Prevalence of Iron Deficiency and Iron-Deficiency Anemia in US Females Aged 12-21 Years, 2003-2020

Abstract

This study examines prevalence of iron deficiency among females aged 12 to 21 years to inform future screening strategies for iron deficiency and iron-deficiency anemia.

Iron deficiency and iron-deficiency anemia are common, underappreciated conditions with significant morbidity and mortality despite widespread availability of effective treatment. Historically, the focus of screening has been preschool-aged and pregnant persons. The Centers for Disease Control and Prevention recommends anemia screening for nonpregnant female adolescents and women every 5 to 10 years,¹ whereas the US Preventive Services Task Force does not address screening for these populations.¹

Although screening for anemia by measurement of hemoglobin level is recommended, there is benefit in identifying and treating iron deficiency in those without anemia because supplementation improves exercise performance and reduces fatigue, and iron deficiency is associated with increased all-cause mortality.^2,3 We examined iron deficiency prevalence among females aged 12 to 21 years to inform future screening strategies.

Methods

This study used National Health and Nutrition Examination Survey (NHANES) cycles from 2003-2010 and 2015–March 2020 (ferritin level was not measured in 2011-2014). NHANES is a series of nationally representative surveys consisting of interviews and physical examinations. Response rates ranged from 51% to 80%. The study protocol was approved by the ethics review board of the National Center for Health Statistics and participants provided informed consent.

Data were extracted for nonpregnant females aged 12 to 21 years. Individuals were excluded for missing data, inflammation, and kidney or liver dysfunction (additional information appears in the eMethods in Supplement 1). The proportion of the population with iron deficiency (ferritin <25 μg/L)⁴ was described; and the ferritin cutoffs of 15 μg/L and 50 μg/L were assessed as sensitivity analyses. The prevalence of iron-deficiency anemia (hemoglobin <12 mg/dL by World Health Organization definition and ferritin <25 μg/L) was examined as well as using the hemoglobin cutoffs of 12.5 mg/dL and 13 mg/dL, given debate around this definition.⁵ Quasibinomial models were used to generate independent adjusted odds ratios to assess the associations among race and ethnicity, income, food security, menstruation, and body mass index and having iron deficiency or iron-deficiency anemia. Self-reported race and ethnicity (using categories defined by NHANES) were collected to evaluate for associations between social determinants of health and iron deficiency.

The models restricted to menstruating individuals were generated to evaluate the association with years menstruating. The counts were unweighted and the percentages were weighted to account for nonresponse. A 2-sided α < .05 was considered statistically significant. Analyses were conducted using the survey package in R version 4.2.2 (R Foundation for Statistical Computing).

Results

There were 4052 individuals who met inclusion criteria and 3490 who had complete data. Of these 3490 individuals, 188 were premenarchal (5.4% [95% CI, 4.2%-6.6%]) (Table 1). The overall prevalence of iron deficiency was 38.6% (95% CI, 35.8%-40.9%); 17% (95% CI, 15.4%-19.2%) using a 15-μg/L ferritin cutoff and 77.5% (95% CI, 75.7%-79.3%) using a 50-μg/L cutoff. Premenarchal individuals had a prevalence of iron deficiency of 27.1% (95% CI, 17.1%-37.0%) using a 25-μg/L ferritin cutoff.

Table 1. Sample Description Stratified by Iron Deficiency and Iron-Deficiency Anemia Status.

Characteristics	Total (N = 3490)		Iron deficiency (n = 1413)		Iron deficiency anemia (n = 271)
	No.	% (95% CI)a	No.	% (95% CI)a	No.	% (95% CI)a
Age, median (IQR), y		16 (14-18)		16 (14-18)		17 (15-18)
Race and ethnicityb
Non-Hispanic Black	968	14.1 (11.9-16.3)	399	34.2 (27.1-41.2)	143	34.3 (27.2-41.4)
Hispanic	1224	19.0 (16.0-22.1)	544	21.6 (14.1-29.1)	69	21.7 (14.2-29.3)
Non-Hispanic White	1039	59.4 (55.7-63.1)	364	38.0 (28.9-47.0)	42	37.6 (28.5-46.7)
Otherc	259	7.4 (6.1-8.8)	106	6.3 (2.9-9.7)	17	6.3 (2.9-9.8)
Income-to-poverty ratio, median (IQR)		2.3 (1.1-4.1)		2.3 (1.0-3.9)		1.8 (1.0-3.4)
Below 130% of federal poverty level	1519	32.0 (28.9-35.0)	655	35.0 (30.7-39.3)	125	37.4 (29.8-45.0)
Food insecurity	1355	29.0 (26.5-31.6)	547	29.5 (25.6-33.3)	128	42.8 (35.8-49.8)
Premenarchal	188	5.4 (4.2-6.6)	41	3.8 (2.1-5.6)	3	0.4 (<0.01-0.9)
Body mass index, median (IQR)d		22.3 (19.8-26.2)		22.3 (19.9-25.8)		23.1 (20.6-27.0)

^a Unless otherwise indicated. The percentages were weighted to account for nonresponse.

