A global picture of pain
A study in Nature Medicine pulled together self-reported pain data from 6,125,459 people in 118 countries and territories. The analysis covered 902 population-based data sources collected between 1990 and 2025, with participants aged five to more than 100. About 55 percent were female.
The researchers built reference curves for pain at 11 body sites across the whole lifespan. They found that pain prevalence ranged from 2 percent for facial pain to 40 percent for back pain. Women reported higher prevalence than men at every site.
Pain does not rise in a straight line
The largest sex differences appeared in facial pain, which was 83 percent higher in women, headache at 74 percent higher, and stomach or abdominal pain at 60 percent higher.
Age patterns were more complex than expected. Pain at seven of the 11 sites peaks in mid-to-late adulthood and then declines. Headache, abdominal pain, neck or shoulder pain, elbow pain and facial pain all peak earlier in life. Musculoskeletal pain in the lower body behaves differently. Back, hip and knee pain rise fastest between ages 20 and 55 and keep climbing, reaching their highest levels at or beyond 75 years.
Overall, the share of people reporting any bodily pain increased from about 30 percent at age five to 70 percent at age 100 or older. High-intensity pain peaked around age 50, while generalised pain peaked around 70.
Where a person lives shapes the curve
From age 40, pain prevalence rose more steeply in countries with low scores on the Human Development Index. At the oldest ages, low back pain was nearly twice as common in low-HDI countries as in high-HDI countries.
Smoking, obesity and low household income were linked to more pain. Together the three factors accounted for about 18 percent of site-specific pain burden worldwide, with the share ranging from 12.6 percent in sub-Saharan Africa to 27.1 percent in eastern Europe.
The authors say prevention may need to start earlier in life and be tailored to each body site. They also note a limitation: most datasets were cross-sectional, so the curves compare age groups rather than tracking individuals over time.