Type 2 Diabetes Causes
Type 2 Diabetes Causes
Type 2 diabetes is produced by two things in sequence: the body's cells stop answering insulin as readily as they did, and the pancreas — which has been producing more insulin to compensate — eventually cannot keep up. This page sets out what drives each half, why where fat is carried matters more than how much there is, what is inherited and what is not, and which medicines and conditions push glucose up in someone who already has the condition.
Insulin resistance — the first half
Insulin is the signal that tells muscle, liver and fat tissue to take glucose out of the bloodstream. Insulin resistance is the state in which those tissues stop answering that signal as readily as they used to, so a normal amount of insulin moves less glucose than it should, and glucose stays in the blood.
The body's answer is elegant and it works for years: the pancreas senses the glucose is not falling and produces more insulin. Glucose stays in the ordinary range, the person feels entirely well, and nothing about the condition is visible from the outside. This is the stage at which most people are for a long time, and it is the reason there is no onset to point at.
What makes resistance worse, and what improves it, is mostly a matter of where energy is stored and how much muscle is being used. Muscle is the largest tissue that clears glucose from the blood, so using it changes the picture directly — which is why resistance training has a stronger claim here than its reputation allows. Fat inside the abdomen and around the liver releases signals and fatty acids that interfere with insulin's action, and reducing it improves resistance measurably. Untreated sleep apnoea, shortened sleep and long stretches of sitting all push in the other direction.
Resistance on its own is not diabetes. Plenty of people carry substantial resistance and never cross the line, because their pancreas keeps compensating. What turns resistance into the condition is the second half.
Beta cells — the half that decides when it becomes diabetes
The beta cells of the pancreas make insulin, and in type 2 diabetes their output falls gradually across the years. It is the combination that makes the diagnosis: resistance sets the demand, and the point at which supply no longer meets it is the point at which glucose rises.
Two consequences follow, and both explain decisions that otherwise look arbitrary. The first is that the condition tends to need more treatment over time rather than less, because the underlying supply is declining. That is not treatment failure and it is not the person's doing; it is the natural history, and it is the reason plans are reviewed rather than fixed.
The second is that the years count. Control achieved early appears to carry benefit forward, which is why guidelines treat the period around diagnosis as the most valuable time to get the number down rather than the least urgent. It is also why a person whose glucose has been only slightly high for a decade is not in the same position as someone whose glucose has been much higher for two years: both duration and level contribute, and duration is the part that cannot be recovered.
None of this is a reason for alarm and it is not a countdown. The rate varies enormously between people, some maintain good control on the same treatment for many years, and nothing about the mechanism implies a schedule. What it argues for is continuity: the same series of measurements, reviewed at intervals, so that change is noticed when it is small.
Body fat: where it is matters more than how much
Excess weight is the largest single modifiable driver of type 2 diabetes, and the qualifier in that sentence is doing real work. What drives insulin resistance is not weight in the abstract but fat in particular places: around the abdominal organs, and inside the liver itself. Fat stored under the skin elsewhere is far less metabolically active, which is why two people of the same weight can carry very different risk.
This is also why the number on the scale is a poorer guide than the tape measure, and why a modest loss can produce a disproportionate improvement. Fat in the liver is mobilised early when intake falls, and insulin resistance in the liver improves before the person has lost very much at all. Waist measurement is worth recording alongside weight for that reason, and a falling waist with a steady weight is not a failed attempt.
The other direction deserves care. Not everyone with type 2 diabetes carries excess weight, and in several populations the condition appears at weights that would be called unremarkable in Europe or North America. Telling a person of ordinary weight that their condition is their weight's fault is both wrong and counterproductive. What matters for them is the same thing that matters for everyone else: activity, sleep, the medicines review, and the possibility that the type has been mislabelled.
Where weight is part of the picture, doing it with support rather than alone matters. Our endocrinology department runs a medical weight-loss clinic that builds programmes from an endocrine and metabolic standpoint, with attention to reducing the chance of regaining weight afterwards — because the familiar pattern is not failing to lose it, but losing it and putting it back.
Family history, age and ethnicity — the part nobody chose
The largest share of the risk in type 2 diabetes is not chosen and not earned. It is inherited, it accumulates with age, and it varies by population. Stating this is not politeness; it is the reason the condition should not be discussed as though it were a verdict on a person's habits.
| Factor | How it acts | What follows from it |
|---|---|---|
| A first-degree relative with type 2 diabetes | The strongest single risk factor, and it runs through both resistance and beta-cell capacity | Screening earlier and more often, and the rest of the family knowing their own position |
| Increasing age | Resistance rises and beta-cell output falls with age, independently of weight | Screening becomes worth doing on a schedule rather than waiting for a symptom |
| Ethnicity | Higher risk at lower body weights in South Asian, African, African-Caribbean, Middle Eastern and East Asian populations | The threshold at which screening starts is lower, and the type deserves checking in someone young and lean |
| Gestational diabetes in a previous pregnancy | A strong marker of the beta-cell reserve a person has, and of risk later on | Retesting after the pregnancy, and at intervals thereafter |
| Polycystic ovary syndrome | Travels with insulin resistance, independently of body weight | Glucose checked as part of the condition rather than years later |
| Low birth weight, and growth patterns in childhood | Associated with reduced beta-cell capacity in adult life | A piece of the history worth giving, rather than a factor to act on |
What that table argues for is screening rather than worry. None of these can be changed, and knowing about them changes exactly one thing: how early and how often the test is done. Where the result is in the raised-but-not-diabetic band, that is the point at which something can still be done about it, which is described on the monitoring page.
Medicines, conditions and habits that push glucose up
There is a long list of things that raise glucose in someone who already has the condition, and it is worth knowing because "my control has slipped and nothing has changed" usually turns out to be one of them.
Medicines. Corticosteroids are by far the most important, and they raise glucose substantially, predictably and dose-dependently — whether taken by mouth, by injection, by inhaler at high dose or into a joint. Several antipsychotics raise it, as do some immunosuppressants used after transplantation, some diuretics at high dose, and several others. Glucose should be checked when any of these is started, and the plan adjusted while the course lasts rather than after it.
Conditions. Anything that produces a stress response raises glucose, which is why an infection, a surgical admission or a heart event can destabilise a plan that had been steady for years. Diseases of the pancreas — pancreatitis, pancreatic surgery, and the conditions that damage it — cause a form of diabetes with its own management. Hormonal conditions such as Cushing's syndrome and acromegaly raise glucose as a consequence of the hormone rather than as a primary problem, and haemochromatosis does it through iron deposition.
Sleep. Untreated obstructive sleep apnoea is common in people with type 2 diabetes and it worsens insulin resistance nightly. Snoring, waking unrefreshed, morning headache and daytime sleepiness are the things to mention, and treating the apnoea often improves glucose and blood pressure together.
Habits. Long stretches of sitting, shortened sleep, heavy alcohol use, and smoking — which does not raise glucose directly but adds vascular risk that the condition has already raised. And a febrile illness, where the rise in glucose is expected and the plan should already say what to do about it: the sick-day rules are on the day-to-day page.