Health Waves: Two Sources of Hope in 2025

Inspiration and hope are close companions; we rarely find one without the other.

This week I learned about Opal Sandy, a toddler who lives in Oxfordshire, England.

Opal was born deaf, but in March of last year, at the age of 18 months, gene therapy in her right ear, along with a cochlear implant in the left, allowed the girl to hear for the first time.

Within a few months, Opal's parents were contending a new problem: Opal had come to love making noise. Playing drums. Banging spoons on the table. In short, the usual toddler stuff.

Stories like this inspire me – and give me hope. Partly because I have a condition in one ear similar to Opal's. But mainly because gene therapy is creating a brave new world of medical possibilities. It's a stunningly expensive world, at least for now, but Opal's "therapy" consisted of a single, 16-minute procedure. That feels magical.

In this newsletter, I'll be describing two sources of hope for 2025. My focus is on recent "health waves", or signs of recovery from recently worsening trends.

There's a difference between entertaining hope and taking the blue pill. I won't claim that the U.S. is poised to become the healthiest country in the world. We have a long way to go, by any measure. My point is simply that recent data offers grounds for hope.

The chronic disease epidemic

Currently, nearly half of American adults suffer from at least one chronic disease.

(According to the CDC, the 10 most prevalent and costly examples are obesity, hypertension, high cholesterol, coronary heart disease, chronic obstructive pulmonary disease, asthma, chronic kidney disease, diabetes, cancer (excluding skin cancer), and depression.)

Over the past two decades, the prevalence of chronic illness in the U.S. has increased. Our per capita rates are higher than in any of the other 37 wealthy OECD countries, though we spend more on health care. Put simply, we spend more for less.

We've also experienced mostly steady growth in traffic fatalities, drug overdose deaths, and suicides, and on these dimensions too, we lead the OECD.

Grim stuff. And yet, over the past one to two years, we've seen declines in adult obesity, drug overdose deaths, traffic fatalities, and murder rates. In a December 2024 article for The Atlantic, Derek Thomson described these trends as a "health wave."

A health wave?

Mr. Thomson is a brilliant writer – I strongly recommend his Work in Progress newsletter ­­­– but I believe that he mischaracterized this so-called wave.

If anything, we're experiencing multiple waves, including some Thomson didn't mention. At the same time, he and others have exaggerated the extent of improvement in areas such as obesity.

What caught my eye – and partly motivated this newsletter – is that Mr. Thomson's reading of the data reflects some of the most common ways statistics are misrepresented.

I want to share some details with you, because I think that informed hope is the most desirable kind. You'll find some gentle red-pilling here – and some encouraging takeaways.

Traffic fatalities

The latest NHTSA estimate shows quarterly declines in traffic fatalities for eight quarters in a row. The current rate is around 1.13 for every 100 million vehicle miles – still higher than other OECD countries, but moving in a positive direction. So yes, complete agreement with Mr. Thomson on this point.

What's been happening on our roads? Experts are divided about the relative contributions of safer roadways (thanks to NHTSA funding), tougher policing, and new vehicle safety technologies.

(I'm biased toward the new technology account. Two years ago, after trading in my 2012 Prius for a new Accord Hybrid, I became acquainted with all the latest driver assistance features. I'm not sure I could have an accident now, even if I wanted to.)

Murder rates

Here the news is even better than Mr. Thomson lets on. Murder rates have been declining since the height of the pandemic, but this is just one small part of a decades-long trend.

The figure below, from Jeff-alytics, shows murder rates per 100,000 Americans since 1960.

If you only focus on changes since the pandemic, as Mr. Thomson did, you do see what appears to be a dramatic decline since 2020. (Just focus on those last four bars on the right.)

Scan the entire figure though, and a more complete story emerges: Per capita murder rates sharply increased beginning in the 1960s, then declined about as sharply in the 1990s. What's happened over the past decade is a mini-spike that peaked during the pandemic and is now dropping toward earlier levels.

