September Neurochemistry in the Southern Hemisphere · vicente.md
vicente.md/2026-002 [q-bio.NC] · 11 Sept 2026
Preprint · not peer reviewed 8 pages
September Neurochemistry in the Southern Hemisphere
Vicente González B.†
Engineer and Computational Neuroscience Researcher · Santiago, Chile
† correspondence · hi@vicente.md · no institutional funding, no competing interests
Abstract
September marks the beginning of spring in the Southern Hemisphere, when increasing daylight can influence sleep, mood, and brain chemistry. Human studies show seasonal changes in melatonin secretion and in systems involving serotonin, dopamine, and natural opioids. These findings describe different biological processes, rather than a uniform rise in brain activity. Some people experience relief from winter depression, while susceptible individuals may develop manic episodes. Most chemical measurements come from Northern Hemisphere studies comparing winter and summer, so their relevance to September in the south remains uncertain. This narrative synthesis separates measured changes from proposed mechanisms, ranks the evidence without pretending to rank causal impact, and asks what a Southern Hemisphere study would need to measure. The central constraint is that receptor binding, neurotransmitter release, and brain activation are different quantities: a change in one does not, by itself, identify a change in the others.
September opens spring in the Southern Hemisphere. It is a useful moment to ask how increasing daylight interacts with the brain, but a poor substitute for measuring that interaction. The relevant variables include day length, the timing and intensity of light reaching the eyes, sleep timing, and individual susceptibility. A calendar month bundles these exposures together.
The human literature supports seasonal variation in biological timing and several molecular imaging measures. It does not identify a single chemical surge that explains spring. For computational neuroscience, the central problem is identifiability: which hidden biological changes can be recovered from the measurements available?
A seasonal difference in a PET signal is evidence about a molecular measurement. Establishing a change in synaptic transmission, and then a cause of psychiatric symptoms, requires additional observations.
This article is a focused narrative synthesis, not a systematic review or meta-analysis. It prioritizes human longitudinal studies, molecular imaging, an Australian serotonin study, and Southern Hemisphere clinical observations. The ranking below is an editorial ordering of evidence strength and clinical relevance. Causal psychiatric impact cannot be ranked from these studies. No new participant-level data are analyzed; calculated contrasts are identified explicitly.
Clinical seasonality in the Southern Hemisphere
In New South Wales, a study covering more than 20,000 patients admitted for mania over 15 years found a spring peak. Admissions began rising after day length started increasing in winter. Models incorporating photoperiod and its rate of change explained approximately 20% of admission variance across the year. That statistic is neither a 20% rise in personal risk nor evidence that daylight caused 20% of episodes. Diagnoses, geographic exposure estimates, admission delays, and unmeasured social factors limit interpretation.1
Clinically, manic episodes can involve reduced need for sleep, increased activity, rapid speech, racing thoughts, and elevated or irritable mood.2 Winter-pattern seasonal affective disorder (SAD) often remits in spring; the less common summer pattern may instead involve insomnia, poor appetite, anxiety, and agitation.3
Historical Santiago-region data also found more suicide deaths during October–January, with a December maximum. This is a population association, not a September-specific effect or a demonstrated molecular pathway.4 Clinical seasonality establishes a phenomenon worth explaining. It does not determine which neurotransmitter moved, in which circuit, or in which direction.
Chile's September clock change
Chile adds an abrupt social-time shift to the gradual spring photoperiod change. In 2026, Santiago and most mainland regions advanced from UTC−4 to UTC−3 at midnight ending September 5. Aysén and Magallanes remain at UTC−3 year-round; Rapa Nui and Salas y Gómez advanced at 22:00 local time to UTC−5.5The clock advance adds no daylight: it moves fixed work and school schedules one hour earlier relative to solar time.
The plausible contribution is a mismatch between biological timing and scheduled sleep. Keeping the same clock-time alarm requires earlier waking while circadian adjustment may lag. A European study tracking 50 people around both transitions found incomplete spring adjustment of activity timing, especially in late chronotypes.6 In a US study of 35 adolescents completing the protocol, actigraphy showed 32 minutes less sleep per school night in the week after the spring advance (p=0.001), together with greater sleepiness and poorer vigilance.7 These estimates are cohort-specific, not measurements in Chile.
