This is an interesting general question. It applies not just to Ramadan in Muslim communities but to almost any community since most communities practice both formal and informal routines of fasting and feasting. Christians have an extended fasting period (Lent) and may fast or feast on particular dates (e.g. Orthodox Christians fast on the day before Christmas) or days (e.g. Catholics may remove meat from their diet on Fridays, Orthodox Christians may remove meat and dairy products from their diet on Wednesday and Fridays). In my own, post-Chistian, culture there remains a tradition of Christmas and Easter feasts, “fish suppers” on Fridays (a meat-free day) and increased intake and diversity on Sundays (the “sunday lunch” at which meat, egg-batters, egg-custards, and sweet puddings are almost always eaten). Members of clubs, associations, trade unions, political parties may also have feast days. In my own village we have secular but communal feast two days a year in which all households contribute a dish to a feast. Weddings and funerals are usually feast days. National days also tend to be feast days. All major religions have fasting and feasting days (see [url=http://en.wikipedia.org/wiki/Fasting]this Wikipedia page[/url] for a brief review. Regarding IYCF and WDDS during Ramadan … this should be "robust to Ramadan” as fasting is obligatory for adults (not children) but pregnant and lactating women (most likely to be in our surveys) are usually exempt. Regarding DDS during Ramadan … it is common for non-Muslims to think of Ramadan as a period of fasting. It is, however, and period of both fasting and feasting. In some settings, the iftar (evening fast-breaking) meal may be more energy dense and diverse than outside of Ramadan. Since feasting and fasting are universal, IYCF and DDS indicator sets are routinely collected in communities and at times when people are feasting or fasting. It is, I think, important to know if this is happening. I usually ask: Was yesterday or is today a celebration or feast day when you eat special foods or eat more food than usual? Was yesterday or is today a fasting day when you eat special foods, avoid eating particular foods, eat less food, or have fewer meals than usual? as part of the DDS question set. I would then back this up with qualitative work (outside of the main survey) so that I can make a judgement with regard to the size and magnitude of bias that is introduced. This is then documented in the survey report. This is not really an issue of "scientific papers". It is an issue of standard epidemiological practice. The short answer … Yes but you need to apply some intelligence. I hope this is of some use.

Thanks Mark. But what if the entire data collection period would fall on the Ramadan period (28 June to 27 July) does it affect IYCF and Anthropometric Surveys?

I am sure that there will be some effect. The complication is working out what this will be (see below). The fact that the entire data collection period will fall within Ramadan is a mixed blessing. On one hand, it makes things easier as you will not have the complication of having two very different periods WRT mealtimes / diversity / &c. removing the need to make separate reports for data collected during and after Ramadan. On the other hand, a mixed period would make it easier to see the magnitude and direction of effect (i.e. you can examine the data from the Ramadan and post-Ramadan period (also see below)). Working out the effect … there are a few ways of doing this: (1) Have an additional “hypothetical” question set. This is the same questions preface by something like “Imagine that I were asking these questions outside of Ramadan for a normal day on which you were not feasting or fasting …”. There are problems with this approach. It is a hypothetical / counterfactual and we tend to avoid these unless we have to use them as they can confuse respondents. Also, we are unsure what is being reported … is a “normal day” a day without hunger or is it some recalled day. You will need to work out the wording of the preface to (e.g.) be indicate a recall for the week before Ramadan (this has its own problem of long and varying recall period). Another issue is that this makes for much longer interviews. (2) A simpler approach to (1) is to ask about differences between Ramadan and non-Ramadan. Something like “Are there foods that you eat during Ramadan that you do not usually eat at other times?”, “If so, what?” (with a checklist and free-text field), “Are there foods that you do not eat during Ramadan that you usually eat at other times?”, “If so, what?” (with a checklist and free-text field) and then use this data to “correct” the DDS / IYCF. (3) Do some qualitative work to get at the data that might be collected above. You might also do other semi-quantitative work on food-security, intra-HH differences in consumption, &c. at the same time. You could use both approaches. I would be tempted to do (3) before the survey to inform questionnaire design in the survey and use the simplest approach in (2). Adjustments are then based on two sets of data from disparate sources. Another approach would be to have a small survey after Ramadan that provides “calibrating" data. If you use scales / scores such as HDDS, WDDS or an IYCF score (e.g. ICFI as described in FEX 48) then you will only need a small sample size to look at differences in mean (or median) scores (WHO IYCF indicators would require a much larger sample size - note that you can apply ICFI to WHO IYCF data). A sample size of n = 60 would likely be enough to do this as well as to look for large Ramadan and post-Ramadam differences in consumption of individual food groups. The survey would only need to collect your DDS and IYCF (and perhaps anthropometry) data to do this. WRT Anthropometry … weight, height, and MUAC are affected by hydration so your adult anthropometry (assuming a day-time take) will be biased downwards (upwards for prevalence). The degree of bias will likely alter with time of day. BMI may be OK as height is lost during the day which may compensate for loss of weight due to dehydration but we cannot guarantee this (I am not a fan of BMI anyway). These should not be problems for child anthropometry as they are not likely to be deprived of water (as they are exempted from fasting). Older children may partially fast but I guess you will sample children 0-23 or 0-59 months. I hope this helps.

