Appearance
22.1 — What a Drug Actually Does
A tablet is a lump of chemical you swallow. It dissolves, crosses into your blood, spreads through your entire body, and somehow ends up relieving a headache and not, say, stopping your heart.
How does it know where to go?
It does not. The drug reaches essentially every tissue you have. What makes it selective is that only some cells carry the molecular target it can bind to.
That single idea — the drug goes everywhere, the target is only in some places — explains both how medicines work and why side effects exist.
The lock and the key
Most drugs work by binding to a protein.
The shape of the drug molecule fits the shape of a binding site on that protein, the way a key fits a lock. Bind, and the protein's behaviour changes.
Four kinds of target account for the great majority of medicines:
Receptors — proteins that normally receive a signal from a hormone or neurotransmitter (Chapter 1.8).
Enzymes — proteins that catalyse a chemical reaction.
Ion channels — pores that let charged particles across a membrane (Chapter 11.1).
Transporters — pumps that move substances across membranes.
And a smaller group works differently: antibiotics targeting structures bacteria have and we do not; chemotherapy attacking DNA replication; antacids acting by simple chemistry; and antibodies binding a specific molecule and removing it.
Agonists and antagonists
Two words that unlock most of pharmacology.
An agonist binds a receptor and activates it — it does what the body's own signal would do.
Salbutamol is an agonist at beta-2 receptors on airway muscle (Chapter 8.1). The body's own adrenaline does the same thing. Salbutamol is a chemical that fits the same lock, so the airway relaxes.
An antagonist binds the receptor and blocks it — it sits in the lock and stops the real key working.
Propranolol is an antagonist at beta receptors. Adrenaline can no longer act on them, so the heart does not race.
And the difference in how they behave follows directly:
Antagonists do nothing on their own. Someone with no adrenaline circulating notices very little from a beta blocker. The drug has an effect only in proportion to the signal it is blocking.
Which is why stopping a beta blocker suddenly is a problem. The body has increased its receptor numbers to compensate for the blockade. Remove the blocker and all those receptors are exposed at once, so the heart rate and blood pressure overshoot. This is why beta blockers are tapered rather than stopped.
Partial agonists — binding and activating, but only weakly.
Which produces genuinely useful behaviour: a partial agonist acts as a mild agonist when nothing else is present, and as an antagonist when a full agonist is around, because it occupies the site and delivers less effect than the full agonist would have.
Buprenorphine is a partial agonist at opioid receptors (Chapter 20.10). It relieves withdrawal, and it blunts the effect of a full opioid taken on top, which is exactly what you want in opioid substitution treatment.
Inverse agonists — binding and producing the opposite of the normal effect, because some receptors have activity even with nothing bound.
Where selectivity comes from
Receptors come in subtypes, and drugs can be built to prefer one.
Beta-1 receptors are mainly on the heart. Beta-2 receptors are mainly on airway and blood vessel smooth muscle.
Bisoprolol is relatively selective for beta-1, so it slows the heart with less effect on the airways — which is why it can be used more safely in someone with asthma than propranolol, which blocks both and can cause bronchospasm.
And "relatively" is the honest word. Selectivity is a preference, not an absolute, and it falls away at higher doses. A cardioselective beta blocker at a large dose still blocks beta-2.
This is the general rule for the whole of pharmacology: selectivity is a matter of degree, and it degrades as the dose rises. Which is the deepest reason side effects increase with dose.
Why side effects happen
And there are exactly three reasons, which is a satisfyingly short list.
1. The target is in more than one place.
Antihistamines block histamine receptors, which is what you want in the nose. The same receptors in the brain are involved in wakefulness, so the older antihistamines cause drowsiness. Newer ones were designed not to cross into the brain, which fixed the problem without changing the target.
2. The drug binds something else as well.
Amitriptyline was designed as an antidepressant and also blocks acetylcholine receptors, histamine receptors and alpha-adrenergic receptors — producing dry mouth, constipation, blurred vision, drowsiness and dizziness on standing (Chapter 11.9). Every one of those is a different unintended lock the same key happens to fit.
3. The intended effect is too large.
A blood pressure drug that works too well causes fainting. An anticoagulant that works too well causes bleeding. Insulin that works too well causes hypoglycaemia.
These are not really side effects; they are the main effect, at the wrong magnitude.
And that distinction matters practically: a type 1 problem may be solved by a more selective drug, a type 2 by a different drug in the same class, and a type 3 only by adjusting the dose.
The therapeutic index
The gap between the dose that works and the dose that harms.
A wide therapeutic index means the two doses are far apart. Penicillin, most antihistamines. You can be imprecise and stay safe.
