Do National Longevity Rankings Miss the Most Important Story
- Lowell Sheppard
- Feb 5
- 6 min read
A rabbit hole I fell into after finishing a book on longevity in Japan
I’ve just finished writing a book on longevity lessons from Japan, and the manuscript is now in my agent’s hands. For various reasons, I can’t share much from the book itself at the moment. But since submission, I found myself down a rabbit hole.
Before writing it, I spent fifteen months moored in Okinawa, long celebrated as one of the world’s most famous “Blue Zones.”
While there, I made an uncomfortable discovery: Okinawa had fallen from its Blue Zone perch. It had become the only prefecture in Japan where the average life expectancy was declining.
That discovery prompted a year-long search in 2025, travelling across Japan, searching not for myths or marketing labels, but for current longevity hot spots, places where people were still managing to live long lives with a greater degree of independence.
My research was qualitative rather than quantitative. I was observing, listening, and spending time in communities. But one thing quietly guided where I went next: Japan’s prefectural longevity tables. Those tables became my map.
And once you start living with data at the prefectural level, 47 different aging profiles inside a single country, it raises an obvious question you hadn’t planned on asking. If Japan looks this different internally, how does it really compare with other countries once we stop hiding behind national averages?
That’s the rabbit hole I fell into after submitting the manuscript.
Armed with detailed Japanese prefectural data, I wanted to see how Japan measured up against North America, particularly the United States and my home country, Canada, not just in how long people live, but how long they remain healthy, and how long decline tends to last.
What follows isn’t an excerpt from the book, nor a sneak preview of its conclusions. It’s simply what happened when curiosity got the better of me again. (This happens more often than I admit, and usually at inconvenient times.)
Why national rankings fall short
Most international discussions about longevity rely on national rankings.
Japan is near the top. The United States is well below. Canada is somewhere reassuringly in between. But after months of working with Japan’s 47 prefectural tables, those national comparisons began to feel strangely blunt—like trying to understand a country by averaging its weather.
So I tried a different lens. What happens if, instead of comparing countries, we compare prefectures to states and provinces?
How do we even define a “longevity hot spot”?
There isn’t one correct answer. Different metrics illuminate different things.
Three commonly used lenses are:
1. Centenarians per capita (the classic “Blue Zone” style metric)
2. Life expectancy at birth (how long people live)
3. Healthy life expectancy (HALE) (how long people live in good functional health)
I looked at all three. What surprised me wasn’t that they told different stories, that part was expected, but how rarely they pointed to the same places.
Over time, one measure began to interest me more than the others. Not how long people lived, but how long decline lasted. Before getting there, it’s worth seeing how the tables actually line up.
How to read the tables below
To avoid mixing signals, each table uses one metric applied consistently across Japan, the United States, and Canada.
Where definitions or data availability differ, particularly for centenarians in Canada, this is clearly noted. The goal here isn’t false precision. It’s comparative clarity.
Table 1
Centenarians per 100,000 Population
(“Blue Zone”–style metric: cohort survival)
Japan — Prefectures (2024–2025)
Prefecture | Centenarians / 100,000 |
Shimane | 168.7 |
Kōchi | 157.2 |
Tottori | 144.6 |
Kagoshima | 136.5 |
Nagano | 133.9 |
Lowest: Saitama | 48.5 |
Source: Japan Ministry of Health, Labour and Welfare (summarised by Nippon.com)
United States — States (2020 Census)
State | Centenarians / 100,000 |
Hawaii | 44.4 |
Rhode Island | 39.7 |
South Dakota | 38.7 |
Connecticut | 36.4 |
North Dakota | 34.8 |
Lowest: Utah | 10.4 |
Source: U.S. Census Bureau (Centenarians: 2020)Note: Census reports per 10,000; values converted here to per 100,000.
Canada — Provinces (Data Limited)
Canada does not publish regularly updated provincial centenarian rates. The most complete standardised dataset remains the 2011 Census.
Province (2011) | Centenarians / 100,000 |
Saskatchewan | ~31 |
Quebec | ~29 |
Ontario | ~25 |
Source: Statistics Canada, 2011 Census. Included for context only; not comparable in timeliness to Japan and U.S. data.
What this table already tells us
· Japan’s top prefectures operate on an entirely different centenarian scale from North America.
· Even within Japan, the spread is more than threefold.
· Centenarian counts tell us who reaches 100, not how life is lived on the way there.
