Affluent people are 50 per cent more likely to take part in bowel cancer screening than their deprived counterparts, according to a study published in the British Journal of Cancer (BJC).
The Cancer Research UK study, which looked at how many people in the London area returned their home testing kits for bowel cancer, found that only one in three people in the most deprived areas participated.
This compared with one in two people in the wealthiest areas.
The figures are part of the National Awareness and Early Diagnosis Initiative (NAEDI) supplement published in the BJC, which for the first time brings together research underpinning the initiative.
Professor Jane Wardle, lead author from Cancer Research UK's Health Behaviour Research Centre based at University College London, said: "There's a real danger that bowel cancer could increasingly become a disease of lower social class if these figures hold true across the UK.
"We know that more than 90 per cent of bowel cancer patients survive if the disease is caught at the earliest stage, compared with around six per cent for cases detected at the latest stage.
"Screening helps to spot early signs of bowel cancer, as well as pre-cancerous growths that don't cause any symptoms, so it's important that everyone who receives a testing kit takes part."
Bowel screening kits - faecal occult blood (FOB) tests - are sent through the post every two years to people aged between 60 and 74 years old2. People take their own stool samples at home and then send the kit back for testing. The results are posted out within two weeks.
Professor Wardle continued: "Our study reveals a worrying pattern, but we don't know why social class has played such a major role in screening uptake.
"Bowel cancer screening may not be pleasant, but this seems to be less of a deterrent for those who are better off.
"Future research will need to discover what the key barriers are."
The study monitored the response to the roll out of bowel cancer screening across London in the first 30 months of the programme.