Appendix I
Dietary Intake Data and Calculation of the Target Median Intake for Iron
This appendix contains information on how schoolchildren’s dietary intakes compare with Estimated Average Requirements and data and a description of the use of the probability method to calculate the Target Median Intake for iron for adolescent females.
LIST OF TABLES
-
Table I-1 Estimated Average Requirements (EARs) for Schoolchildren and Reported Nutrient Intakes at the 5th Percentile and Median by Age-Grade Group and Gender
-
Table I-2 Iron Intake Distribution for 14–18-Year-Old Female Participants (mg/d)
-
Table I-3 Iron Intake Distribution for 11–13-Year-Old Female Participants (mg/d)
-
Table I-4 Iron Requirement Distribution for 14–18-Year-Old Females (mg/d)
-
Table I-5 Iron Requirement Distribution for 9–13-Year-Old Females (mg/d)
-
Table I-6 Estimated Iron Requirement Distribution for 11–13-Year-Old Females (mg/d)
TABLE I-1 Estimated Average Requirements (EARs) for Schoolchildren and Reported Nutrient Intakes at the 5th Percentile and Median by Age-Grade Group and Gender
Nutrient |
6–10 years |
11–13 years |
14–18 years |
|||
Males (n=295) |
Females (n=317) |
Males (n=342) |
Females (n=342) |
Males (n=506) |
Females (n=512) |
|
Protein (g/kg/d) |
|
|
|
|
|
|
EAR |
0.76 |
0.76 |
0.76 |
0.76 |
0.73 |
.71 |
Intake at 5th |
1.5 |
1.5 |
1.1 |
0.7 |
0.9 |
0.5 |
Median Intake |
2.4 |
2.3 |
1.8 |
1.4 |
1.5 |
1.1 |
Vitamin A (μg RAE/d) |
|
|
|
|
|
|
EAR |
343 |
333 |
445 |
445 |
630 |
485 |
Intake at 5th |
352 |
367 |
373 |
236 |
280 |
175 |
Median Intake |
631 |
614 |
689 |
529 |
635 |
439 |
Vitamin C (mg/d) |
|
|
|
|
|
|
EAR |
29 |
29 |
39 |
39 |
63 |
56 |
Intake at 5th |
36 |
48 |
43 |
24 |
32 |
19 |
Median Intake |
83 |
90 |
92 |
73 |
90 |
67 |
Vitamin E (mg αT/d) |
|
|
|
|
|
|
EAR |
7.2 |
7.2 |
9.0 |
9.0 |
12.0 |
12.0 |
Intake at 5th |
4.9 |
3.4 |
4.1 |
2.6 |
4.2 |
2.6 |
Median Intake |
6.0 |
5.2 |
6.5 |
5.4 |
7.2 |
5.3 |
Thiamin (mg/d) |
|
|
|
|
|
|
EAR |
0.6 |
0.6 |
0.7 |
0.7 |
1.0 |
0.9 |
Intake at 5th |
1.0 |
1.0 |
1.2 |
0.7 |
1.1 |
0.7 |
Median Intake |
1.6 |
1.5 |
1.7 |
1.4 |
1.9 |
1.3 |
Riboflavin (mg/d) |
|
|
|
|
|
|
EAR |
0.6 |
0.6 |
0.8 |
0.8 |
1.1 |
0.9 |
Intake at 5th |
1.4 |
1.4 |
1.6 |
1.0 |
1.4 |
0.8 |
Median Intake |
2.3 |
2.2 |
2.5 |
2.0 |
2.6 |
1.7 |
Niacin (mg/d) |
|
|
|
|
|
|
EAR |
7.2 |
7.2 |
9.0 |
9.0 |
12.0 |
11.0 |
Intake at 5th |
13.9 |
12.9 |
15.1 |
10.8 |
18.1 |
9.6 |
Median Intake |
20.5 |
19.9 |
22.5 |
19.6 |
27.1 |
18.2 |
Vitamin B6 (mg/d) |
|
|
|
|
|
|
EAR |
0.6 |
0.6 |
0.8 |
0.8 |
1.1 |
1.0 |
Intake at 5th |
1.2 |
1.0 |
1.2 |
0.8 |
1.5 |
0.7 |
Median Intake |
1.7 |
1.6 |
1.9 |
1.6 |
2.2 |
1.4 |
Folate (μg DFE/d) |
|
|
|
|
|
|
EAR |
196 |
196 |
250 |
250 |
330 |
330 |
Intake at 5th |
310 |
322 |
415 |
228 |
361 |
219 |
Median Intake |
553 |
536 |
640 |
477 |
647 |
442 |
Vitamin B12 (μg/d) |
|
|
|
|
|
|
EAR |
1.2 |
1.2 |
1.5 |
1.5 |
2.0 |
2.0 |
Intake at 5th |
2.5 |
2.5 |
3.9 |
2.0 |
3.0 |
1.4 |
Median Intake |
5.1 |
4.6 |
6.0 |
4.5 |
6.1 |
3.8 |
Iron (mg/d) |
|
|
|
|
|
|
EAR |
4.8 |
4.7 |
5.9 |
5.7a |
7.7 |
7.9 |
Intake at 5th |
8.5 |
8.6 |
10.9 |
6.9 |
10.6 |
6.0 |
Median Intake |
14.6 |
13.9 |
16.2 |
13.3 |
17.9 |
11.8 |
Nutrient |
6–10 years |
11–13 years |
14–18 years |
|||
Males (n=295) |
Females (n=317) |
Males (n=342) |
Females (n=342) |
Males (n=506) |
Females (n=512) |
|
Magnesium (mg/d) |
|
|
|
|
|
|
EAR |
146 |
146 |
200 |
200 |
340 |
300 |
Intake at 5th |
165 |
172 |
181 |
134 |
182 |
110 |
Median Intake |
253 |
236 |
266 |
223 |
291 |
206 |
Zinc (mg/d) |
|
|
|
|
|
|
EAR |
7.3 |
5.2 |
5.2 |
7.0 |
7.0 |
8.5 |
Intake at 5th |
6.9 |
6.5 |
8.7 |
