Food consumption studies
Food consumption studies Arnold Sondern 26.11.98
Re: Arnold Sondern's email Tonia Marek 27.11.98
a comment on dietary measurement Beatrice L. Rogers 10.12.98

From: "Arnold Sondern" <>

Date: Thu, 26 Nov 1998 08:35:40 +0300

Subject: Ngonut: food consumption studies


Dear Sir, Mrs, Ms,,

In the framework of refining some elements of our Food Economy Analysis in Somalia and of the periodic food needs assessments, we would like to look more into depth into variations in food consumption for the different food economy groups.

The result should be a form of a seasonal calendar of food consumption. It should allow us to have an idea of variations in caloric & protein intake and also of the variations in quality of the diet with regard to main micronutrients (vit A, C). The methodology that suits us is one that could be carried out over a short period of time, giving information about a one year period.

Resulting data should permit to be used to refine adjustment of period food need requirements that we submit to WFP and donors.

The same accounts for physical activity calendars that influence food needs of populations.

It is the idea that we should establish within a short period something of a baseline, that over time using another, finer methodology , could be refined and checked. Food economy permits some data collection about it (determining sources of income and sources of food).

I have the following questions:

1. Do you have the title(s) of (a) publication(s) about food consumption measurement and/or physical activity measurement methodologies that might suit our target? Or even have an electronic copy?

Related to this issue:

2. Are there databases with nutrition publications that can be consulted through Internet, in general but also databases where I likely can get more info on these subjects?

Thanks in advance for your assistance


Arnold Sondern

Food Security Assessment Unit for Somalia Nairobi-Kenya

Date: Fri, 27 Nov 1998 13:10:47 +0000

From: Tonia Marek

Subject: Ngonut: Re: Arnold Sondern's email


Dear Mr Sondern,

A while back I undertook surveys using the "food frequency" method, which would give you a good indication of the diversity of the diet and its evolution (i) over time and between seasons; (ii) among different socio-economic groups. I am sorry, but I don't have my papers with me, however frequency counts and analysis is no rocket science. It will tell you if the conditions improve for the poorer/vulnerable groups or not. The advantage of this method is that it's easy to carry out, you go to each household for an interview on food recall with a standard questionnaire, and you do it at every season.

This method won't tell you much about the amounts of calories, proteins, vitamins ingested. If you want to know the amount of calories, you need to also weigh the food they eat and translate that into nutrients with a regular food table (don't forget to ask about wild food gathered and eaten).

However, if your objective is "to need requirements that we submit to WFP and donors", I would recommend the following:

i) find out what the nutritional status of children/women is in your population and identify the malnourished groups (quantify them); there probably is some kind of data available from surveys you can use. I'd use anthropometric indicators and if you suspect micronutrient deficiency do a survey for that too (vitamin A through clinical check for Bitot spots, for example)

ii) use the usual food requirement tables to assess the needs of this target group (don't undertake a physical activity measurement study).

iii) undertake a rapid food recall survey to assess the amount/variety of food eaten by the target group to find out approximately the calories they eat and any major lack of micronutrients (as first explained above).

iv) translate the gap into tons of food for donors.

Don't forget that the best is sometimes the ennemy of the better. If you want to act quickly, don't try to have perfect exact figures, make estimates based on your experience and that of other colleagues, on the quick surveys, on available information, and you'll probably be right (plus or minus 10-15%).

Now, all this if for non-emergency situations. In emergency you will use slightly different methods, and I assumed you were not in an emergency situation.

I also forwarded your email to a colleague of mine, Thierry Brun, who might give you more specific references.

I hope I answered parts of your questions. Let me know if you think I can be of further help.


Tonia Marek, Senior Public Health Specialist

Date: Thu, 10 Dec 1998 09:50:02 +0000

From: "Beatrice L. Rogers" <>

Subject: Ngonut: a comment on dietary measurement


I read Tonia Marek's comments, and completely agree on the differing uses of food frequency methods versus more rigorous quantification. But I wonder if weighing is always the preferred method for doing the quantification when it is needed.

My colleague Anne Swindale (Academy for Educational Development, Washington DC) has written a manual that documents a method for quantifying food intake using volumetric measures of food (cooked or uncooked, depending on how the respondent can best provide the information)...for each food, conversion factors are calculated to go from the cooked to the raw form if needed, and convert from volume to weight in order to get gram amounts for conversion to nutrients. The volumetric estimates are obtained using rice to estimate quantity of solids like beans, rice, oatmeal; water for liquids. The respondent puts these substances into her own receptacles (bowls, cups, spoons) for more accurate judgment of the quantity prepared or served, and then these amounts are poured into graduated beakers for accurate milliliter measurement. Obviously, there are different conversion factors from volume to raw weight for each food. We have used this method in a couple of surveys, and she has used it in several more. It avoids the very intrusive need to weigh food as it is being prepared or consumed. (Weighing also takes more time than the volume method.) Solids like meat and cheese are estimated using clay, which the respondent forms; the shape is then submerged in a beaker of water; the displaced water gives the volumetric measure. Irregular shapes like fish, cassava, are estimated using two dimensional cutouts.

There are problems with every method I know of, but my sense is that weighing can be quite cumbersome compared to this recall method of estimation.