Sleep disruption may matter particularly in susceptible people with bipolar disorder: 20% of 3,140 participants retrospectively reported sleep loss as a trigger of mania or hypomania.8 That proportion is not a risk estimate for the clock change. A Finnish register study covering 1987–2003 found no significant increase in hospital-treated mania in the two weeks after the spring transition (p=0.480).9A contribution through sleep and circadian disruption is plausible; a causal explanation of Chile's spring psychiatric pattern remains unproven. These studies do not demonstrate a Chile-specific clock-change effect on 5-HT, dopamine, or melatonin concentration.
For modeling, treat the transition as a dated change in social schedules alongside continuous photoperiod, and measure sleep and personal light exposure around it. Aysén and Magallanes offer possible comparison regions, but their latitude and seasonal light profiles require adjustment; they are not automatically equivalent controls.
Disorder sensitivity to seasonal change
Seasonal mood disorders have the clearest clinical signal; neurological conditions require separate, diagnosis-specific interpretation. “Sensitivity” here means a tendency for symptoms or episodes to vary with season. It can include improvement: winter-pattern depression commonly recedes as spring arrives.
The tiers below summarize confidence in clinically relevant seasonality and its applicability to the spring transition. They are an editorial evidence map, not a validated scale of biological sensitivity, illness severity, or individual September risk. Diagnoses within a tier are not ordered. Admission counts, symptom scores, and relapse timing have different denominators and cannot establish a universal “most affected” ranking.
Tier key: A — clearest seasonal mood patterns; B — condition-specific neurological evidence; C — limited support for a general spring inference.
Tier
Disorder or phenotype
Seasonal response and evidence boundary
A
Bipolar disorder (BD), especially mania
Manic admissions peak in Australian spring.1 In Finland, summer manic admissions had an observed/expected ratio O/E=1.10 (95% CI 1.06–1.13).10 This concerns episode timing and service use; it does not mean every person with BD worsens in spring.
A
Depression with a seasonal pattern (SAD)
Winter-pattern depressive symptoms usually decrease in spring; the less common summer pattern can worsen in spring/summer.3 The seasonal subtype is selected for recurrent seasonality and must be distinguished from depression overall.
B
Multiple sclerosis (MS), relapses
A registry analysis of 9,811 people and 32,762 relapses found spring-associated relapse timing in both hemispheres.11 This is evidence about inflammatory disease activity; it does not establish a seasonal monoamine mechanism.
B
Cluster headache (CH), bouts
In a Korean cohort, 88/200 patients reported seasonal propensity; spring was the most frequent season among those 88 (37.5%).12 The result describes a subgroup and a northern setting, not a 37.5% spring risk increase or a prediction for Chile.
Here O/E denotes observed admissions divided by the study's expected count, not a patient's probability of an episode. The Finnish estimates come from one national dataset covering 1987–2017; its large sample does not remove geographic or admission-selection effects.10
Timing remains model-dependent. In the MS study, raw southern monthly counts peaked in November, while a pooled harmonic model estimated September 5 (95% CI August 10–September 26). Those are different summaries of the data, not interchangeable dates for an individual's relapse.11 This distinction is particularly relevant to computational models that compress a seasonal curve into one peak.
These clinical tiers cannot be mapped directly onto the neurochemical ranking below. The cited studies do not identify a spring change in dopamine as the cause of ADHD symptoms, or a transmitter-mediated pathway from season to MS relapses. A lower tier means less support for a general spring inference; it does not exclude a strong seasonal pattern in a particular person.
Ranking the human evidence
Unless specified otherwise, directions below compare winter or autumn/winter with summer or spring/summer. They must not be read as measurements of September in Chile. Molecular imaging evidence is largely northern; hemispheric transfer is a hypothesis about comparable exposures, not a simple relabeling of months.
