mitochondria and proteostasis. If males run higher PINK-1 transcription in substantia nigra, and if females emphasize maturation gene programs, you’re looking at sex-skewed set points for mitophagy and differentiation. No one has shown sex-differential mitophagy flux in human dopaminergic neurons in vivo.

That measurement is still out of reach. But the messenger RNA differences, the estrogen-tilted redox protection via mitogen-activated protein kinase and extracellular signal-regulated kinase and phosphoinositide three-kinase and Akt, and the catalog of sex-biased autophagy and lysosomal genes together argue for regulatory, not hard-coded, dimorphism in mitochondrial maintenance and protein handling. That matches the clinical pattern: broadly similar genetic risk, different trajectories and modifiability across the life course.

The immune and glial axis slots in here too. Estrogen doesn’t just touch neurons; it shapes astrocyte and microglial states. Early-life aromatization to estradiol molds astrocyte morphology and glutamatergic tone.

Later in life, estradiol supports anti-inflammatory phenotypes and antioxidant defenses. Schaffner and colleagues synthesize this into a convergent view: sex-biased transcriptional and epigenetic set points, modulated by hormone status, tilt inflammatory cascades, antigen presentation capacity, and oxidative stress handling in substantia nigra. That tilt is not necessarily protection for females across the board—recall the more extensive methylation remodeling reported in females—but it is a different starting line and a different trajectory under load.

Now, a reality check about scale and context. Parkinson’s remains more prevalent in men—roughly one point four times as many men as women carried the diagnosis between nineteen ninety and twenty sixteen—although more recent analyses suggest the gap might be narrowing. Globally, prevalence rose dramatically from nineteen ninety to twenty nineteen, and projections point to tens of millions affected by twenty fifty.

Those numbers set the public health stage, but they don’t overturn the molecular message: sex is not a covariate to be neutralized; it’s a biological system organizing risk. And this biology is hierarchical—chromosomes to hormones to chromatin to transcripts—and contingent on timing.

You’ll notice how consistently the findings converge on a few methodological imperatives. Balance the cohorts. Don’t treat sex as a nuisance variable to be adjusted away.

Use covariate adjustment when that’s all the data allow, but, where powered, run sex-stratified models and explicit sex-by-genotype or sex-by-hormone interaction terms. Midlife hormonal states matter—menopausal status, hormone replacement therapy, contraceptive use, parity—so collect them. In transcriptomic and epigenomic designs, put sex in the primary design matrix, not in the footnotes.

And because stratification chews up power, build multi-center collaborations and lean on public data to get the sample sizes needed to avoid chasing ghosts.

There are limitations worth pausing on. Under-representation of females is not a trivial footnote; it’s a recurrent bias that can dampen signal detection or inflate apparent male-specificity. The androgen literature is noisy and inconsistent, making it hard to draw conclusions about testosterone beyond “context matters.” Animal toxin models are invaluable for mechanism but imperfect proxies for human disease, so the impressive estrogen effects in six-hydroxydopamine or MPTP paradigms are suggestive, not dispositive.

And the most tempting mechanistic links—say, from higher PINK-1 messenger RNA in males to actual differences in mitophagy capacity—sit a couple of experimental steps away from direct evidence in human nigral neurons.

Let’s stitch the threads. Sex chromosomes and early hormonal organization install different baselines in the dopaminergic system—differences in astrocyte architecture, synaptic glutamate dynamics, and the transcriptional programs of neurons that will live in a high-oxidative, high-calcium environment. As adolescence and early adulthood pass, those programs settle, helped along by sex-specific mosaics from X-inactivation escape and imprinting and, in males, by Y-linked regulation such as SRY’s hand on dopamine biosynthesis.

Midlife arrives, and with it a collapse in estradiol in women and a more gradual androgen decline in men. The same polygenic risk, which looks nearly identical across sexes on a Manhattan plot, now funnels through different gene-hormone-chromatin contexts. In the substantia nigra, that translates into distinct patterns of differential expression—hundreds of genes shifting in opposite directions by sex—and sex-skewed epigenetic remodeling.

In blood and brain, immune and glial readouts echo those biases. Add in pockets of sex-specific variant associations, like the female-only hits near the LRRK2 gene and the SNCA gene in the Korean dataset, and interactions with reproductive timing, and you have a coherent, layered account of why Parkinson’s looks the way it does across sexes without invoking radically different genetic architectures.

One more layer, because the field often asks about lifestyle. Gendered exposures—pesticides, smoking, head injury, even caffeine—do interact with this biology, but in Schaffner’s synthesis they are the frame, not the canvas. The caffeine story is a good example of a three-way conversation: the protective association of caffeine varies with estrogen receptor genotypes, the ESR1 and ESR2 genes, and with postmenopausal hormone status.

That doesn’t negate the core hormone-gene mechanisms; it shows how they plug into lived environments. Useful, but peripheral to the central point that sex-linked regulation is the organizing principle here.

So where does this leave practice and the next round of studies? First, timing is biology. If estradiol’s neuroprotection is real in humans—and the clinical and mechanistic data point that way—the window matters.

The benefit signal shows up when hormone therapy is started early in the disease course, and age at menopause tracks with age at onset. Trials that lump perimenopausal, postmenopausal, and non-menopausal women together will blur effects. Second, the substantia nigra is not the only site of sex-biased regulation, but it is ground zero.

Single-cell atlases that resolve dopaminergic subtypes, astrocytes, and microglia by sex, menopausal status, and genotype can turn the current transcriptional and epigenetic associations into causal pathways. Third, gene-hormone-lifestyle interactions need formal tests with prespecified hypotheses—interactions between the ESR1 and ESR2 genes and caffeine is a template, not an outlier.

And if you’re building cohorts or pipelines, treat sex as you would a strong prior. Balance enrollment. Capture granular hormonal history—menarche, pregnancies, menopause, hormone therapy.

Include sex-stratified analyses in the statistical plan, with realistic power calculations for interaction terms. Make single-nucleus RNA and assay for transposase-accessible chromatin sequencing designs explicitly sex-aware. And if resources force a choice, prioritize sampling at the life-course edges where the levers move—adolescence and midlife transitions—because that’s where gene regulation is likely to tip.

The closing thought is simple. Similar heritability does not mean interchangeable biology. Men and women carry largely the same polygenic risk for Parkinson’s disease, but sex chromosomes, hormone dynamics, and sex-biased regulation recalibrate how that risk is expressed, especially in the substantia nigra.

That recalibration shows up as higher SNCA and PINK-1 messenger RNA in male nigral neurons, as broader maturation programs in females, as hundreds of differentially expressed genes that flip direction by sex, as methylation patterns that remodel more extensively in women, and as a handful of sex-specific variant associations tightly linked to reproductive timing. It also shows up in the clinic—later onset with later menopause, lower symptom severity with early estrogen therapy—and in the immune system, where women with Parkinson’s show blood signatures tuned to B cell signaling. None of this requires inventing a sex-specific genetic architecture.

It asks us to read the same genome through a sex-specific regulatory lens, and to design our experiments—and eventually our interventions—accordingly.