Personality Pharmacology

Part I · Foundations Draft

Chapter 3

The Polygenic Neurochemical Set-Point

Why do people have characteristic psychological starting points at all?

The intuitive answer is that brains differ. The useful answer requires more precision, because there is no single variable called brain chemistry and no reason to expect a person’s baseline to be reducible to one concentration.

Consider dopamine as an example. Effective dopaminergic signaling depends on precursor availability, synthesis, vesicular packaging, release probability, tonic and phasic firing, autoreceptors, receptor subtype density, receptor sensitivity, transporters, enzymatic metabolism, intracellular signaling, local circuitry, long-range connectivity, hormonal state, prior drug exposure, learning history, sleep, stress, and age. Each component is influenced by many genes and by development. The same is true, with different machinery, for serotonin, noradrenaline, GABA, glutamate, acetylcholine, histamine, endogenous opioids, and the endocannabinoid system.

The result is not a set of neat chemical knobs. It is a regulatory regime assembled from many interacting parts.

The term set-point is useful if handled carefully. It should not be imagined as a thermostat set to exactly 21 degrees. A biological system rarely defends a single scalar value in that way. A better picture is an attractor: a range of states toward which the system tends to return under ordinary conditions. The person may fluctuate around that region with sleep, meals, stress, hormones, social experience, and circadian rhythms while still showing a recognizable long-term pattern.

dotted: what a thermostat would holdstressshort sleepthe range it returns todays →
Figure 3.1An attractor, not a thermostat. A set-point is better pictured as a range the system keeps returning to than as one value it defends.

This helps reconcile stability with variation. A person can be characteristically threat-sensitive without experiencing the same anxiety every day. They can be characteristically reward-driven without being equally motivated in every context. The regulatory system has a preferred operating regime, not a frozen state.

Why should these operating regimes differ between people? Because virtually every determinant differs. Common genetic variants have small effects across thousands of loci. Rare variants can have larger effects. Development changes receptor expression and connectivity. Early experiences shape threat learning and attachment. Chronic stress modifies endocrine and autonomic systems. Repeated drug exposure produces tolerance or sensitization. Inflammation and metabolic disease alter signaling. Sleep patterns influence arousal and reward. Culture and family environments reinforce some behaviors and extinguish others.

Personality can therefore emerge from a feedback loop. A child with a slight biological bias toward novelty may seek more stimulating environments. Those environments provide different learning experiences, social networks, risks, and rewards. The resulting learning then alters future choices, and some of those experiences change the biological system itself. A small initial difference can be amplified without ever being genetically predetermined.

A slight biastoward noveltySeeksstimulationmore stimulatingplaces and peopleLearnsdifferentlynew risksBias growschoices and biologyshiftthe loop amplifies
Figure 3.2A small difference, amplified. A slight bias shapes which environments a child seeks; the environments shape learning; learning shapes later choices and even the biology. No genetic destiny is needed.

This also explains why the same outward personality can hide different mechanisms. Two people may both appear calm. One may have low threat reactivity. The other may have high threat reactivity combined with unusually strong inhibitory control and practiced cognitive regulation. Their behavior converges, but their internal architecture differs. A perturbation that weakens control may barely affect the first person and dramatically destabilize the second.

That possibility is central to personality pharmacology. Surface traits are imperfect summaries of hidden control strategies.

Person Athreat reactivityregulatory controllooks calmstill calmPerson Bthreat reactivityregulatory controllooks calmanxiety breaks throughcontrol weakenedIdentical on a questionnaire, different under challenge
Figure 3.3Same surface, different machinery. Two people can look equally calm for opposite reasons. A perturbation that weakens control barely touches one and unmasks the other.

An early version of the theory placed considerable weight on the central limit theorem. If many small, partly independent causes contribute to a trait, the aggregate often tends toward a roughly normal distribution. This gives an appealing picture of a population spread continuously from low to high on biological dimensions, with many people in the middle and fewer at the tails.

The intuition is useful, but it needs one refinement. The central limit theorem says that many small causes produce a bell curve on the scale on which those causes add up1. Most molecular causes do not add fixed amounts. A genetic variant raises or lowers the expression of a gene by a proportion; a stressor changes enzyme activity by a proportion. Proportional effects add up on a logarithmic scale, so the molecules behind tone should tend towards a log-normal distribution, with a long tail on the high side23. The brain, however, largely reads those molecules on a logarithmic scale itself, since receptors respond to the logarithm of concentration and pathways saturate. Functional tone can therefore come out more symmetric than the molecules that produce it, though not necessarily as a perfect bell curve. Chapter 5 returns to this.

Interactive figure: Many small causes.

Figure 3.4Many small causes. Each simulated person is the product of a number of small, independent influences. Change how many there are and how they combine. When they scale one another, the molecules come out skewed; read the way the brain reads them, the spread becomes roughly symmetric again. Illustrative model, not measured data.

The conclusion must still be held loosely. Biological causes are not perfectly independent. Feedback creates nonlinearities. Sex, age, developmental stage, and disease can create mixtures. Selection can truncate or maintain tails. Measurement can distort distributions. Clinical samples are rarely representative. Some variables may be log-normal, skewed, multimodal, or bounded. The robust claim is not that every neurochemical trait forms a bell curve. It is that many relevant variables should differ quantitatively across people rather than existing only as discrete normal and abnormal types.

Even that claim is difficult to measure. A blood concentration of serotonin is not a measure of serotonergic function in the prefrontal cortex. Cerebrospinal-fluid metabolites integrate multiple sources and can be affected by sampling procedures. PET receptor availability is a proxy, not a complete description of signaling. Magnetic resonance spectroscopy captures broad pools of metabolites, not synaptic transmission. There may never be a single laboratory value called “dopamine tone” that can be placed beside a personality score.

This does not make the construct useless. Many scientifically useful variables are latent. Intelligence, socioeconomic status, inflammation, and even depression are inferred from imperfect measurements. The important question is whether different indicators converge, show within-person stability, distinguish individuals, predict behavior, and respond to perturbation in expected ways.

A rigorous set-point theory would therefore make several predictions. Repeated measurements should reveal some stable between-person variance after accounting for state. Multiple biological proxies should sometimes cluster into latent functional factors. Those factors should relate to computational behaviors such as reward learning, threat sensitivity, effort allocation, or inhibitory control. They should help predict responses to pharmacological challenges. And they should explain something beyond what can already be predicted from ordinary clinical observation.

If they do not, the set-point language should be abandoned or narrowed.

The most important shift is conceptual. We should stop imagining a person’s baseline as the concentration of a transmitter and start imagining it as a characteristic regulatory regime. Such a regime can be partly inherited, partly learned, partly state-dependent, and still stable enough to shape a life.

That is the substrate personality pharmacology needs. The next question is what, exactly, it means for a signaling system to have a characteristic “tone” when so many components can compensate for one another.

Sources

  1. Fisher RA (1918). The correlation between relatives on the supposition of Mendelian inheritance. Transactions of the Royal Society of Edinburgh. doi:10.1017/S0080456800012163
  2. Mohammadi P, Castel SE, Brown AA, et al. (2017). Quantifying the regulatory effect size of cis-acting genetic variation using allelic fold change. Genome Research. doi:10.1101/gr.216747.116
  3. Limpert E, Stahel WA, Abbt M (2001). Log-normal distributions across the sciences: keys and clues. BioScience. doi:10.1641/0006-3568(2001)051[0341:LNDATS]2.0.CO;2