The paleo diet narrative is held up by two big-picture ideas: meat was essential to our evolution, and grain was a blow to our fitness. I've already shown how the "meat made us human" half of this meme fails to hold up under scrutiny. The meat-eating:big brain correlation looks less like causation every day. But how about the flip side of that claim: was grain really a death(ish) blow to the human race?
The basic claim here is that when Homo sapiens adopted grains into our lives through the invention of agriculture, it had a species-wide negative impact on our health and fitness. As evidence, grain blamers point to the infamous "Neolithic decline," the point in the fossil record when humans are markedly smaller than in previous times, and when researchers begin to note an increase in evidence of widespread nutritional stress and dental problems. This decline coincides geologically with the transition from the Pleistocene (the so-called "Ice Age") to the Holocene (the current geological period), which in turn coincides with increased grain consumption and invention of agriculture by humans (ca. 10,000 - 7,000 years ago). Hence, agriculture generally and grains specifically caused the decline.
But is correlation really causation in this case?
Here's how grain blamer Mark Sisson summarizes the evidence:
We in the Primal health community are quick to point out that agriculture reduced physical stature. Generally speaking, bone records indicate that Paleolithic (and, to a lesser extent, Mesolithic) humans were taller than humans living immediately after the advent of agriculture. Multiple sources exist, so let’s take a look at a couple of them before moving on:
According to one study on remains of early Europeans, prior to 16,000 BC, European males stood 179 cm tall, or 5’10.5″, and females stood 158 cm, or 5’2″. Between 8,000 to 6,600 BC, average heights had dropped to 166 cm for males. Heights fell even further in Neolithic populations, dropping down to 164 cm for males and 150 cm for females, only reaching and surpassing 170 cm at the end of the 19th century.
Another source found that Paleolithic humans living between 30,000 and 9,000 BC ran almost 5’10″, which is close to the average modern American male’s height. After agriculture was fully adopted, male height dropped to 161 cm, or 5’5.4″. Females went from 166.5 cm to 154.3 cm under the same parameters.This is a pretty dramatic change, to be sure. But take a close look at the data from Sisson's first source: the Paleolithic-to-Mesolithic stature decline in males was 13 cm. From the Mesolithic to the Neolithic, male stature decline was only an additional 2 cm. No data for Mesolithic females is offered, so it's a bit harder to evaluate the claim based on that data set; nonetheless, we can reasonably assume a similar pattern held for female humans across the Paleolithic-Mesolithic-Neolithic transition. If you read carefully, you should have noticed that most of the stature decline (at least in males) happened before the invention of agriculture.
There's a similar pattern in Sisson's second source: from the Late Paleolithic to the Mesolithic, average male stature declines a total of 4.6 cm. Similarly, female stature from the same transition drops a total of 6.8 cm. But from the Mesolithic to the Neolithic, male decline is only 2.9 cm, while female decline is 4.2 cm. Again, in both cases, there's a bigger decline from the Paleolithic to the Mesolithic than there is from the Mesolithic to the Neolithic. People started getting smaller before the invention of agriculture.
To be generous, one could still argue that the stature decline was a consequence of the invention of agriculture, since the Mesolithic was a transitional period, and agriculture probably wasn't the result of a single eureka moment, but rather a long process of aggregate discoveries or inventions building on previous ones. As we got more sedentary and plant-dependent, we also got smaller. That is, maybe the two processes were a feedback loop.
Or perhaps something else was going on...
We know these changes to height also reflected worsened health, because with shortness came dental pathologies like caries, plaque, and decay, signs of arrested growth indicating instances of severe malnutrition, and skull abnormalities that stem from iron deficiency. People got shorter, sicker, and less healthy. Height wasn’t a cause of poor health, of course, but it was an indicator.To his credit, Sisson doesn't say getting shorter is what made us sicker. His careful use of language here is commendable. But while it may be true that height is an indicator of health, this doesn't necessarily mean, as Sisson implies, that there's a causal relationship between agriculture, health and stature in prehistoric humans. Correlation ain't causation!
