Dietary assessment and physical activity measurements toolkit

Dietary assessment - Diet variation

The term habitual intake is often referred to when describing the goal of dietary assessment; it is a key objective of many dietary assessments to estimate average long-term intake of particular foods or nutrients for the group or individuals under study (Rutishauser & Black, 2002). Habitual intake is exceedingly difficult to measure and only a robust dietary history or FFQ is designed to measure it. A prospective diary, or repeated 24 hour recalls can only be described as providing a ‘snapshot in time’ which is often used to reflect usual intake but may or may not be representative of habitual diet. Assessing more days increases the likelihood of capturing habitual diet.

In developed countries dietary intakes vary widely from day to day for many individuals through variations in meal patterns, household composition, the wide range of foods available, consumption of food outside the home, and working patterns. If the aim of a dietary assessment is to capture individual variation in intake only the diary method or repeated 24 hour recalls are suitable to take account of day to day variation, and only if sufficient days of intake are assessed. The number of days of measurement will be influenced by the dietary constituent being assessed and its intra-individual and inter-individual variability (Nelson et al, 1989) as well as whether group or individual analysis is required. Food frequency questionnaires and diet histories provide a summative description of intake over a period of time and therefore cannot detect day-to-day variation.

Diets may also vary due to seasonal effects.  In a study seeking to detect dietary change for example after a dietary intervention, it is important that assessments are undertaken at similar times of the year.  In cross sectional population surveys, or prospective cohort studies dietary assessments may be taken throughout the year to reduce the effect of seasonal variation.  It is possible that with the large amount of imported fruit and vegetables now available, seasonal variation is less than in years gone by.  However a dietary assessment undertaken in the few weeks of a year where certain fruit is widely available is likely to give a distorted view of micronutrient intake.

Between- and within-individual variation
Variance of the mean intake of a nutrient is dependent on between- and within-individual variation which may differ at different ages. Between- and within-individual variability can be estimated statistically using analysis of variance, providing two days or more of dietary intake are available in at least a subset of the population. If between-individual variation is large relative to within-individual variation, individuals can be readily distinguished so that the usual nutrient intakes of individuals can be characterised (Gibson, 2005). Unfortunately, for most nutrients there is greater variation within-individual than between (Beaton et al, 1997), which explains why the mean intake of a group is usually more readily assessed than individual. The use of a sufficiently large sample size, which is representative, reduces the effect of between-individual variation on group mean intakes.

Between-individual variation is likely to be higher if there is a wide age range in the study population; this is likely to be a greater effect than sex.  Hence, national surveys results are analysed and reported in age bands.

Within-individual variation is a measure of daily dietary variation and is often expressed as the coefficient of variation. Within-individual variation is reduced by increasing the number of days of assessment i.e. the greater the number of days the less the variation.

In 17 studies summarised by Bingham (1987), mean within-individual Coefficient Variation (CV) for various nutrients in adults were:

- Energy 23% - Carbohydrate 23% - Protein 27% - Fibre 31%
- Calcium and Iron 34% - Thiamine 39% - Riboflavin 44%
- Ascorbic acid 63% - Vitamin A 131%. 

The limited data on the weekly variation in intake was summarised by Bingham (1987) giving average CV for

- Energy 11% - Carbohydrate 11% - Protein 13% - Fat 15%    
- Iron 14% - Thiamine 16% - Riboflavin 18% - Vitamin C 25% - Vitamin A 52%.

The wider the variation the greater the number of days required to study. 

Duration of dietary records - how many days?
To capture the most representative data of usual intake as possible, the more days that can be recorded the better. However, an increased number of days can affect both recruitment or compliance. The first point to consider is whether analysis is to be at individual or group level, and then secondly to decide if a mean intake of the macronutrients is the primary outcome or is there a specific micronutrient of interest.

The classical review of Bingham (1987) is still referred to for guidance on the number of days to be assessed in adults and its dependence on the nutrient of interest.  The recommendations made in this review (Bingham, 1987) for the number of days recording (this could be in a prospective diary or repeated recall) are provided in Table 1; this is for group comparisons.

Macronutrients require the shortest number of recording days, whilst micronutrients, particularly those which occur in a very limited number of foods e.g. omega 3 fatty acids, vitamin A, require a much longer period of assessment. The use of a complementary dietary assessment such as a FFQ may capture these regular but infrequent food groups (Emmett, 2009). If mean intakes of individuals are required, more days recording is required, e.g. for 90% of the population 13 days of recording are necessary to determine the mean energy intake with a standard error of ± 10% (Bingham, 1987).

Table 2 shows the number of days of dietary information needed to classify 80% of individuals into the correct third of distribution.

In the UK, the current National Diet and Nutrition Survey has been reduced from seven days of assessment to four days (and changed from weighed to estimated) to improve recruitment. In a comparison study prior to the main study there was no difference in nutrient intakes when a four day estimated diary was compared to four repeated 24-hour recalls (Stephen et al, 2009).
The start day of a prospective diary or the recall days should be randomised, this means that when estimating means of reported intakes day-to-day variation has been accounted for. A 7-day diary overcomes the concern of day-to-day variation, but may increase individual burden.

To reduce the effect of variation as large a population as possible should be studied for the longest time that is feasible.

Day of the week
The biggest difference in dietary intake has often been assumed to be between weekdays and weekends, the assumption being that Saturday and Sunday were similar. Analysis from the most recent UK National Diet and Nutrition Surveys of adults and young people have shown that these two days vary greatly from each other, suggesting that both days may need to be assessed (Thane & Stephen, 2006a,b). 

In studies using interviewer administered dietary assessment, attention must be paid to the days of the week. Interviewers who do not work weekends or respondents who do not want to be interviewed on weekends for instance will cause a bias in the dietary data collected. 

In order to account for both within- and between-individual variation in dietary intake, assessments over as long a time period as is possible, and in as large a cohort as is feasible are the ideal. A clear focus on the primary outcome of the dietary assessment will dictate the importance of diet variation.



Web design by Studio 24