5.9 |
8.0 |
4.7 |
Median Intake |
11.1 |
10.0 |
12.4 |
9.9 |
14.2 |
9.1 |
Phosphorus (mg/d) |
|
|
|
|
|
|
EAR |
665 |
665 |
1,055 |
1,055 |
1,055 |
1,055 |
Intake at 5th |
874 |
917 |
1082 |
636 |
971 |
597 |
Median Intake |
1,376 |
1,281 |
1,483 |
1,171 |
1,622 |
1,087 |
NOTES: αT = α-tocopherol; d = day; DFE = dietary folate equivalents; g = gram; kg = kilogram; mg = milligrams; n = sample size; RAE = retinol activity equivalents; μg = micrograms. aThe committee used a reference value of 7.5 mg for girls ages 11–13 years, as explained under “Iron Status” in Chapter 3. SOURCES: Weighted tabulations of data from the third School Nutrition Dietary Assessment study (SNDA-III) (USDA/FNS, 2007a); adapted from Table VI.16 in Volume II and tables in Appendix J to Volume II. Dietary intake data (24-hour recalls) were collected during the 2004–2005 school year and do not include intakes from dietary supplements (e.g., multivitamin-multimineral preparations). The personal computer version of the Software for Intake Distribution Estimation (PC-SIDE; ISU, 1997) was used to estimate the usual nutrient intake distributions and the percentage of children with usual intakes below the EARs. The EARs used in the analysis were from the DRI reports (IOM, 1997, 1998, 2000a, 2001, 2002/2005). EARs shown for the males and females ages 6–10 years are weighted averages of two DRI age groups. Bolded numbers indicate that intake at the 5th percentile is below the EAR. |
CALCULATION OF THE TARGET MEDIAN INTAKES FOR IRON
The Probability Approach for Calculating the Prevalence of Inadequacy
The distribution of iron requirements has been estimated using factorial models based on component losses and the deposition of iron. Since it was expected that the distribution was not normal, the distribution was estimated using simulation of a population of 100,000 individuals (IOM, 2000b, p. 569). A consequence of the nonnormality of the requirement distribution is that the Estimated Average Requirement (EAR) cut-point method does not provide a sufficiently accurate estimate of the prevalence of inadequacy, particularly for menstruating women. The recommended alternative is to use the probability approach (IOM, 2000b, pp. 205–208).
The basic idea underlying the probability approach is most easily
understood in terms of a large population of individuals with known intakes. For each individual, the probability of inadequacy is calculated from the requirement distribution (i.e., the probability that the requirement is greater than the individual’s intake). These probabilities are averaged over all individuals in the population to give the prevalence of inadequacy.
The two inputs for the calculation are the intake distribution and the requirement distribution. Let FR(r) and FI(i) denote the cumulative distribution functions for requirement and intake, respectively. The prevalence of inadequacy is the probability that the intake, I, is less than or equal to the requirement, R, that is, P(I ≤ R). In terms of the cumulative distribution functions, we have the following expression for the prevalence of inadequacy:
Let x1 ≤ x2 ≤ L ≤ xn denote an ordered set of intakes that span the range of the distribution. The probability of inadequacy can be approximated by
Intake and Requirement Distributions
Calculations using this method were performed for 14–18-year-old females and 11–13-year-old females. The intake distribution was based on National School Lunch participants included in the third School Nutrition Dietary Assessment study. For the 14–18-year-old female participants, the intake distribution is in Table I-2.
