You’ve probably seen a recent media blast about a study¹ that concluded that low carb diets shorten lifespans, saying they’ll take 4.5 years off of your life. Just what does “healthy” mean? Take a look back for a second through history and you’ll see that it’s varied wildly.
- Nows – Eat whole foods and exercise
- 2000s – Low carb! Atkins! Crossfit!
- 1990s – Aerobics! Low fat! Extra low fat!
- 1950s – Eat sugar and all these frozen dinners, especially what lobbyists told the government is healthy
- 1900s – Let’s put cocaine in everything to forget about Spanish Flu
- 1800s – Do I have a disease that will kill me? Nope? Awesome
- 1500s – I survived the disease that killed most of my town
- 1000s – We dodged arrows today and managed to not drink water from where we peed
- 500s – We’ve got food and now we’re Christian
- 000s – We’ve got food
- 100bc – “Let food be thy medicine and medicine be thy food” – Hippocrates tl;dr eat whole foods and exercise
Though this is of course a flippant glance at human mortality over the centuries, it’s reasonably accurate. For most of human history, the primary concern was getting enough food to eat. After that, healthy literally meant not being sick, as even a small infection could kill you. So examining human development over the years even in an extremely cursory way, you see that as we advance as a species, we die off differently. As human population swelled, war from encountering other civilizations took many lives, as well as spreading new and interesting diseases. So we’ve developed immunizations, antibiotics, antivirals, and all other manner of fighting infectious diseases. Border clashes, all-out war, and accidental ingestion of lead through hats has diminished significantly. It seems, however, that as our mortality rates from war, infectious disease, and occupational hazards has gone down, our incidences of diabetes and heart disease have gone up (though thankfully deaths have decreased)². Cancer rates increased in the 20th century, but cancer deaths have gone down as we approached the 21st century³ Healthy has had varied meaning, and today is no different. We exercise with the most recent science in our mind, stuff our faces with supplements, and walk a thin line between conventional nutrition wisdom and emerging data that flies completely in the face of that traditional wisdom. The traditional “wisdom” is that fat and cholesterol lead to death, but that’s not likely so: it wasn’t until the last decade or two that we were provided with research that refuted the notion that dietary cholesterol increases your blood cholesterol.⁴
Tl;Dr: the study
It’s inflammation – chronic and rampant – that creates the situation in our bodies that leads to heart attack, and chronic inflammation comes from insulin resistance (and diabetes), among other factors. Based on this, cutting sugar and carbs seems logical, right? So what’s the hullabaloo with this study? Well, part of it is media frenzy. Whenever a study comes out that claims something fairly contradictory to traditional wisdom or current cultural zeitgeist, it’s news-worthy. I’ve seen this article shared from a dozen different websites, but only a few actually have a link to the study. It’s much easier to piggy-back off of what other people have stirred up rather than dissecting the science yourself. It’s not a particularly long study, as far as these things go, so it reads alright. That said, you probably don’t want to comb through it (though if you do, it’s linked at the bottom), so I’ll do it for you. Is keto unhealthy? Let’s look at the data.
The study looked at 15,428 adults aged 45-64 in 4 U.S. communities: North Carolina, Minnesota, Maryland, and Mississippi. Between 1987 and 1989, they were asked 66 questions about their dietary habits, and then again 6 years later. The study then followed those people over 25 years to gather mortality rates.
While they controlled for a lot, including socioeconomic status, smoking status, diabetes status, age, and biological sex. They also only included self-reports of “non-extreme caloric intake” (which they refer to as above 500 and below 3600 calories a day). They gathered data well and included observations of variables that could potentially confound their results, including where the protein intake for each person originated (plant or animal).
They concluded that people on the extremes of the carb-intake spectrum were more likely to die of all-cause mortality than those in the middle. This is called a “U-shaped” association, because, well, it looks like a U.
Well, it’s been screamed everywhere from buzz-news outlets; low carb will kill you. Not really, because the study concludes directly that people with moderate carb intake – between 50-55% of their total energy intake – are the healthiest. There’s that word again.
