Dietary Reference Standards 123tainable at population level. If the standard were set
at the point of the average of all individual require-
ments, then half the population would have require-
ments in excess of the standard. In a normal distribu-
tion, some 2.5% of points lie at the upper and lower
tails outside that range of the mean plus or minus 2
SDs. Thus, by setting the RDA to this point of the
mean plus 2 SDs, we are setting the standard for
97.5% of the population. The consumption of most
nutrients at levels somewhat greater than actually
required is generally not harmful; hence, setting rec-
ommendations at the population average require-
ment plus a notional 2 SDs is logical if the aim is to
describe an intake that is adequate for almost every-
one. However, this is spectacularly inappropriate in
the case of recommendations for energy intake, since
even relatively small imbalances in energy intake over
expenditure will lead, over time, to overweight and
ultimately obesity, an increasing problem in most
populations. Recommendations for energy intake are
therefore given only as the estimated population
average requirement.
Thus, for almost half a century, these were the
terms used and the underlying conceptual approaches.
However, since the 1980s, changes have occurred in
both of these areas.Changes in terminology
Two basic changes occurred with regard to terminol-
ogy. The fi rst was that the term “recommended dietary
allowance” was altered and the second was that new
terms were introduced so that the adequacy of diets
could be evaluated from several perspectives. The
reason for changing the terminology was in effect to
re-emphasize some of the basic concepts underlying
the term RDA. “Recommended” has a prescriptive air
about it and there were concerns that consumers
might see this as something that had to be met daily
and met precisely. The term “allowance” reinforces
the perception of a prescriptive approach. Thus, the
UK adopted the term dietary reference value (DRV),
the EU introduced the term population reference
intake (PRI), the USA and Canada introduced the
term dietary reference intake (DRI), and Australia
and New Zealand now use the term nutrient intake
value (NIV). All are precisely equivalent to the origi-
nal concept of the RDA, a term that many countries
prefer to continue to use.
Two new terms were introduced: a minimum
requirement and a safe upper level. The minimum
requirement represents the average requirement
minus 2 SDs (point a in Figure 7.1). A defi nition
describing this point is given in Figure 7.1 along with
the various terms used to defi ne this point (Box 7.1).
The concept of an upper safe limit of intake has
gained importance in view of the increased opportu-
nity for people to consume high levels of nutrients
from fortifi ed foods or supplements. The recently
revised US DRI set “tolerable upper intake” levels
(ULs) that are judged to be the highest level of nutri-
ent intake that is likely to pose no risk of adverse
health effects in almost all individuals in a group. The
current European and UK recommendations also
address this concern in the case of those nutrients for
which toxic levels have been reported. The terms used
by different recommending bodies to describe the
various points on the distribution of individual
requirements for a nutrient are given in Box 7.2, while
precise defi nitions may be found in the relevant pub-
lications referred to.
The World Health Organization (WHO) has taken
a rather different approach, defi ning population safe
ranges of intake. “Normative requirement” is nowNumber of individualsNutrient requirementsa2.5%bc95% 2.5%Figure 7.1 Frequency distribution of individual requirements for a
nutrient. (a) The mean minus a notional 2 standard deviations (SDs);
intakes below this will be inadequate for nearly all of the population.
(b) The mean; the midpoint of the population’s requirement. (c) The
mean plus a notional 2 SDs; the intake that is adequate for nearly all
of the population. Note that, in practice, because insuffi cient data exist
to establish reliable means and SDs for many nutrient requirements,
the reference intakes describing the points a and c on the curve are
generally set, in the case of a, at the level that is judged to prevent
the appearance of signs of defi ciency (biochemical or clinical), and, in
the case of c, at the level above which all individuals appear to be
adequately supplied. Thus, it is unlikely that even 2.5% of the popula-
tion would not achieve adequacy at intake level c.