(Looking at the final four bars vs. the entire figure illustrates the point that time trends can seem very different depending on what time period is being examined. Always check the x-axis for that!)

Experts like Jeff Asher see the decline in murder rates as leveling off in 2024-2025, but I'm hoping it will continue.

During the 2024 presidential election, violent crime – or crime in general – was often depicted as spiraling out of control. Republicans are best known for pushing this narrative, but it remains a somewhat bipartisan misconception. (In a 2023 Gallup poll, for instance, 92% of Republicans, 78% of Independents, and 58% of Democrats claimed – incorrectly – that crime rates in the U.S. were higher than the previous year.)

National rates for most types of crime (including violent crime) have been dropping since the 1990s. Public perceptions don't fully align with the data.

Earlier this year I discussed this disconnect, as well as some of the more than 35 (!) different explanations for the decline of crime since the 1990s. Whatever the cause, this trend almost surely constitutes good news.

Drug-overdose mortality

In 2023, for the first time since 2018, drug overdose fatalities declined in the U.S. Specifically, the CDC reported a 3% reduction, much greater than seen from 2017 to 2018. New data shows that the decline continued through 2024.

This is an encouraging trend, although we still lead the world in per capita drug overdose fatalities. Keep in mind too that the figure below doesn't capture the steady rise in drug overdoses since the 1980s, when age-adjusted rates were one-fifteenth or less of current levels. (Again, the x-axis matters.)

CDC data show that opioid-related fatalities are mainly responsible for the recent decline, but experts don't agree about what's responsible. By my count, more than two dozen potential contributors are being discussed and debated. Most of them fall under one of three headings:

• changes to local drug policies (greater availability of naloxone, more access to addiction and harm reduction programs, etc.)

• changes to the population of drug users (former users either recovered or died)

• changes to the drugs themselves (less potent fentanyl; more dilution of fentanyl with xylazine, which keeps people sedated longer)

Although Mr. Thomson accurately described the drug overdose statistics, his conclusions seem a little blue-pilled. I see grounds for hope, but not in every conceivable scenario.

For instance, drug policies do seem to impact overdose rates, but not everyone agrees on which policies to pursue.

Liberals tend to advocate for rehabilitation, greater availability of naloxone, more needle-exchange and fentanyl test strip resources, and so on.

The more conservative approach – which we'll almost surely be seeing over the next four years ­– emphasizes stricter drug laws and enforcement.

The latter approach, even if effective in some respects, can backfire. This is illustrated by the Richard Cowan's "iron law of prohibition": As drug law enforcement becomes more intense, the potency of the drugs increases. (Stiffer penalties incentivize drug producers to sell more concentrated products; these products take up less space to transport while bringing in more revenue.)

In short, stricter drug laws, combined with the greater border security that Donald Trump promises, may result in drug cartels shifting back to more concentrated fentanyl, which could in turn reverse the decline in overdose fatalities we've seen.

I remain hopeful, because we don't fully understand yet why overdose fatalities have been declining in the first place. Even the most informed hope contains an element of faith. I'm happy if my red pill has some traces of blue.

Obesity

Our streets have been growing safer in recent years, at least in the sense that deaths from traffic accidents, violent crimes, and overdoses have been declining. These are largely independent health waves, driven by mostly different causes.

At the same time, calling them health waves relies on a broad conception of "health". In other respects, the U.S. may not be getting healthier.

In the Atlantic article that inspired this newsletter, Derek Thomson notes a recent decline in adult obesity. The evidence? An often-cited CDC report from last September.

Sad to say, the data in this report contradicts Mr. Thomson's reading of it. Here's the key figure:

Mr. Thomson focused on the right side of the graph, the period from 2017-March 2020 through August 2021-August 2023. But really, no matter where you look, those lines are pretty flat.