^b Self-reported and defined by the National Health and Nutrition Examination Survey.

^c American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, non-Hispanic multiracial, or other or not listed.

^d Calculated as weight in kilograms divided by height in meters squared.

The overall prevalence of iron-deficiency anemia was 6.3% (95% CI, 5.2%-7.4%); 11.0% (95% CI, 9.5%-12.6%) using a 12.5-mg/dL hemoglobin cutoff and 17.2% (95% CI, 15.3%-19.1%) using a 13-mg/dL cutoff. Among individuals with iron deficiency, it was not associated with iron-deficiency anemia for 83.6% (95% CI, 80.8%-86.4%).

In multivariable analyses, non-White race, Hispanic ethnicity, and menstruation were associated with iron deficiency and iron-deficiency anemia. Lower body mass index and poverty were associated with iron deficiency. Food insecurity was associated with iron-deficiency anemia (Table 2). When restricting models to menstruating individuals, the number of years menstruating was not associated with iron deficiency (adjusted odds ratio, 1.05 [95% CI, 0.98-1.13]) or iron-deficiency anemia (adjusted odds ratio, 1.05 [95% CI, 0.92-1.21]).

Table 2. Adjusted Associations Between Demographic and Menstruation Characteristics and Odds of Iron Deficiency or Iron-Deficiency Anemia.

Characteristics	Iron deficiency		Iron deficiency anemia
	Adjusted OR (95% CI)	P value	Adjusted OR (95% CI)	P value
Age	0.99 (0.95-1.02)	.57	1.04 (0.98-1.11)	.20
Race and ethnicitya
Non-Hispanic Black	1.37 (1.06-1.68)	.02	4.05 (2.59-6.33)	<.001
Hispanic	1.34 (1.07-1.68)	.01	1.62 (0.97-2.72)	.07
Non-Hispanic White	1 [Reference]		1 [Reference]
Otherb	1.20 (0.81-1.76)	.35	1.35 (0.67-2.72)	.40
Below 130% of federal poverty level
No	1 [Reference]		1 [Reference]
Yes	1.24 (1.02-1.51)	.03	0.86 (0.60-1.21)	.37
Food insecurity
No	1 [Reference]		1 [Reference]
Yes	0.87 (0.70-1.09)	.22	1.50 (1.03-2.18)	.04
Menstruation status
Premenarchal	1 [Reference]		1 [Reference]
Menstruating	1.90 (1.12-3.12)	.02	11.21 (3.33-37.7)	<.001
Body mass index	0.98 (0.96-0.99)	.004	1.00 (0.97-1.03)	.97

Abbreviation: OR, odds ratio.

^a Self-reported and defined by the National Health and Nutrition Examination Survey.

^b American Indian or Alaska Native, Asian, Native Hawaiian or Pacific Islander, non-Hispanic multiracial, or other or not listed.

Discussion

Among 12- to 21-year-old US females between 2003 and 2020, iron deficiency affected almost 40% and iron-deficiency anemia affected 6%, with variation by the ferritin or hemoglobin thresholds used. Menstruation was a risk factor for both, but more than one-quarter of premenarchal individuals had iron deficiency.

Limitations of this study include limited granularity of the race and ethnicity data and potential overfitting of the iron-deficiency anemia model because few premenarchal participants had iron-deficiency anemia. However, removing the menstruation variable from the model had minimal effects on other adjusted associations.

Given the high prevalence of iron deficiency found with the majority not associated with iron-deficiency anemia, current screening guidance may miss many individuals with iron deficiency. Although annual screening is recommended for higher-risk patients, risk factors (extensive menstrual blood loss,⁶ low iron intake, prior diagnosis of iron deficiency) are not clearly defined and likely result in inconsistent screening.

The frequency of universal screening for iron deficiency and iron-deficiency anemia in menstruating persons and the best ferritin and hemoglobin thresholds should be evaluated. Further study is needed to identify risk factors and inform screening practices among premenarchal individuals.

Section Editors: Jody W. Zylke, MD, Deputy Editor; Kristin Walter, MD, Senior Editor.

Supplement 1.

eMethods

Supplement 2.

Data Sharing Statement