Rank
System and observable
Defensible finding
Main inference boundary
1
Melatonin secretion duration
Shorter summer secretion in a SAD cohort; controls showed no significant change
Duration is not concentration or a demonstrated mania mechanism
2
Serotonin turnover; SERT binding
Australian turnover estimate covaries with sunshine; transporter seasonality is cohort-dependent
Neither gives a universal sign for synaptic 5-HT
3
Dopamine synthesis capacity; D₂/₃ binding
Lower putamen FDOPA influx in spring/summer; lower receptor availability with longer days
Synthesis and binding do not identify net dopamine release
4
MAO-A distribution volume
Seasonal decrease in healthy controls, significant across pooled regions
Enzyme-related PET signal is not measured monoamine catabolic flux
5
μ-opioid receptor binding
Nonlinear relationship with day length; spring peak in Finnish data
Receptor availability is not endogenous opioid release
Melatonin: the biological night can change
Wehr and colleagues measured melatonin over 24 hours in both seasons. In 55 people with SAD, secretion duration declined from 9.0±1.3 to 8.4±1.3h (mean ± SD; p=0.001). In 55 controls it was 9.0±1.6 versus 8.9±1.2h, with no significant change.16
Figure 1. Melatonin secretion duration. Group means ±1 SD; each cohort was assessed in both seasons. Contrasts do not describe every individual's trajectory. Redrawn from summary statistics reported by Wehr et al. (2001).
This is evidence for a seasonal change in the duration of a hormonal signal, not a universal decrease in melatonin concentration. Separately, experimental exposure to natural light–dark cycles showed a longer biological night in winter than summer; seasonal differences were less evident under modern electrical lighting. That result makes personal light exposure important to the model, rather than treating season as an adequate exposure measurement.17
Serotonin: turnover and transport are different measurements
The most directly relevant southern molecular evidence comes from Melbourne. Lambert and colleagues sampled blood from the internal jugular veins of 101 healthy men to estimate cerebral serotonin turnover. Turnover was lowest in winter; the estimate correlated with bright-sunlight duration (r=0.294, p=0.010). It was an inference from a metabolite in blood draining the brain, not a direct assay of synaptic serotonin or local neuronal firing.18
In a Canadian cross-sectional PET study of 88 healthy participants, serotonin transporter (SERT) binding was lower in spring/summer across the prefrontal cortex, anterior cingulate cortex, caudate, putamen, thalamus, and midbrain. Caudate binding potential averaged 0.93 versus 1.06 in autumn/winter: approximately 12% lower when calculated relative to the autumn/winter mean.19
That direction is not universal. A longitudinal study of 23 people resilient to SAD found lower raphe and global SERT binding in winter.20 A separate longitudinal study linked seasonal transporter regulation to depressive symptom severity in SAD.21 The useful conclusion is that serotonin regulation varies with season and phenotype. The evidence does not license a population-wide “serotonin rises in September” statement.
Dopamine: synthesis capacity and receptor availability
Eisenberg and colleagues studied 86 healthy adults using [18F]FDOPA PET. Mean bilateral putamen influx, Ki, was 0.0073 in autumn/winter and 0.0070 in spring/summer (p=0.038). Calculated from those rounded means, the latter is approximately 4.1% lower. The caudate showed no significant seasonal effect. This was a cross-sectional synthesis/storage-related tracer measurement, not a measurement of dopamine release.22
Sun and colleagues subsequently analyzed 291 baseline [11C]raclopride scans. Their primary analysis included 227 participants younger than 40. In the left caudate, a one-SD increase in day length—4.26 hours—was associated with a reported 2.8% decrease in D2/3 receptor availability. Day length included civil twilight. Age, sex, and scanner effects were modeled; whole-sample analyses were supplementary.23
The denominator matters: the 2.8% estimate should be attached to the primary analysis, not presented as a September change in 291 people. These studies support seasonal variation in dopamine-related observables. They do not establish a common direction of dopaminergic signaling across circuits.
