US Mortality Gains Slow to 4% for Men as Ages 25-29 Improve 17.6%

The 2024 slowdown in US mortality improvement happened almost entirely at ages 65 and over. Men in that band improved 2.1% in 2024 against 7.6% a year earlier, while men aged 20 to 44 improved 12.5%, up from 8.0% (SOA Research Institute, October 2026). The all-ages figure of 4.0% averages a young-adult population that has cleared its pandemic excess with an older population whose catch-up has ended.

On October 15, 2026 the SOA Research Institute published U.S. Population Mortality Rates 2000-2024 and its single-age data file. Read as one number, the release says improvement halved. Read by age, term and group life writers insure the ages that gained most, while payout annuity and pension reserves sit in the band that gained least.

Key Takeaways

4.0% and 2.4%: all-ages male and female improvement in 2024, down from 8.0% and 7.1% in 2023; under-65 improvement was 7.2% and 5.2% against 2.1% and 1.6% at 65 and over (SOA Research Institute, October 2026).
17.6%: male improvement at ages 25 to 29, the largest single-year gain for that band in the 25-year series, where the prior high was 11.5% in 2022 and again in 2023. Females at the same ages improved 16.0%.
8.7% below 2019: male mortality at ages 25 to 29 in 2024 (0.00147 against 0.00161), computed from the single-age file, while age 45 males sit 13% above 2019. The pandemic excess has cleared at both ends of the age range and persists in the middle.
0.65% a year: annualized improvement for men aged 65 to 84 from 2019 to 2024, against 0.87% a year for the same band from 2009 to 2019, so two catch-up years still leave the band behind its pre-pandemic trend.
2013 to 2023: the historical window of the 2026 VM-20 improvement scale LATF adopted on September 17, 2026, which leaves 2024, the first year the older-age catch-up ended, outside the fit (SOA MI Life Working Group, September 2026).

Two Populations Inside One 4.0% Figure

Tables 3.1 to 3.6 of the report decompose the headline by five-year band, broad band and the under-65 split, for each sex. "In 2024, there was improvement in every five-year age group, with improvement percentages ranging from 1.3% (ages 90 to 94) to 17.6% (ages 25 to 29)" (SOA Research Institute, October 2026). The 2000-2023 edition had reported the equivalent spread for 2023 as 5.4% to 10.4% (SOA Research Institute, November 2025). In one year the range widened from five points to sixteen.

Older ages carry the slowdown. Men aged 65 to 84 improved 2.2% in 2024 against 8.1% in 2023, and women in that band 2.0% against 7.2%. Men aged 45 to 64 fell to 5.2% from 8.8%. Men aged 20 to 44 moved the other way, to 12.5% from 8.0%. Every band from 60 upward improved by less than 4%, and every male band below 45 improved by more than 8%.

The CDC's final 2024 data, on an age-adjusted all-ages basis, lines up with the SOA split. The national age-adjusted death rate fell 3.8% to 722.1, with the rate at ages 25 to 34 down 15.9%, from 148.1 to 124.5, while ages 65 to 74 fell 2.0% and 75 to 84 fell 1.9% (CDC NCHS, January 2026). Unintentional injuries, the leading-cause category that carries overdoses, fell 14.4%, the largest decline among the ten leading causes.

Level Against 2019: Where the Pandemic Excess Has Cleared

Improvement rates say how fast mortality moved. The single-age file says where it stands. Averaging the crude rates across each band and comparing 2024 with 2019 gives the table below.

Age band	Males, 2024 rate vs 2019	Females, 2024 rate vs 2019	Males, 2019-2024 annualized improvement	Females, 2019-2024 annualized improvement
25 to 29	8.7% lower	11.3% lower	+1.80% a year	+2.37% a year
35 to 44	12.0% higher	4.2% higher	-2.29% a year	-0.83% a year
45 to 54	2.4% higher	2.5% lower	-0.47% a year	+0.51% a year
65 to 84	3.2% lower	2.3% lower	+0.65% a year	+0.47% a year
85 to 94	3.3% lower	1.5% lower	+0.66% a year	+0.29% a year

Computed by actuary.info from the SOA single-age crude mortality rate file: unweighted average of single-age rates within each band, 2024 against 2019; annualized improvement is the geometric rate over five years. Lower mortality is favorable. The SOA's published improvement rates are weighted to 2012 population and differ slightly.