Which brings us to the next lens.
Table 2
Life Expectancy at Birth (Years)
(Longevity metric: how long people live)
Japan — Prefectures (2022–2023)
Prefecture | Life Expectancy |
Nagano | ~84.9 |
Shiga | ~84.7 |
Fukui | ~84.6 |
Lowest: Aomori | ~81.0 |
Source: Japan Ministry of Health, Labour and Welfare
United States — States (2019, pre-pandemic)
State | Life Expectancy |
Hawaii | 80.9 |
California | 80.9 |
New York | 80.7 |
Lowest: Mississippi | 74.4 |
Source: CDC / National Center for Health Statistics
Canada — Provinces (2021–2022)
Province | Life Expectancy |
British Columbia | ~83.0 |
Ontario | ~82.7 |
Quebec | ~82.5 |
Lowest: Newfoundland & Labrador | ~79.5 |
Source: Statistics Canada
What life expectancy smooths over
Life expectancy narrows international gaps, but it doesn’t erase regional variation. Canada and the U.S. overlap substantially at the provincial and state level. Japan retains an advantage, but place still matters, sometimes more than national policy.
Still, this measure doesn’t answer the question that kept nagging at me.
Table 3
Healthy Life Expectancy (HALE / HALE-Equivalent, Years)
(Quality-of-life metric: how long people live in good functional health)
Japan — Prefectures
(Official “healthy life expectancy,” defined as years lived unassisted)
Prefecture | Healthy Life Expectancy |
Nagano | ~75.3 |
Shimane | ~74.8 |
Yamagata | ~74.5 |
Lowest: Aomori | ~71.0 |
Source: Japan Ministry of Health, Labour and Welfare
United States — States
(IHME / Global Burden of Disease estimates)
State | HALE |
Minnesota | ~70.1 |
Colorado | ~69.8 |
Utah | ~69.6 |
Lowest: Mississippi | ~64.2 |
Source: IHME, Global Burden of Disease
Canada — Provinces
(Health-adjusted life expectancy, HALE-equivalent)
Province | HALE-Equivalent |
Quebec | 71.1 |
British Columbia | 69.1 |
Ontario | 68.9 |
Lowest: Nova Scotia | 65.7 |
Source: Statistics Canada (2019 baseline)
Note: HALE definitions vary slightly by country; values are most reliable for within-country comparison.
The pattern that finally caught my attention
Seen together, these tables make a few things hard to ignore:
Different metrics highlight different places.
Very few regions perform near the top across all measures.
The timing and length of late-life decline vary more than lifespan itself.
Seen together, these tables did something I didn’t quite expect. They didn’t deliver neat answers; they raised better questions.
Different metrics highlight different places. Very few regions sit near the top across all of them. And the timing and length of decline appear to vary far more than lifespan itself.
One thing is worth stating clearly, though. Even when you move below the national level, Japan continues to perform exceptionally well. While prefectures vary widely across the country, Japan’s regional outcomes still rank near the top in global comparisons. When small, exceptionally wealthy city-states are set aside, Japan remains the world’s longest-lived large society by most standard measures, and among the strongest when healthy years are taken into account. The variation inside Japan is real, but it starts from a remarkably high baseline.
Once I noticed that, I stopped asking which country “wins” the longevity tables and started asking a quieter question: Where does decline arrive later—and where does it linger longer?
That question rarely produces a tidy conclusion. It tends to raise an eyebrow instead. And once you’ve raised an eyebrow, it’s usually only a short slide into another rabbit hole.
So consider this a field note rather than a verdict. If I’ve misread a number, missed a source, or if you see the data pointing somewhere I haven’t yet looked, I’d genuinely welcome the correction. I’m still learning, and the numbers have a habit of saying more than we first hear, especially when we slow down long enough to listen.
Which leaves me, once again, standing at the edge of other rabbit holes, wondering which one to dive into next.
Sources (consolidated)
· Japan Ministry of Health, Labour and Welfare — Prefectural Life Expectancy & Healthy Life Expectancy
· Nippon.com — Summary of MHLW Centenarian Data
· U.S. Census Bureau — Centenarians: 2020
· CDC / National Center for Health Statistics — State Life Tables (2019)
· IHME — Global Burden of Disease, State-level HALE
· Statistics Canada — Provincial Life Expectancy & Health-Adjusted Life Expectancy