Interpreting that conclusion
What this study did was ask people to recall their diets over the last 6 years, and then it followed them until they did or didn’t die. More accurately, it observed a TON of data, which is good, but it’s also somewhat flawed. Let’s start with what the study did right:
- Identified potentially confounding variables in health disease risks and their correlation with socio-economic status, school completion rates, income, diabetes risk, smoking history, and location.
- Had a varied and large sample size, excluding extremes of caloric intake – they figured out their target demographic and ensured adherence to that demographic.
- A thorough dietary intake questionnaire that also gathered information on lifestyle, including exercise, amount of sedentary time, etc.
They thoroughly gathered data and pieced it apart well. Now let’s look at what about the study was wrong:
- This is a cohort study, not a clinical trial, and as such, they relied on self-reported data, six years apart. I can’t remember what I ate a few days ago, let alone on average over the last 6 years. In addition, self-reports, even when done anonymously (these weren’t) tend towards response bias, or a tendency for people to lie. Sometimes it’s because they are embarrassed, or because they want to fall within what they expect the tester is looking for ideally. It’s difficult in this type of study to exclude that bias.
- The study doesn’t measure anything in regards to alcohol intake. The degree that alcohol intake can affect the body is pretty significant, especially metabolic practices within the liver. This can have a profound impact on overall health and is definitely a confounding variable they didn’t include for whatever reason. In addition, carbohydrate intake is one of the first things to drop when a person increases alcohol intake.
- Over 25 years, a person’s diet can vary wildly. If many of those people started out low carb and then went back to their standard American diets, it’s difficult to say they were truly eating low carb. People on low carb diets tend to vary in adherence as well, fluctuating between periods of low carb eating and “traditional” eating, fluctuating with stress, holidays, mental health, etc.
- Roughly 33% of all participants actively smoked – and in a time when smoking indoors was commonplace – and 30% more were previous smokers. That’s nearly 65% of all participants who had a history of smoking. It’s difficult to say how much this impacted the low carb group, but it’s definitely a large factor.
- People who were adhering to a low carb diet were likely overweight to begin with, and at 45, it’s difficult to assess to what degree atherosclerosis had already established itself in relation to when they started low carb. The study of course couldn’t assess that, as coronary calcium scanning didn’t exist then.
Modern studies indicate that ketosis – at least when established by a well-formulated, whole-foods diet that includes plants – has a direct, positive result on:
- Insulin sensitivity⁶
- Cardiovascular risk factors (in women within the target age group of the study)⁷
- Systemic inflammation (the likely leading cause of heart disease)⁸
Anecdotally I personally went from 450 pounds to 250 using a ketogenic lifestyle. I went from a cholesterol in the 300s to a normal cholesterol of around 101, increased hdl, lowered ldl, and I went from hypertensive to having great blood pressure. I’m not an outlier, though – plenty of people have the same results, and some in clinical trials (as above).
Conclusion (of this post)
Does low carb eating shorten your life? Probably not, especially if you go from overweight with cardiovascular and diabetic risk factors to having a healthy weight, reversing diabetes and inflammation, and lowering cardiac risk factors. Not smoking, low alcohol consumption, a diet consisting of varied and natural, whole foods, exercise, and stress reduction are all a part of a long, healthy life. Is keto healthy? Signs and studies point to yes. Also, if the media is passing around an article that claims something sweeping through correlation and not direct, observable causation, it’s probably bullocks. Resources ¹ https://doi.org/10.1016/S2468-2667(18)30135-X (the study in question) ² https://www.cdc.gov/mmwr/preview/mmwrhtml/mm4830a1.htm ³ https://www.cancer.org/latest-news/facts-and-figures-2018-rate-of-deaths-from-cancer-continues-decline.html ⁴ https://www.ncbi.nlm.nih.gov/pubmed/19751443 ⁵ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2928459/ ⁶ https://www.ncbi.nlm.nih.gov/pubmed/24176230 ⁷ https://www.ncbi.nlm.nih.gov/pubmed/26178720 ⁸ https://www.ncbi.nlm.nih.gov/pubmed/24021709