At the young end the recovery is complete. Male mortality at ages 25 to 29 averages 0.00147 in 2024 against 0.00161 in 2019, and the female rate is 0.00058 against 0.00065. Mid-life ages have not recovered: age 45 males sit at 0.00375 against 0.00332, and the 35 to 44 band is 12% above 2019. At the old end, 65 to 84 males are 3.2% below 2019, but that is five years at 0.65% a year against 0.87% a year from 2009 to 2019. Two catch-up years of 7.7% and 8.1% plus 2024's 2.2% left the band short of its pre-pandemic trend line.

For payout annuity and pension reserves, 65 to 84 is the band that carries the improvement assumption, and its observed five-year rate now runs a quarter below the pre-pandemic trend. Annuitants dying slightly faster than a scale projects releases reserve, with SPIA sales at a record $4.0 billion in Q2 2026, up 12% year over year (LIMRA, July 2026). The site's reading of the SOA-LIMRA payout annuity study cut against assuming it: that study's 2023-2024 annuitant cells pointed to survivor selection running the other way. One year at 2.1% after two years above 7% is a thin basis for trimming a long-term improvement rate.

For term and group life the issue ages split. At 25 to 29 the gain is a cause-of-death event: overdose deaths fell 26.9% in 2024, to an estimated 80,391 from 110,037 (CDC NCHS, May 2025), and the death rate at ages 25 to 34 fell 15.9%. At 35 to 44, where male mortality sits 12% above 2019, an expected basis set on 2019 experience carries a 12-point excess that 2024's improvement of 11.5% at 35 to 39 and 8.8% at 40 to 44 did not close.

The 2026 VM-20 Scale Was Fit on Data Ending in 2023

The individual life HMI and FMI scales LATF adopted on September 17, 2026 take their historical component from a geometric average of insured mortality improvement over 2013 to 2023, and their future component from the SSA 2026 Trustees intermediate projection over 2023 to 2043 (SOA MI Life Working Group, September 2026). So 2024, the year the older-age catch-up ended, lies outside the historical window and enters only through the SSA projection.

At ages 26 to 40 the FMI assumes a return to zero improvement in 2028, zero through 2031, then a grade to the long-term rate by 2036, and the site's reading of the September exposure traced a -1.28% male band at ages 32 to 39. The population table carries 17.6% at 25 to 29 and 15.6% at 30 to 34, with the 25 to 29 level below 2019. Scale and data point in opposite directions at the ages where the scale's judgment is sharpest.

That disagreement is deliberate. Appendix B of the scale document records the 2022 industry principle that a shock "should be reflected in valuation mortality only to the extent it is expected to continue in the future," and a July 2026 LATF presentation that proposed bridging 2023 to 2025 with general-population data directly was set aside before adoption.

The series gives that caution a basis. From 2009 to 2019, male mortality at 25 to 29 deteriorated 1.64% a year, so the 2024 rate of 0.00147 still sits above the 0.00129 the same band recorded in 2000. A record single-year gain has recovered ground lost over the opioid decade, and a cause-specific decline of that size has a different reversion profile from improvement driven by cardiovascular or cancer treatment.

There is a measurement problem underneath as well. Population counts for 2020 to 2024 come from the Census Bureau's Vintage 2025 estimates, whose April 2020 base is now built "almost entirely from the 2020 Census via the Modified Age and Race Census (MARC) file," a change the Bureau calls "a notable departure" from the vintages produced since 2020 (Census Bureau, May 2026). According to the SOA, the new counts raised male rates at ages 26 to 64 and lowered them at 20 to 25 for 2020 through 2023.

Its own 2023 improvement figure for men aged 25 to 29 moved from 10.4% in last year's edition to 11.5% in this one. A company that benchmarked an improvement study to the 2000-2023 file cannot reproduce it from the 2000-2024 file, and part of the record 17.6% is a 2023 denominator restated upward before 2024 was measured against it.

US Mortality Gains Slow to 4% for Men as Ages 25-29 Improve 17.6%

Key Takeaways

Two Populations Inside One 4.0% Figure

Level Against 2019: Where the Pandemic Excess Has Cleared

The 2026 VM-20 Scale Was Fit on Data Ending in 2023

Further Reading

Sources