Sunday, November 30, 2014

Did Hunter-Gatherers Eat at least 100g of Fibre per Day?

One figure that gets mentioned occasionally (for example, by David Gatz, Jeff Leach and Marlene Zuk) is that hunter-gatherers consumed an average of 100g of fibre per day.  It’s often mentioned to argue that hunter-gatherers ate a plant-based diet and/or that Paleo dieters aren’t eating a ‘true’ Paleo diet unless they’re eating 100g of fibre
Estimating Fibre Intake of Hunter-Gatherers
Where does this figure come from?  Here’s what I came up with following a brief look at the literature:
·         In Eaton and Konner’s origin paper, ‘Paleolithic Nutrition — A Consideration of Its Nature and Current Implications’, they estimated that hunter-gatherers had an average fibre intake of 45.7g based on nutrient analysis of wild plant foods and plants providing 65% of total calories [1]
·         However, in a more recent paper Eaton, et al estimate hunter-gatherers consumed an average fibre intake of 86g [2]
·         “Analysis of vegetable foods consumed by foragers in this century (Table 1) and evaluation of archaic native American coproliths suggest that ancestral human fiber intake exceeded 100 g/d (Eaton 1990). Rural Chinese consume up to 77 g/d (Campbell and Chen 1994), rural Africans up to 120 g/d (Burkitt 1983)” [2]
·         Australian Aboriginals have been estimated to eat 40-80g of fibre based on the nutrient analysis of native plant foods and plants providing 20-40% of total calories [3]
·         Jeff Leach (Human Food Project) and Kristin Sobolik estimated a hunter-forager group in the northern Chihuahuan Desert consumed an average of 150-225g of fibre per day, with 135g coming from “inulin-like fructans” from plants such as agave, sotol and onion.  However, this hunter-forager group may be an outlier as those plants are unusually high in fructans.  The paper estimates that 25% of the fructans would be degraded by cooking [4]
It seems like there is some evidence to support the idea, but there is huge inconsistency too.  Also, these estimates should be taken with a grain of salt (quite literally, because the estimated sodium intake of hunter-gatherers is really low):
·         These estimates may not adequately factor for cooking and other processing
·         They may not approximate what plant foods hunter-gatherers actually ate.  For example: simply averaging all wild plants may overestimate vegetable intake, thereby overestimating fibre intake.  “This category included the leaves, leaf buds, pith and stalks of various plants. As a group they played a fairly minor role in AA diets”* [3]
·         Another reason to be somewhat sceptical is that other estimates of hunter-gatherer nutrient intake are unrealistic.  For example, the estimates for protein are very high – 19-37% [5], 34% [1], ~50% [3] – considering that maximal urea synthesis equates to a protein intake of 29.7–40.9% of total calories in a 3000 calorie diet [5].
Putting 100g of Fibre into Perspective
The Australian adequate intake (AI) for fibre is 30g for men and 25g for women.  This is based on the median population intake + a few grams because resistant starch isn’t measured as fibre [6] (this is consistent with measurements from the ABS [7]).  The suggested fibre intake to reduce chronic disease risk is 38g for men and 28g for women, which is intake of the 90th centile [6].
The US AI** for fibre is 14g/1000kcal (28g in a 2000 calorie diet, 42g in a 3000 calorie diet) [8], which is similar to the Australian recommendation.  However, average fibre intake in the US is ~15g [9], quite a bit lower than Australia
The current fibre recommendations and the average intake in both Australia and the US is a long way off 100g, but is this a realistic target?
Using data from the USDA nutrient database and the NUTTAB nutrient database I’ve estimated what it will take to meet 100g of fibre.  The numbers are the medians for a category of food and the number is in red if a food category exceeds 2000g or 3000 calories.  These numbers may be a slight overestimation if fructans and galactans are classified as sugars or starches in the nutrient databases rather than fibre.  I also excluded categories of food with low/no fibre (animal foods, beverages, junk foods) to improve readability.
Data from ~200 foods (raw) in the USDA nutrient database.  Download here
Weight (g)
Non-Starchy Vegetables
Starchy Vegetables
Sugary Fruit
Whole Grains
Immature Legumes
Mature Legumes
Data from the entire NUTTAB nutrient database.  Download here
Weight (g)
Herbs, Seasonings and Spices
Bread and Bread Products
Breakfast Cereals
Flours, Grains and Starches
Nuts and Seeds
Data from the entire USDA nutrient database.  Download here.  The USDA nutrient database may overestimate weight and calories needed to meet 100g of fibre because food categories often include both whole foods and food products.
Weight (g)
Spices and Herbs
Breakfast Cereals
Fruits and Fruit Juices
Vegetables and Vegetable Products
Nut and Seed Products
Legumes and Legume Products
Cereal Grains and Pasta
American Indian/Alaska Native Foods
It’s quite evident how extremely difficult it would be to get 100g of fibre with commonly available foods.  Almost all plant foods would need to be eaten in huge quantities, with a few exceptions.  Even still, to be most efficient, you would need to eat ~1000g and almost ~2000 calories of mature legumes to meet 100g of fibre, leaving little room for animal foods.  One might argue that plant foods these days are higher in sugar/starch and lower in fibre than those in the Paleolithic, but that doesn’t seem to be case [10].  So in conclusion I think it’s unlikely the average hunter-gatherer group consumed 100g fibre, particularly with plant foods providing 35% of total calories on average [5]
* Unfortunately Cordain, et al didn’t give a fibre estimate in the paper ‘Plant-animal subsistence ratios and macronutrient energy estimations in worldwide hunter-gatherer diets’.  Though I suspect it would have been similar to the estimate of Australian Aborigines given that the macronutrient composition of wild plant foods came from that paper [5] and because the both papers had a similar estimated energy intake from plant foods 20-40% [3] vs. 35% [5]
** The US adequate intake (AI) is defined as: the recommended average daily intake level based on observed or experimentally determined approximations or estimates of nutrient intake by a group (or groups) of apparently healthy people that are assumed to be adequate—used when an RDA cannot be determined.” 
*** Denis Burkitt may have been another source of the idea that hunter-gatherers consumed 100g of fibre per day, but unfortunately I can’t track down his peer reviewed papers.  His hypothesis that diverticulosis and colorectal cancer are due to low fibre and can be improved by increasing fibre is not well supported [11] [12] [13]