As per my very limited experience/observation, the household budget/expenditure during Ramadan is exceptionally very high compared to other months. As a result, the dietary consumption in terms of quality, quantity or variety is better. so, in view of this, is it advisable to conduct dietary assessment during this month? can we use the findings to compare with other months or seasons?

All survey assessments are snapshots. The survey discussed here will give a snapshot of a special time of year. The trick is to work out what makes it special and report that too.

Another issue raised by James Lual: It is always advisable not to consider survey during special occasions, to me special days like distribution, wedding and festivals should be exceptional because you might not find targeted groups at residents. Thanks James Lual

I agree with Mark, the results will only be a snapshot of a special period. This special period is a part of the diet consumed by your population in as much as it would be the period prior to a food distribution, or any other period. One has to provide the contextual information to better understand the findings. For what is worth, we collected food security and IYCF indicators from four Saharawi refugees camps in 2010 and 2012; both at a similar times (Oct/Nov), but with a relevant difference. In 2010 we finished the three weeks data collection period a day before the Ramadan period; whilst in 2012 we started the three weeks data collection the day after the Ramadan period. Both results looked comparable and we did not observed any systematic patterns in each of the surveys such as declining food consumption scores as the 2012 survey progressed, or a systematic increase of food-related IYCF indicators. Together our assessments represent (very crudely) a set of snapshots of a 6-weeks period with the special period in the middle, and we found no evidence of Ramadan affecting the patterns observed. My tuppence

@ Carlos Grijalva-Eternod – In 2010, you collected the data three weeks before the Ramadan period whereas in 2012 you collected your data after the Ramadan period. This means you collected before and after Ramadan. Do have data during Ramadan as this was the main agenda of the discussion. The other issue I want raise is regarding your study subjects. It was in refugee camp where the population is usually dependent entirely on aid. In Refugee setting, season might not be a major factor. I think the discussion should exclude refugee settings.

Thanks Kiross for clearly delimiting what issues should or should not be discuss in this tread. One correction to my last post. I wrote Ramadan but I meant to write Eid, an important but short celebration, the “Feast of the Sacrifice”. Saharawi refugees, despite having a strong reliance on food assistance for survival, have also access to food markets and have developed, during the last 39 years as refugees, an economic base where food commodities play an important role; and this economic base is seriously activated during feast periods like Eid. Now, I am no expert on assessing the impact feast rituals have on food security or IYCF indicators, but my example (and I insist, a crude one) was to show that the assumption that food rituals change food security or IYCF indicators might not be correct or as straightforward as it is sometimes assumed. For this reason, I am agreeing with Mark that one needs to “dig a bit more” to better understand what your results mean. I think James mentioned something more important, that is, during festivals you might have significant population movement. Consequently, you might find yourself in situations such as the people you’ll find in the household are not the usual residents or that people inhabiting the household are too busy preparing for the celebration that they have little time or energy to welcome your team to collect data. All these aspects should be discuss with the community before you start gearing up to implement the survey.

I did say it was an interesting question ... James : What you say makes a lot of sense but it is often not practical to avoid fasting and feasting days during a survey. An example is a survey of a sub-city or Addis Ababa which I worked on in March / April this year. It was a mixed Christian / Muslim area. Working five days a week (Monday to Friday) sampling would have been on fasting days on 2 from 5 days for Christian households. For Islamic household we would have had a varying part of the 24-hour recall period on the Sabbath. This is ignoring other days special to each community and special days for individual households (e.g. weddings, birthdays, christenings). On any day there will be a fraction of the population (and of your sample) for which it is special. We have, I think, no choice but to compromise but we should know that we are compromising and account for that. That said, given the choice and if practicable we should try to avoid special periods. Sometimes that is just not possible. Kiross : I do not think we can exclude camp populations. Special days happen in camps too. It is often assumed that camp populations are homogeneous with regard to (e.g.) dietary diversity and seasonality is unimportant. In my experience, this is often not the case. I tend to find the similar ranges of dietary diversities in camps and in "free-living" populations. This is because of wealth disparity, smallholdings, cash or food for labour both inside and outside of the camp, collection of wild foods, and sales or swaps of rations / relief items (in one case I saw a peak in DDS coincide with a bednet distribution). Some of these factors related to DDS in camp settings are seasonal. Carlos : I think you are right to pick up on James' point about absences and visitors. The rules of thumb that I use are that absences and refusals are recorded and reported. Visitors are included if they slept in the dwelling on the previous night.

Thank you guys for your inputs. This is of great help to us. : )

As we all know, majority of Muslims typically consume two meals per day during this period ; commonly one just before down and the other after sun set. According to the publications I had gone through there is no clear conclusions about the changes related with RAMADAN. Same concluded that it can lead to a decline in energy intake, loss of body weight and a state of dehydration, whereas the others reported that they did not find any significant changes in body weight or body composition during Ramadan. For some others the change was compounded by climatic changes as it mostly occur during summer. For example according to Chaouachi et al, there was no changes in BM or body fat during Ramadan in elite judo athletes. So, as to me, its better to expand study period and consider alterations during and after this particular period. Thanks