MAO-A: enzyme-related signal is not metabolic flux
In a longitudinal [11C]harmine PET study, 27 healthy controls showed lower monoamine oxidase A (MAO-A) distribution volume, VT, in spring/summer. A decrease was not significant in 24 participants with SAD. Seasonal effects were significant when regions were evaluated together, not in individual regions separately. Crucially, significance in controls and non-significance in patients does not itself establish a between-group difference; the study did not find a significant group difference in the seasonal contrast.24
MAO-A participates in monoamine catabolism. A lower enzyme-related PET signal may suggest altered breakdown capacity, but it does not quantify reaction flux, substrate availability, or a resulting increase in extracellular serotonin or dopamine. This is a plausible regulatory component with an incompletely identified functional consequence.
μ-opioid receptors: a nonlinear seasonal association
A Finnish dataset of 204 healthy participants, scanned with [11C]carfentanil, showed an inverted-U relationship between day length and μ-opioid receptor availability. Binding peaked in spring, with associations spanning cingulate, orbitofrontal, insular, and other cortical regions. An accompanying rat experiment supported a causal effect of day-length manipulation on receptor availability.25
The human evidence remained observational. It did not establish that spring increases endorphin release, pleasure, or sociability. The high-latitude setting and nonlinear exposure relationship also make a direct calendar translation to Santiago particularly uncertain.
Separating latent states from observations
A minimal description of the inference problem separates a latent biological state, x(t), from each modality's observation function:
ym(t)=hm(x(t),θm)+εm(t).
Here m indexes modalities, θm their measurement parameters, and εm their errors. This is a conceptual formulation, not a fitted model. A seasonal change in ym need not identify a unique change in x.
For reversible receptor radioligands, the standard binding-potential relation makes the ambiguity concrete:26
BPND=fNDKDBavail.
Bavail denotes binding sites available to the tracer, KD its equilibrium dissociation constant, and fND the free fraction in nondisplaceable tissue. Binding potential is a composite quantity. An observed decrease cannot, on its own, distinguish altered receptor availability, affinity, or other contributing factors. For tracers sensitive to endogenous competition, occupancy is another possible contributor. These alternatives require additional constraints or measurements.
Figure 2. Three observation domains. Molecular PET measures tracer-dependent binding or kinetics; synaptic concentration is a latent target in these seasonal studies; BOLD measures a hemodynamic response. The separation denotes different quantities and observation models, not an absence of biological coupling. Conceptual schematic; no fitted effects are shown.
For the same reason, a percentage change in binding cannot be compared with a percentage change in secretion duration to rank psychiatric impact. The measurements have different units, biological meanings, and exposure contrasts. A numerical leaderboard would hide these differences.
BOLD seasonality is task-dependent
In a Belgian fMRI study of 28 healthy participants, sustained-attention responses peaked near the summer solstice in regions including the thalamus, amygdala, and frontal cortex. Working-memory responses instead peaked near autumn and reached their minimum near spring, involving prefrontal, insular, and thalamic regions. Behavioral performance was stable across seasons.27
Lower spring BOLD in one task therefore did not demonstrate worse cognition. Nor did higher summer BOLD in another establish a global increase in neural activity. Task demands and the hemodynamic observation model belong in the interpretation.
What a Southern Hemisphere study should estimate
The next useful experiment would observe the same people across the transition, measure actual light exposure and sleep, and prespecify molecular and clinical endpoints. “September” would become a time coordinate; it would not stand in for a mechanism.
A candidate longitudinal association model could distinguish day length from its rate of change:
yit=αi+βLLit+βΔLL˙it+γ⊤zit+εit.
αi captures participant-specific baseline differences. Lit is a prespecified photoperiod measure; L˙it is its change per day. The covariate vector zit could include personal light exposure, sleep phase and duration, scan time, medication, temperature, and scanner variables. Exposure variables must be defined before analysis; strongly correlated candidates should not all be included without addressing identifiability.
This equation is a study-design proposal, not an established biological law. A linear term may be inadequate for a nonlinear receptor association; lagged responses and interactions with susceptibility may matter. Model comparison and out-of-sample evaluation should decide whether the added structure improves prediction. Establishing mediation from light through a molecular state to symptoms would require temporal ordering and stronger causal assumptions than a seasonal correlation provides.
The main outcome to seek is a reproducible within-person response to measured exposure, with uncertainty. Generalization should then be tested across latitude, hemisphere, and clinical phenotype. Merely shifting a northern peak by six months would leave the central question unanswered.