During that 2017-2023 time period, severe obesity (the lower, green line) increased from 9.2% to 9.7% of the population, a small but statistically significant increase. Obesity in general (the upper, dark blue line) decreased from 41.9% to 40.3%. According to CDC analyses, this change was not significant.

Mr. Thomson made two mistakes here: Ignoring the slight but significant rise in severe obesity, and overstating the importance of that slight, nonsignificant 1.6% drop in obesity overall.

Roughly speaking, "nonsignificant" in this context means there's too much uncertainty around the data to conclude that the change was genuine. We could just be seeing random fluctuation, measurement error, a shift in the composition of the sample, or something else.

This is a common mistake. Change is not the same as statistically significant change. If you take an IQ test and score 121, the next time you take the test you might score 120 or 122, but your IQ probably hasn't changed, nor would a statistical analysis reveal a significant difference in score. Analogously, the CDC data shows a decline in overall obesity, but not a significant one, and so we can't infer that obesity is generally declining.

A JAMA Health Forum study published last month revealed slightly different figures but a similar trend: A decline in adult obesity in the U.S. from 46% in 2022 to 45.4% in 2023. (Statistical comparisons weren't made.)

If there's progress here, it's that the CDC data and the JAMA paper both show that in 2023, for the first time in over a decade, adult obesity rates didn't increase over the previous year.

This leveling off is a source of hope, though it's not quite the same as saying that obesity rates are falling. The real takeaway here is that 40% or more of Americans are classified as obese. This is worrisome.

That 40% statistic isn't perfectly accurate, since it derives from BMI data, and the BMI is increasingly understood to be fallible (I discuss why here), but the general message is clear: Obesity is major public health problem in the U.S. Our current rates are approximately twice that of the OECD average.

The true rates are probably declining now, owing to the impact of Ozempic and other new weight-loss drugs, a trend that should become increasingly visible as data from 2024 and 2025 are explored. Although I find some hope in this, I also have mixed feelings, because I suspect that for some people, what's needed most is a change in lifestyle rather than the drugs.

Cherrypicking

Though I agree with much of what Mr. Thomson wrote, he did cherrypick his evidence. This is among the most common ways data ends up misrepresented.

If you want to claim that public health is improving, you can always find examples. But it's just as easy to find examples of growing problems. Acute illnesses like whooping cough, bird flu, and COVID-19 are on the rise now, as are some enduring problems like chronic pain. Can you really speak of a "health wave" when health is improving in some respects while declining in others?

The other problem with cherrypicking arises when you don't pick all the cherries. Here are three important health waves that Mr. Thomson didn't mention:

1. COVID-19 mortality.

As of 2024, COVID-19 was no longer among the top 10 leading causes of death in the U.S. Current rates are around 7 deaths per 100,000 people, down from 11.5 in 2023, and 44.5 in 2022.

2. STIs.

Two months ago, the CDC reported that sexually transmitted infections are declining after a nearly two-decade increase. (This is mainly attributable to a sharp reduction in gonorrhea, coupled with virtually no changes in syphilis rates after years of double-digit percentage increases.)

3. Cervical cancer.

Among the best of the good news from last year was a late November JAMA study showing a 65% rate of decline in cervical cancer among women under 25 from 2012 through 2019. This is the first cohort of women who were eligible for the HPV vaccine, and the majority did get vaccinated. The HPV vaccine deserves a place alongside many others (smallpox, polio, MMR, COVID-19, etc.) in the public health hall of fame.

Final thoughts

My takeaway is that the U.S. isn't experiencing some sort of overarching health wave, but rather a number of specific waves driven by many different forces. I'm encouraged by these changes, even if none of them are large in scope or fully understood. I hope that in at least some respects, 2025 is a good year for public health (and not just in the U.S.).

Next week I'll share what makes me hopeful with respect to personal health. I'll be discussing scientific progress as well as the unanticipated benefits of the MAHA dialogue. And I'll revisit the story of Opal Sandy, who's probably banging on something at this very moment.

Thanks for reading!