  1. Excellent post Steven.

    You, unlike Jeff Leach, seem to hold dear the idea of Falsification.

    It's truly amazing that the 'non-essential' aspect of fiber to humans is never clearly addressed by Jeff Leach et al. Now, does that mean we shouldn't eat fiber or it can't be conditionally beneficial? Absolutely not!

    For example, I think fiber may play a seasonal/periodical/conditional role in human immune modulation - maternal GOS being the obvious example. Essentially, I hold the hypothesis that many non-caloric or low-caloric plants may play a primary role for us as medicines...and a secondary role as foods. This hierarchy would be subject to change depending on the environment & availability of the essential proteins & fats.

    PS: anecdotally speaking, I eat a VLC &/or ketogenic diet by including substantial amounts of vegetables, a few nuts and fruit every few days (more in the summer). This is primarily for taste (hedonism, yay!) and variety. In a better food system variety would be more easily even with little to no plant matter. An interesting hypothetical.

    1. Thanks raphi. I agree that non-essentiality doesn't say too much about the health effect of fibre/etc. I'm with you, there's a lot of unknowns in nutrition science, but based on the role of gut bacteria I think some soluble fibre is likely to be health promoting. 'How much' would be a very difficult question to answer (although distended bellies and excessive gas = probably too much).

      I stayed clear of the health effects of fibre in this post, but there could be good fibre RCTs (for another day), though I suspect none of them would even approach the 100g that Jeff Leach recommends. He should do a trial, it would be very interesting

  2. I hope that (better quality) RCT comes about soon too.

    I used to speculate that in an Arctic environment the near absence of plant fiber might be 'OK' considering the different bacterial profile & overall load that is expected in such freezing temperatures. However, some Hadza & Mongol villages/tribes who also eat a nearly all meat/fish diet are in environments with much higher bacterial loads & profiles than we'd expect in the Arctic - so....could it be that thing like the cartilaginous soluble fibers of animals provide play that (still speculative but probable) role for gut bacteria substrate?

    It's interesting to consider whether we're simply obliged to live with a few pounds of bacteria in/on us & thus remain healthy DESPITE them (i.e. we manage them) or that their colonization is a causal factor in our health (providing us with a clear evolutionary advantage)? Could our obligate management be that evolutionary advantageous adaptation?

    More questions than answers...

    1. Are you thinking about:

      Germ free mice can provide some insight to that. They tend to be resistant to several diseases (not surprising that bacteria can be a weak link) but have lower BDNF and cognitive impairments too. So probably some adaptation/compensation and mutual backscratching involved.

  3. yes that post also comes to mind.

    GF-mice are an interesting case - 'mutual backscratching' is a nice way of putting it hehe

  4. I think the 100g fibers is from a study about the hadza peoples gut bacteria. Cant find it now but maybe you have better google skills than me.