Interpretation
Human seasonal neurobiology is measurable. The strongest practical conclusions concern changing biological timing, phenotype-dependent serotonin regulation, and seasonal differences in several PET observables. The evidence for a single neurochemical explanation of spring psychiatric change is much weaker.
For an individual, a change in sleep or mood cannot be reverse-engineered into a serotonin, dopamine, or opioid level from the season alone. For a modeler, the productive question is narrower: which exposure changes which observable, through which identifiable mechanism, in which people?
Supplementary material
Download the compact 9:16 evidence card. It summarizes the same measurement boundaries; the article supplies the methodological context and source references.
References
Footnotes
Parker, G., Hadzi-Pavlovic, D., Bayes, A., & Graham, R. (2018). Relationship between photoperiod and hospital admissions for mania in New South Wales, Australia. Journal of Affective Disorders, 226, 72–76. doi:10.1016/j.jad.2017.09.014. ↩↩2
National Institute of Mental Health. Bipolar Disorder. Clinical background; accessed September 2026. ↩
Retamal, P., & Humphreys, D. (1998). Occurrence of suicide and seasonal variation. Revista de Saúde Pública, 32, 408–412. doi:10.1590/S0034-89101998000500002. ↩
Armada de Chile, DIRECTEMAR/SHOA (2026). Cambio de hora en Chile Continental e Insular Occidental. Official notice implementing Decreto Supremo N.º 98 (July 2, 2026). September 5, 2026 transition and regional exceptions. ↩
Kantermann, T., Juda, M., Merrow, M., & Roenneberg, T. (2007). The human circadian clock's seasonal adjustment is disrupted by daylight saving time. Current Biology, 17, 1996–2000. doi:10.1016/j.cub.2007.10.025. ↩
Medina, D., Ebben, M., Milrad, S., Atkinson, B., & Krieger, A. C. (2015). Adverse effects of daylight saving time on adolescents' sleep and vigilance. Journal of Clinical Sleep Medicine, 11, 879–884. doi:10.5664/jcsm.4938. ↩
Lewis, K. S., et al. (2017). Sleep loss as a trigger of mood episodes in bipolar disorder: individual differences based on diagnostic subtype and gender. The British Journal of Psychiatry, 211, 169–174. doi:10.1192/bjp.bp.117.202259. ↩
Lahti, T. A., Haukka, J., Lönnqvist, J., & Partonen, T. (2008). Daylight saving time transitions and hospital treatments due to accidents or manic episodes. BMC Public Health, 8, 74. doi:10.1186/1471-2458-8-74. ↩
Törmälehto, S., et al. (2022). Seasonal effects on hospitalizations due to mood and psychotic disorders: a nationwide 31-year register study. Clinical Epidemiology, 14, 1177–1191. doi:10.2147/CLEP.S372341. ↩↩2↩3↩4
Spelman, T., Gray, O., Lucas, R., & Butzkueven, H. (2015). A method of trigonometric modelling of seasonal variation demonstrated with multiple sclerosis relapse data. Journal of Visualized Experiments, 106, e53169. doi:10.3791/53169. ↩↩2
Moon, H.-S., et al. (2017). Clinical features of cluster headache patients in Korea. Journal of Korean Medical Science, 32, 502–506. doi:10.3346/jkms.2017.32.3.502. ↩
Traffanstedt, M. K., Mehta, S., & LoBello, S. G. (2016). Major depression with seasonal variation: is it a valid construct? Clinical Psychological Science, 4, 825–834. doi:10.1177/2167702615615867. Cross-sectional population findings do not directly test recurrence within diagnosed seasonal cases. ↩
Vogel, S. W. N., et al. (2019; first published online 2016). Seasonal variations in the severity of ADHD symptoms in the Dutch general population. Journal of Attention Disorders, 23, 924–930. doi:10.1177/1087054716649663. ↩
Hammen, T., et al. (2024). The influence of climatic factors on the provocation of epileptic seizures. Journal of Clinical Medicine, 13, 3404. doi:10.3390/jcm13123404. ↩
