A modern take on the Clara M. Davis paper

You know, the one I said you should read because it is unbelievably interesting and most of you never bothered because you’re too busy steaming carrot sticks? Honestly, what are you mothers of young babies like? It’s almost like you can’t find the time to read half-bonkers old research that is only tangentially linked to BLW…

So I asked my good friend Amanda Kvalsvig, who is the proud possessor of a fine scientific mind as well as being a confirmed BLWer, to have a gander at the documents and interpret them from a more modern standpoint. Because I am aware that I approached them from more of a gawping ‘eeeeuw, cod liver oil… yuk yuk sheep’s brains etc’ perspective and am therefore unqualified to comment.

This report is based on two papers:
* Davis CM. Results of the self-selection of diets by young children. Can Med Assoc J 1939;41:257-61
* Strauss, Stephen. Clara M. Davis and the wisdom of letting children choose their own diets. CMAJ 2006 175: 1199

What was Dr Davis’ experiment?

Her main research question was “What will babies eat if they’re given free choice?”

To answer this, she observed 15 children over a period of 6 years. Every day, they were offered a selection of 33 different foods, and observers noted what they chose to eat and how much they ate. A key point of the experiment was that mealtimes were set up to prevent the staff from influencing what the babies ate.

The diets of the children were then analysed and compared with the standard diet recommended for children at the time (the 1920s). The children’s health was monitored by physical and biochemical examinations and by X-rays of their bones.

What did she find?

When the children’s food choices were analysed, they were found to be very different from each other and from the standard diet of the time. But, interestingly, when the diets were examined as a whole, they showed similar patterns of energy intake and proportions of proteins, carbohydrate and fats.

Dr Davis reports that the children were very healthy and were remarkably free of the common childhood diseases of the time. Five of the children had rickets at the start of the experiment, and the rickets resolved without medical treatment – apart from one child with severe rickets who spontaneously drank cod liver oil. The radiologist at the children’s hospital was very excited at the “beautiful calcification” of the bones on the children’s X rays.

Although some of the children were undernourished when they arrived, this evened out until there were no notably fat or thin children.

How does the experiment look to us in 2007?

Ironically, one of the things that makes this experiment very interesting and important is the fact that there is no chance of it ever being repeated, because experimenting on institutionalised children in this way is now agreed to be totally unethical.

There are other problems with her study from the science point of view. One of the main problems is the very small sample size: only 15 children were studied. What this means is that we need to be very cautious about applying the findings to all children. It’s possible that in a larger group of children there would have been some who didn’t thrive.

Another difficulty is the lack of a control group. It’s likely that factors other than diet contributed to the health of these children. There should have been a control group of children selected in the same way, living in the same place, with lives identical to the experimental children in every way except diet. That way, any differences between the groups could clearly be attributed to diet.

Clara Davis and other health professionals were very impressed with the health of the children, but a modern researcher would be concerned about the possibility of observer bias. Humans tend to see what they expect to see, especially if they have a personal interest in what they’re seeing. In a modern study, the children’s health would be monitored by independent observers who didn’t know anything about their diet.

In addition to the design and conduct of the study, to a modern researcher there are concerns about the way the results were reported. Dr Davis describes the study and gives us her conclusions, but doesn’t present the results in any detail. Would we have drawn the same conclusions from her results? We can’t tell.

This is important because there have been many changes in knowledge and practice of nutrition. A modern researcher looking at Clara Davis’ results might well put a different interpretation on them.

Should we just disregard the study altogether then?

Absolutely not. Even with the above reservations, it’s very interesting to see that these children managed to choose a healthy diet on their own. Although all the foods on offer were considered healthy (some wouldn’t be recommended nowadays), the children could easily have selected an unhealthy diet by eating just 2 or 3 favourite foods – yet they didn’t.

There’s a strong biological plausibility to the idea that children will instinctively choose a balanced diet. In fact, the survival advantage is so clear that it would be surprising if children didn’t do this. The only problem is that modern society has developed a number of foodstuffs which might deceive that instinct by being, for example, intensely sweet without having the nutritional benefits of naturally sweet foods such as fruit and root vegetables.

Clara Davis’ experiment worked because the food selected by the children had already been pre-selected to include only natural flavours. This 2-tier system is similar to what happens in baby-led weaning: babies choose whether or not to eat something, but their parents determine what’s on offer.

So was Clara Davis a BLW-er?

Yes and no. In one respect, this was very definitely baby–led weaning, because the babies had a free choice of the type and quantity of food that they ate.

However, Dr Davis was primarily interested in appetite and diet (what they ate) rather than feeding (how they ate). The food was “finely cut, mashed and ground” to make it suitable for spoonfeeding (although it seems some babies just used their hands anyway). This meant that the babies weren’t limited by their ability to pick up or chew the food they chose. In the end this might not have made much difference, as these babies were weaned at an older age than modern babies and had (as far as we know) normal development, so it’s likely that they would have coped just as well if presented with lumpy food.

12 Responses to “A modern take on the Clara M. Davis paper”

  1. Andrea says:

    Hi there,

    this is actually quite interesting.
    If you don’t mind I would translate it for my blog (with a link to yours, although I doubt it will attract very many new customers:).
    It should make interesting reading

  2. Shannon says:

    Clara Davis was my great-grandmother and it’s amazing to me to see her work analyzed in a modern setting like this. I think the points you’ve made about the flaws in her experiment are very insightful however your comment about these babies being weaned at an older age are not 100% true. Several of the babies were very young when they came to her study. My grandfather, for instance, was barely 6 months old when he and his older brother were turned over to my great-grandmother for her experiment.

    • Aitch says:

      My heavens, Shannon, what an extraordinary woman to have in the family! And your folks were involved in the experiment, that’s amazing! Thanks for swinging by, great to have your insight on the paper (for the purposes of our interest, weaning at 6 months is a-okay). Do tell us anything else you know about her…

  3. I am the Stephen Strauss of the paper cited at the beginning of this piece and I would love to connect with Shannon. I am (finally) going to publish a book on Clara’s research in the coming year which will, I think, address numbers of the questions raised here. It should also introduce some Davis points people aren’t familiar with.

  4. Serina says:

    Any idea what the 33 foods were?

    • Aitch says:

      yes, it’s on the link to the paper itself. Lambs brains were, if i recall correctly, prominently features. Blerch. Davis herself acknowledged that they were broadly healthy, that she wasn’t offering junk.

  5. Nina Planck says:

    I wrote about the Clara Davis experiments in Real Food for Mother and Baby and advise mothers to give babies a range of foods and early.

    I would so much like to be informed when the new papers come out, and what a thrill to hear from her great grand daughter.

    Count me in the conversation if you can. (My agent’s email is on my site at http://www.NinaPlanck.com, and she’ll send you to me.) Good luck! Nina

  6. […] to cultural norms over her body. I think it is amazing that if given a selection of healthy food babies will choose to eat balanced and healthy over the course of a week, meeting all of their dietary needs. I think they much more in tune with […]

Leave a Reply