For the 11–13-year-old female participants, the intake distribution is shown in Table I-3.
The requirement distributions for iron are given by IOM (2001). For 14–18-year-old females, the requirement distribution is shown in Table I-4. For 11–13-year-old females, the requirement distribution is not available but the requirement distribution is given for 9–13-year-old females (see Table I-5).
Because the 11–13-year-old females will have a higher percentage of menstruating females than the 9–13-year-old females, the iron requirements
TABLE I-2 Iron Intake Distribution for 14–18-Year-Old Female Participants (mg/d)
TABLE I-3 Iron Intake Distribution for 11–13-Year-Old Female Participants (mg/d)
TABLE I-4 Iron Requirement Distribution for 14–18-Year-Old Females (mg/d)
Percentile |
0.025 |
0.05 |
0.1 |
0.2 |
0.3 |
0.4 |
0.5 |
0.6 |
0.7 |
0.8 |
0.9 |
0.95 |
0.975 |
Requirement |
4.63 |
5.06 |
5.61 |
6.31 |
6.87 |
7.39 |
7.91 |
8.43 |
9.15 |
10.03 |
11.54 |
13.08 |
14.8 |
NOTE: mg/d = milligrams/day. |
TABLE I-5 Iron Requirement Distribution for 9–13-Year-Old Females (mg/d)
Percentile |
0.025 |
0.05 |
0.1 |
0.2 |
0.3 |
0.4 |
0.5 |
0.6 |
0.7 |
0.8 |
0.9 |
0.95 |
0.975 |
Requirement |
3.24 |
3.6 |
4.04 |
4.59 |
4.98 |
5.33 |
5.66 |
6.0 |
6.36 |
6.78 |
7.38 |
7.88 |
8.34 |
NOTE: mg/d = milligrams/day. |
for 11–13-year-old females are higher than those for 9–12-year-old females and the shape of the distribution is likely to be skewed to the right, as is the distribution for 14–18-year-old females. Therefore, an estimated requirement distribution for 11–13-year-old females was computed by setting the EAR at 7.46 (versus 5.66 for 9–13-year-old females and 7.91 for 14–18-year-old females) and using the shape of the distribution for 14–18-year-old females. Thus, the estimated requirement distribution for 11–13-year-old females was obtained by subtracting 0.45 (7.91–7.46) from each of the percentiles of the requirement distribution for 14–18-year-old females. Table I-6 presents the resulting distribution.
Modeling the Distribution
Normal quantile plots indicated that the intake distribution for 14–18-year-old females is skewed to the right. Taking logs and making similar plots suggested that the distributions were fairly close to lognormal but were slightly less skewed. A cubic equation gave a very accurate description of the relationship between the normal score and log iron seen in the normal quantile plot for log iron. Therefore, the cumulative distribution for intake was determined by an equation of the form
where the constant, A, B, and C were estimated by least-squares. This cubic function is used to compute the cumulative distribution for the iron intake distribution needed for the probability approach for calculating the prevalence of inadequacy. The modeled percentiles agreed with reported percentiles exactly when rounded to the reported percentiles. The modeled intake distribution for 14–18-year-old females is
where Φ−1 is the inverse of the normal cumulative distribution function and i is the requirement. The situation was similar for the intake distribution of 11–13-year-old females. The modeled intake distribution is
The requirement distributions were somewhat more skewed. The method used for the intake distributions gave similarly accurate fits. For 14–18-year-old females, the modeled requirement distribution is
For 11–13-year-old females, the modeled requirement distribution is
These approximations for the intake and requirement distributions are very accurate when applied to values within the range of the reported percentiles. In the calculations used for the Target Median Intakes, the modeled values given above are used for intakes between the 0.5 percentile and the 99.5 percentile (tabled values for intakes are given for the 1.0 percentile and the 99 percentile) and for requirements between the 1.25 percentile and the 98.75 percentile (tabled values for requirements are given for the 2.5 percentile and the 97.5 percentile). For values outside these ranges, the cumulative distributions are set to zero for low values and one for high values.
Using the Probability Approach and the Modeled Distributions to Find Target Median Intakes
The probability approach was used with the modeled distributions to determine the prevalence of iron inadequacy for 14–18-year-old females and 11–13-year-old females. Alternative intake distributions were assumed to be of the same distributional form but shifted to higher or lower values. Computationally, this was accomplished by adding a constant to the value of i in FI(i). The prevalence of inadequacy was computed for a range of values of the constant and the value corresponding to a 5 percent prevalence of inadequacy was determined. The value represents the shift in the intake distribution needed to achieve a 5 percent prevalence of inadequacy. The Target Median Intake is the median of the shifted distribution.
For 14–18-year-old females, the Target Median Intake is 15.92 mg/d; and 11–13-year-old females the Target Median Intake is 15.53 mg/d.