Wehr, T. A., et al. (2001). A circadian signal of change of season in patients with seasonal affective disorder. Archives of General Psychiatry, 58, 1108–1114. doi:10.1001/archpsyc.58.12.1108. ↩
Stothard, E. R., et al. (2017). Circadian entrainment to the natural light-dark cycle across seasons and the weekend. Current Biology, 27, 508–513. doi:10.1016/j.cub.2016.12.041. ↩
Lambert, G. W., Reid, C., Kaye, D. M., Jennings, G. L., & Esler, M. D. (2002). Effect of sunlight and season on serotonin turnover in the brain. The Lancet, 360, 1840–1842. doi:10.1016/S0140-6736(02)11737-5. ↩
Praschak-Rieder, N., et al. (2008). Seasonal variation in human brain serotonin transporter binding. Archives of General Psychiatry, 65, 1072–1078. doi:10.1001/archpsyc.65.9.1072. ↩
Mc Mahon, B., et al. (2018). Seasonality-resilient individuals downregulate their cerebral 5-HT transporter binding in winter: a longitudinal combined ¹¹C-DASB and ¹¹C-SB207145 PET study. European Neuropsychopharmacology, 28, 1151–1160. doi:10.1016/j.euroneuro.2018.06.004. ↩
Mc Mahon, B., et al. (2016). Seasonal difference in brain serotonin transporter binding predicts symptom severity in patients with seasonal affective disorder. Brain, 139, 1605–1614. doi:10.1093/brain/aww043. ↩
Eisenberg, D. P., et al. (2010). Seasonal effects on human striatal presynaptic dopamine synthesis. Journal of Neuroscience, 30, 14691–14694. doi:10.1523/JNEUROSCI.1953-10.2010. ↩
Sun, L., et al. (2024). Seasonal variation in D₂/₃ dopamine receptor availability in the human brain. European Journal of Nuclear Medicine and Molecular Imaging, 51, 3284–3291. doi:10.1007/s00259-024-06715-9. ↩
Spies, M., et al. (2018). Brain monoamine oxidase A in seasonal affective disorder and treatment with bright light therapy. Translational Psychiatry, 8, 198. doi:10.1038/s41398-018-0227-2. ↩
Sun, L., et al. (2021). Seasonal variation in the brain μ-opioid receptor availability. Journal of Neuroscience, 41, 1265–1273. PubMed PMID: 33361461. ↩
Innis, R. B., et al. (2007). Consensus nomenclature for in vivo imaging of reversibly binding radioligands. Journal of Cerebral Blood Flow & Metabolism, 27, 1533–1539. doi:10.1038/sj.jcbfm.9600493. ↩
Meyer, C., et al. (2016). Seasonality in human cognitive brain responses. Proceedings of the National Academy of Sciences, 113, 3066–3071. doi:10.1073/pnas.1518129113. ↩
Cite as: Vicente González B.. “September Neurochemistry in the Southern Hemisphere.” vicente.md/2026-002, 11 Sept 2026.← Archive
C
Major depression without an established seasonal pattern
Findings depend on the endpoint: Finnish unipolar-depression admissions had a small spring excess (O/E=1.03; 95% CI 1.01–1.04),10 whereas a US cross-sectional PHQ-8 survey found no association with season.13 A universal spring direction is not established.
C
Schizophrenia
Finnish admissions were slightly elevated in summer/early autumn (O/E≈1.02) and lower in mid-spring (O/E≈0.98).10 This supports modest seasonality in that setting, not a general spring exacerbation.
C
Attention-deficit/hyperactivity disorder (ADHD)
Among 5,303 Dutch adults, those assessed in spring/summer reported higher total ADHD and inattention scores than those assessed in autumn; hyperactivity was higher in spring.14 This was a comparison between respondents in a general-population sample, not longitudinal worsening in a diagnosed ADHD cohort.
C
Epilepsy
A German study of 9,366 seizure-related admissions found more admissions in autumn/winter than spring/summer (p=0.005).15 Seasonal admissions and weather associations do not establish a general Southern Hemisphere spring increase in seizure susceptibility.