The pattern of pre-Neolithic stature decline isn't limited to Sisson's sources. There's evidence for it all over the Paleolithic fossil record. One study noted that the significant Upper Paleolithic-to-Mesolithic stature decline was correlated with a change in hunting strategies brought on by the invention of more efficient weapons and a shift to smaller, more docile game animals. This would extend the stature decline trend back to at least 50,000 years ago, an estimate that agrees with Ruff (2002), who notes that the trend of a decline in average body mass begins about 50 Ka, and is equivalent whether or not Neanderthals are included. Beals, et al., (1984) examining the fossil evidence from a thermoregulatory perspective, concluded that Homo have been de-encephalizing and getting shorter for about 100,000 years, which puts the start of the decline well into the Middle Paleolithic... quite a long time before the invention of agriculture.
Casting further doubt on Sisson's claim is the fact that this decline also struck culturally- and geographically-isolated hunter-gatherer populations who never invented agriculture. Brown (1987) and Brown (1992) examined a Pleistocene-to-Holocene decline in brain size and body stature in the Australian hominid record that's comparable to the one we see in Europe and central Asia. He noted that even if we can attribute the decline in the Eurasian record to agriculture, another explanation must hold for the Australian one, since agriculture was never invented by the prehistoric humans there.
And lest we be speciesist about this, it's worth noting that other taxa of animals went through a similar decline in stature and robusticity at the same period of time. A contemporaneous trend among bears and canids was the subject of a recent Ph.D. thesis at Duke University. Guthrie (2003) reported a pre-extinction rapid body size decline among Pleistocene horses in Alaska. Smith, et al. (1995) and Smith and Betancourt (1998) found that the Pleistocene-to-Holocene body size decline in woodrats is precise enough to be used as a paleotemperature proxy. Hill, et al., (2008) report that bison body size declined 15 to 20 percent from the late Pleistocene to the Holocene.
This multi-species pattern of body size decline was examined in great detail by Smith, et al. (2010), and their results were graphed on the table below. It covers the whole of mammalian evolution and diversification from our beginnings in the days of the dinosaurs up to the present, so this blog post's period of focus looks relatively small on the graph. Nonetheless, you can still see the pattern, especially if you look at the top half of the graph. The red line is the end-Cretaceous mass extinction that wiped out the non-avian dinosaurs. As you move rightward from this line, you are moving forward in time, approaching zero, which is the modern day. Just as you get to the end of the trendline, you'll see a curve downward. That's the transition from the Pleistocene to the Holocene. Mammals in general, not just humans, got smaller.
So, just what the what was going at the terminal Pleistocene to make mammals come up short? Based on my cited sources, you've probably intuited by now that it has something to do with climate change, and you'd be right... though it's a bit more chaotic than you might suppose.
Popular conception of the "Ice Age" depicts long millennia of worldwide arctic climate, followed by a trend of slow warming, where glaciers retreat in stately fashion back towards the poles and mountaintops, while the earth blooms in their wake, making way for us navel-gazing monkeys to take center stage and invent awesome stuff like spears and rockets and LOLcats. In short, most people think of the Pleistocene-Holocene transition as a move from steady cold to steady warmth. But in reality, it was more like a shift from screamo to Stradivarius.
At right is a snapshot of oxygen-18 isotope signals from a Greenland ice core sample (from Ditlevsen, et al. 1996), showing the climate patterns of the late Pleistocene and the early Holocene.
The low pass filter data shows that the Holocene is much less variable than the Pleistocene on time scales of 150 years or more. The high pass filter data shows the same thing on scales of 150 years or less. Ice cores like these can show climate fluctuations on a scale down to the decade level, giving us a pretty precise idea what climate was like in past times, at least for about the last 400,000 years or so. There's a large body of data from across the world, and using multiple indices and proxies, that corroborates the pattern seen in this graph (much of it cited in Richerson, et al., 2001). Most of the data at this resolution from the Pleistocene-Holocene transition shows that the Pleistocene was a period of chaos, with the climate see-sawing from one extreme to the other, often on human time-scales. The terminal Pleistocene was characterized by rapid climate flickering, with climate shifting decade-to-decade from cold and dry, to less cold and wet, and back again. Under these conditions, widespread droughts, devastating floods, massive wind storms and other climate extremes that modern people experience perhaps once a century would have been a lot more frequent, happening every 10 years or so in many places.
