Not medical or financial advice. Pocket Actuary is an educational picture built on the Social Security Administration 2023 period life table. Your own health, family history, and choices matter more than any table.
You
A partner
Set automatically to the opposite of yours.
You are the person on the Survival tab. Both lives are treated as independent, which is a simplification.
Estimate a hazard ratio
A hazard ratio multiplies your chance of dying in any given year compared with the average person in the SSA table. 1.00 is average, 1.50 means 50 percent higher each year, and 0.70 means 30 percent lower. Insurers use the same idea and call it a percent of standard mortality, with each table rating step adding 25 percent.
Your hazard ratio
This slider follows the estimate above automatically. You can also drag it, for example to match an insurance rating or to allow for a condition the estimator leaves out. A value of 1.00 reproduces the SSA table exactly. Illustrative only, not a health prediction.
Where these numbers come from
Smoking: Jha and colleagues, New England Journal of Medicine, 2013. Current smokers died at about 2.8 times the rate of never smokers for men and 3.0 times for women, ages 25 to 79. Quitting at 25 to 34 gained about 10 years, at 35 to 44 about 9, and at 45 to 54 about 6. The "quit at 45 or later" setting is my interpolation from those gains. About 11.0 percent of US adults smoked in 2023 (CDC).
Weight: Global BMI Mortality Collaboration, Lancet, 2016, never smokers without chronic disease, compared with BMI 20 to 25. Hazard ratios: 1.51 under 18.5, 1.13 for 18.5 to 20, 1.07 for 25 to 27.5, 1.20 for 27.5 to 30, 1.45 for 30 to 35, 1.94 for 35 to 40, 2.76 for 40 and up. NHANES 2017 to 2018: 30.7 percent of adults are overweight, 42.4 percent have obesity, and 9.2 percent have severe obesity.
Physical activity: Arem and colleagues, JAMA Internal Medicine, 2015, compared with no leisure time activity. Hazard ratios: 0.80 below the guideline, 0.69 for one to two times it, 0.63 for two to three times, 0.61 for three to five times. The guideline is 150 minutes a week of moderate activity. CDC reports 47.2 percent of adults met it in 2024.
Diabetes: Emerging Risk Factors Collaboration, 2011, hazard ratio 1.80 for all cause death after adjusting for age, sex, smoking, and BMI. About 15 percent of US adults have diabetes, diagnosed or not (CDC, my calculation from 39.8 million adults).
Family history: a UK Biobank analysis of about 186,000 people aged 55 to 73, reported in an American College of Cardiology release in 2016, found all cause mortality 16 percent lower for each extra decade a mother lived past 69 and 17 percent lower for a father. I use 0.835 per decade for each parent, centered on age 79, which is my assumption for a typical parental lifespan. Deaths before 69 count as 69 and deaths after 95 as 95, because the evidence covers later ages. A living parent counts only from age 80, at the age you enter. Two parents together count at 70 percent of their combined effect, my judgment, because they share habits and the same household. Twin studies put the inherited share of lifespan near 15 to 30 percent and a 2018 family tree study near 7 percent or less, so much of this signal overlaps with the lifestyle factors above. I read the release, not the paper itself.
Marital status: a 2020 meta-analysis of 20 studies and about 7.8 million people (Global Health Research and Policy) found all cause mortality for unmarried compared with married people of 1.46 for men and 1.22 for women. By group, never married was 1.67 for men and 1.46 for women, divorced or separated 1.59 and 1.27, and widowed 1.30 and 1.14. Men carried more excess risk than women. I read a summary of the paper, not the paper itself. The studies are observational, and part of the gap is healthier people marrying and staying married, plus overlap with smoking, income, and activity. So I made three adjustments of my own. Each ratio is divided by the population average, using rough shares I assumed for US adults (50 percent married, 32 never married, 12 divorced or separated, 6 widowed). The result is shrunk to half strength in log terms. Nothing is counted under age 30. Cohabiting, marriage quality, and children were not studied.
Grandparents: no study I could verify gives a hazard ratio for them. A grandparent shares half as many genes as a parent, so I apply half the parent effect in log terms, about 0.91 per decade, centered on an average age at death of 76 (my assumption, because earlier generations died younger) and limited to 65 through 95. It counts only when no parent line applies, since parents already carry most of the signal. Treat it as a small nudge, not a finding.
Age: the weight effect is scaled down after age 60, reaching 45 percent of its strength by age 80, from the 1.52 versus 1.21 per 5 BMI units figures above. Starting the taper at 60 is my assumption. Smoking and the other factors are not tapered because I could not verify age specific ratios.
Combining: Li and colleagues, Circulation, 2018, found five healthy habits together carried a hazard ratio of 0.26 against none, with 12 to 14 more years of life at age 50. Simply multiplying separate ratios overshoots that spread, so the estimator shrinks the combined effect to 65 percent in log terms. That 65 percent is my judgment, chosen so the extremes stay inside what Li found. It is not a published figure.
Rescaling: the SSA table describes the average American, not a never smoker at a healthy weight. Each ratio is divided by the average ratio across the population, so a typical mix of answers lands near 1.00. The splits inside broad CDC groups (for example, half of adults below the activity guideline being fully inactive, and equal shares within each weight band) are my assumptions.
Insurance: rating tables run A to J, each step adding 25 percent to standard mortality. Insurers' standard is their own table, not the SSA average, and each company sets its own rules.
Left out on purpose: existing conditions such as heart disease or cancer, alcohol, diet, blood pressure, a parent's cause of death, family history of specific diseases, and friendships and other social ties. Either I could not verify a usable hazard ratio, or the effect overlaps too much with the factors above. Ratios describe groups, not individuals, and a prior diagnosis can outweigh everything here. Not medical advice.
What your hazard ratio does
Mortality improvement
Death rates have fallen over time. This lowers each future year's death rate by the percent shown, compounding. Zero percent freezes 2023 rates.
Life in weeks
Social Security claiming
Full retirement age 67, which applies to anyone born in 1960 or later. Enter the monthly benefit you would get at 67.
Model check and limitsHow this works
Death rates (qx) for ages 0 to 119 come from the SSA 2023 period life table used in the 2026 Trustees Report. Survivors (lx) start at 100,000 and shrink by each year's rate. Remaining life comes from the area under the survival curve. A hazard ratio changes each rate by q' = 1 minus (1 minus q) raised to the ratio. Improvement multiplies the rate k years ahead by (1 minus r) to the k. The Factors tab can estimate a hazard ratio from smoking, weight, activity, and diabetes using published studies, rescaled to the population average and shrunk for overlap. Fractional ages use straight line interpolation. Break even ages ignore taxes, inflation adjustments, interest, and spousal benefits.