A narrow therapeutic index means they are close together. Warfarin, lithium, digoxin, phenytoin, theophylline, and several chemotherapy drugs.
Which is why these specific drugs require blood level monitoring, careful attention to interactions, and dose changes made cautiously. It is not that they are more dangerous drugs; it is that there is less room between working and harming.
And it explains why an interaction that raises a drug level by 30 percent is trivial for one medicine and hospitalising for another.
Tolerance, dependence and rebound
Tolerance — the effect diminishing with repeated use.
The mechanisms are logical once you know them: receptors are reduced in number, receptors become less responsive, the body increases the enzymes that break the drug down, or a compensating system ramps up.
Dependence — the body having adapted so that stopping causes symptoms. This is physiological and is not the same as addiction (Chapter 20.10).
Rebound — the original problem returning worse than before, because of the adaptation. Rebound insomnia after sleeping tablets, rebound acid after proton pump inhibitors, rebound nasal congestion after decongestant sprays used for more than a few days.
All three are the same underlying phenomenon: the body pushing back against a sustained chemical change.
Why some drugs act in minutes and others in weeks
And this confuses people constantly, so it is worth being explicit.
Fast drugs act directly on an existing target. Salbutamol relaxes airway muscle within minutes because the receptor is there and the muscle responds immediately.
Slow drugs work by changing what cells produce.
Steroids enter the cell, bind a receptor, and the complex travels to the nucleus and changes which genes are transcribed (Chapter 2.4). New proteins must be made, so the anti-inflammatory effect takes hours — which is exactly why steroids are started early in an asthma attack rather than being relied on for immediate relief.
Antidepressants take weeks, probably because the useful change is in neural plasticity rather than in neurotransmitter levels, which change within hours (Chapter 20.7).
Bisphosphonates take months, because bone remodels slowly.
The rule: the closer the drug's action is to changing gene expression or tissue structure, the longer it takes to show.
Placebo, and why it is not nothing
Placebo effects are real, measurable, and physiological.
Placebo analgesia releases the body's own endorphins, and it can be blocked by naloxone — which is strong evidence that something chemical is genuinely happening rather than people simply reporting differently.
Effects are largest for subjective outcomes — pain, nausea, fatigue, mood — and small or absent for objective ones like tumour size or blood glucose.
Which is precisely why trials are placebo-controlled, and why a drug that beats no treatment has proved much less than a drug that beats placebo.
The nocebo effect is the mirror image, and it is practically important: telling someone about a side effect makes them more likely to experience it, which is well demonstrated in statin trials where muscle aches occurred at similar rates on drug and placebo (Chapter 21.6). The symptoms are genuinely felt; the attribution is what is mistaken.
Why individual people respond differently
Genetics — pharmacogenomics.
Around 8 percent of people of European ancestry lack a functioning CYP2D6 enzyme, which metabolises many drugs. Codeine is a striking example: it does nothing until the body converts it to morphine using that enzyme.
So poor metabolisers get little pain relief from codeine, and ultra-rapid metabolisers convert too much too fast and can suffer dangerous opioid effects. This has caused deaths in children after tonsil surgery, and it is why codeine is now restricted in children.
Age — newborns have immature liver enzymes and kidneys; older people have reduced kidney function, altered body composition and more sensitivity to sedatives. Doses frequently differ at both ends of life.
Organ function — the kidneys and liver clear most drugs, so impairment in either raises levels.
Body composition, sex, other illnesses, and everything else being taken at the same time.
Reading a drug's name
And this genuinely helps, because the endings are systematic.
| Ending | Class | Example |
|---|---|---|
| -olol | Beta blocker | propranolol |
| -pril | ACE inhibitor | ramipril |
| -sartan | Angiotensin receptor blocker | losartan |
| -statin | Cholesterol-lowering | atorvastatin |
| -prazole | Proton pump inhibitor | omeprazole |
| -cillin | Penicillin antibiotic | amoxicillin |
| -mab | Monoclonal antibody | ocrelizumab |
| -ib | Small-molecule inhibitor | imatinib |
| -pam / -lam | Benzodiazepine | diazepam |
| -dipine | Calcium channel blocker | amlodipine |
Which means an unfamiliar drug name frequently tells you its class before you look it up.
And on generic versus brand names: the generic name is the drug; the brand name is a manufacturer's label for it. Paracetamol and acetaminophen are the same molecule under different international names, and the single commonest cause of accidental overdose is taking two branded products that both contain paracetamol without realising it (Chapter 22.13).
Learn generic names. They are the only reliable way to know what you are taking.
What the next page fixes
Chapter 22.2 follows a drug through the body — how it gets in, where it goes, how it is broken down, and how it leaves. That journey determines the dose, the timing and most of the interactions.