So, it's not just that the world was getting warmer, and animals getting smaller as a result (see Bergmann's rule and Allen's rule). It's also that the global climate was stabilizing after a long period of upheaval and chaos. The profile of selective pressures faced by our ancestors was changing in spectacular ways. That has some profoundly interesting implications for our understanding of the origins of agriculture, and raises some questions about the second part of Sisson's claim -- "with shortness came dental pathologies like caries, plaque, and decay, signs of arrested growth indicating instances of severe malnutrition, and skull abnormalities that stem from iron deficiency."
And You Thought I Forgot...
Let's get one thing settled before moving on. It's true that the human skeletal record changes dramatically at the Neolithic transition. The symptoms Sisson highlights are really there. The reasons why, though, are a bit more complicated than he makes them out to be.
Taking a few steps back and returning to the end-Pleistocene (aka, the Middle to Upper Paleolithic) for a moment, one thing becomes starkly apparent: agriculture would have been impossible in most places. Richerson, et al., (2001) note that the rapid climate fluctuations, low atmospheric CO2 content, oscillating atmospheric dust content, lower average rainfall, and various other chaotic environmental patterns of the end-Pleistocene posed an insurmountable network of interlocking constraints on any attempt to create a sedentary, pastoralist culture. Farming simply couldn't have been done, even if someone had been inclined to invent it. And there is some Paleolithic evidence that many cultures were, in fact, so inclined.
As Richerson, et al., note, the invention of agriculture in Neolithic societies was always preceded by an intensification of subsistence and increased reliance on the processing of low-quality plant and small game resources. But the Neolithic isn't the only time we see this in the record. An increase in plant-heavy subsistence patterns seems to be associated with climate pulses favoring warmer, wetter conditions across the Paleolithic and Mesolithic. The most famous, cited by Richerson, et al., is the Natufian culture of the Levant, who exploited a brief warm period in the Upper Paleolithic to become a semi-sedentary culture relying heavily on small game and grain foraging, along with nuts and fruits. When the Pleistocene climate flickered again and plunged the Levant back into a cold, dry period, the Natufians reverted to traditional hunting and gathering (Stutz et al., 2009). Only when the climate stabilized in the Holocene was the apparent human proclivity towards agriculture really able to take seed; in some ways, the Neolithic "invention" of agriculture wasn't really new at all.
The stable, warm, wet climate of the Holocene exerted a selective pressure for smaller phenotypes among most animals, not just humans. And humans in this situation did what they seem to have done many times before: shifted their trophic strategy to focus on more plants and smaller animals, a strategy that was probably homologous with their australopithecine and even earlier hominoid ancestors.
So far, so good. But what about Sisson's point? We didn't just get smaller, we also seem to have gotten sicker. Does agriculture, in and of itself, explain this? Were grains, in and of themselves, really to blame here?
Sisson's strongest case in this regard is probably the tooth problems. Most researchers agree that the increased carb intake of Neolithic societies played a role in the observed increase of dental caries, plaque and decay. But, there's a positive consequence here, too: Neolithic teeth have less wear than Paleolithic ones, because eating soft food places less stress on the chewing surfaces. This means less energy expended on food processing and, ultimately, on digestion.
Next is "signs of arrested growth indicating instances of severe malnutrition." Again, we can file this under "true, but misleading." Neolithic agriculture is characterized not simply by the adoption of grains, but (I'd argue) more importantly by over-emphasis on one or a few grain resources to the exclusion of other plants: rice in Asia, millet and sorghum in Africa, wheat in Europe and central Asia, corn in the Americas. Sisson's implied formula of "eat grains = get sick" isn't quite on the mark. It ought to read "specialize diet = improve chances of sickness." As diets diversify and societies become more adept at mixing various foods, health improves. There's no good reason to think a varied modern diet centered on whole foods -- whether omnivore or vegan -- is less healthy than anything cavepeople ate.
Finally, we have iron deficiency. This is actually the weakest of Sisson's points. I suppose he's trying to imply that agricultural foods necessarily increase risk of iron deficiency anemia. I call this a lack of imagination and a clear example of confirmation bias. After all, even in the modern world, helminth parasite infection is one of the three leading causes of iron deficiency anemia. Given that even the most impoverished developing communities today have access to more sanitary conditions than their ancestors did, there's no good reason to think that hookworm and other helminth infection rates would have been any lower among prehistoric peoples; in all likelihood, they were much higher than today. Skull abnormalities and other skeletal problems in Neolithic bodies might be evidence of nutritional deficiency... or they might be evidence of hookworm epidemics. Or, for that matter, of malaria, another leading cause of anemia. Most likely, some combination of these three factors explains the Neolithic skeletal record.
(As an interesting side note, helminth parasites eat cholesterol and counteract atherosclerosis, which might be part of the explanation for the lower lipid profiles of hunter-gatherer peoples.)
But there's another possible explanation for signs of increased iron deficiency in Neolithic skeletons, and ironcially, it points to what's actually pretty great about agriculture from an evolutionary point of view. That explanation is pregnancy.
The Neolithic witnessed a dramatic increase in birth rates among pastoralist populations. Because pregnancy imposes greater iron demands on women, and because women were having more babies in the Holocene, it's fair to hypothesize that the increase in signs of iron deficiency among Neolithic peoples reflects a failure of nutritional innovation to keep pace with improved fertility. And it's fertility that really matters here.
Evolution is fundamentally about sex, not food. Survival to reproductive age and production of offspring are the bottom line, and traits or behaviors that improve chances of success on these fronts will be selected for, while those that don't will be selected against. Selective pressures in the Holocene appear to have favored smaller, less robust human phenotypes, and agriculture is the strategy we invented to overcome those pressures. And we did it dramatically well. Everything else is a trade-off.
Seen in this light, agriculture is probably the single greatest boon humans have ever experienced to our fitness. It's what enabled us to survive the new Holocene selective pressures and expand our range to every corner of the world, inventing awesome stuff like spears and rockets and LOLcats along the way. That's not to say we didn't mess things up, too, and it would be crass to ignore the negative impacts we've had on the world and each other, let alone to justify them. But even these infractions are consequences of our fitness, not a demonstration of our lack of it.
The claim that agriculture caused a decline in human stature doesn't hold up to the evidence, and the indictment of grains, in and of themselves, as a detriment to human fitness is far too simplistic to be taken seriously. Correlation ain't causation, grain blamers.
-- Beals, K.L., Smith, C.L., and Dodd, S.M. (1984). Brain size, cranial morphology, climate and time machines. Current Anthropology 25: 303-330.
-- Brown, P. (1987). Pleistocene homogeneity and Holocene size reduction: the Australian human
skeletal evidence. Archaeology in Oceania 22, 47–71.
-- Brown, P. (1992) Recent human evolution in East Asia and Australasia. Philosophical Transactions of the Royal Society B 337, 235–242.
-- Ditlevsen, P.D., et. al. (1996). Contrasting atmospheric and climate dynamics of the last-glacial and Holocene periods. Nature 379, 810-812.
-- Guthrie R.D. (2003). Rapid body size decline in Alaskan Pleistocene horses before extinction. Nature 426, 169-171.
-- Hill, M. E., M. G. Hill, et al. (2008). Late Quaternary Bison diminution on the Great Plains of North America: evaluating the role of human hunting versus climate change. Quaternary Science Reviews 27, 1752–1771.
-- Richerson, P.J., Boyd, Robert and Bettinger, R.L. (2001). Was agriculture impossible during the Pleistocene but mandatory during the Holocene? A climate change hypothesis. American Antiquity, Vol. 66, No. 3, 387-411.
-- Ruff C.B. (2002). Variation in human body size and shape. Annual Review of Anthropology 31:211–232.
-- Smith, F.A., Betancourt, J.L. & Brown, J.H. (1995). Evolution of body size in the woodrat over the past 25,000 years of climate change. Science 270, 2012–2014.
-- Smith, F.A. and Betancourt, J.L. (1998). Response of bushytailed woodrats (Neotoma cinerea ) to late Quaternary climatic change in the Colorado Plateau. Quaternary Research 50: 1-11.
-- Smith, F.A., et.al. (2010). The evolution of maximum body size in terrestrial mammals. Science 330, 1216-1219.
-- Stutz, J.A., Munro, N.D., and Bar-Oz, G. (2009). Increasing the resolution of the Broad Spectrum Revolution in the Southern Levantine Epipaleolithic (19–12 ka). Journal Of Human Evolution. doi:10.1016/j.jhevol.2008.10.004