THE JOURNAL OF PEDIATRICS • www.jpeds.

com CLINICAL AND LABORATORY

OBSERVATIONS
Mercury Poisoning in a Toddler from Home Contamination due to Skin-
Lightening Cream
Michael R. Ori, MD, Jaiva B. Larsen, MD, and Farshad “Mazda” Shirazi, MD, PhD

A 17-month-old child presented with hypertension, fussiness, constipation, and arthralgia due to mercury toxicity
from a skin-lightening cosmetic used by her mother and grandmother. Blood mercury level was 26 mcg/L and urine
level was 243 mcg/g creatinine. She was chelated with succimer. The home was contaminated and needed remediation.
(J Pediatr 2017;■■:■■-■■).

M
ercury exposure is a challenging and uncommon
diagnosis, with signs and symptoms that in more
severe cases overlap with neuroendocrine tumors.
This report details our experience with a case of mercury ex-
posure from household contamination from a cosmetic skin-
lightening facial cream. Around the globe, mercury-containing
creams are marketed for skin lightening, removal of dark spots,
and evening of skin tone. The desired skin effects are due to
inorganic mercury inhibiting tyrosinase activity and there-
fore reducing melanin production. Since 1973, US Federal law
has prohibited mercury in cosmetics beyond trace amounts
(1 mg/kg) due to concern for toxicity.1 Despite the law, mercury-
containing skin creams are an ongoing public health problem
in the US.2 This case report provides our rationale for mercury
testing, therapy, our approach to household remediation and
evaluation, and provides patient progress over a 200-day period.
Case Presentation
A 17-month-old previously healthy female toddler with normal
development was seen by her pediatrician regarding 3 weeks
of fussiness, constipation, decreased appetite, and tempera-
ture to 37.7°C. The pediatrician obtained a chest radiograph,
which was negative, and a urinalysis, which showed no evi-
dence of urinary tract infection. Two days later, the patient was
taken to the emergency department with similar complaints.
She had rhinorrhea, congestion, fussiness, and fever of 38.3°C
but otherwise had no abnormalities. She was discharged with
a presumptive viral syndrome.
In the week after discharge from the emergency depart-
ment, her symptoms did not resolve and she developed a
limp with tenderness in the right knee. After 1 week, she pre-
sented again to the pediatrician. An abdominal radiograph
showed large stool burden but no other concerning finding,
such as radio-opaque foreign bodies. Radiograph of the knee
was unremarkable. Repeat urinalysis again showed no evi-
dence of urinary tract infection. She was noted to have a
0.5-kg weight loss and new hypertension (Figure 1; available
at www.jpeds.com) above 95th percentile but was afebrile.
NG Nasogastric
PO BID By mouth 3 times per day
She was directly admitted to the hospital (day 1 hereafter),
where an endocrine workup was begun. Endocrine studies were
obtained and were either within the normal range or mod-
estly elevated (Table I; available at www.jpeds.com). Nuclear
medicine single-photon emission computed tomography
imaging of the adrenal glands showed no tumors. Renal func-
tion was normal.
The patient became increasingly fussy and had poor appe-
tite with continued weight loss, reaching a nadir of 11%
(1.14 kg) from admission weight on day 27. She had progres-
sive decrease in ambulation such that she required assis-
tance. After day 1 of hospitalization, rhinorrhea, congestion,
and fever had resolved. It is unclear whether she initially had
a viral syndrome or whether all presenting symptoms were
related to her final diagnosis. She had persistent unexplained
hypertension, weight loss, and inability to walk. Endocrine
workup had not established a diagnosis. Therefore, heavy
metal screening was obtained on day 18. Whole blood mercury
was found to be 26 mcg/L (normal <10 mcg/L) with a random
spot urine mercury level of 243 mcg/g creatinine (normal
<35 mcg/g creatinine) (Table II). Arsenic and lead levels were
unremarkable.
After the discovery of elevated mercury level on day 18, the
patient was further evaluated for clinical signs and symp-
toms of mercury toxicity and exposure history. She was noted
to have mild resting tremor. There was no rash, leukonychia
striata (Mees lines), desquamation of the hands or feet, or gin-
gival discoloration (Burton lines).
Multiple conversations with the patient’s mother were needed
to identify the source of mercury toxicity. Several other mercury
sources were considered but denied by the family. After re-
peated prompting, the patient’s mother recalled that she had
been using a skin-lightening facial cream at bedtime for 4
months, which she stored in the family refrigerator.
The patient is a first-generation Hispanic American living
in a Southwestern US border city. The other members of her
household included the patient’s 29-year-old mother, with
From the Arizona Poison and Drug Information Center, University of Arizona, Tucson,
AZ
The authors declare no conflicts of interest.
0022-3476/$ - see front matter. © 2017 Elsevier Inc. All rights reserved.
https://doi.org10.1016/j.jpeds.2017.12.023

CRP 5.5.0 DTD ■ YMPD9656_proof ■ January 30, 2018

THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume ■■ • ■■ 2017

Table II. Patient and family urine and blood mercury levels by day since admission
Day
Patients and levels Normal 14 18 23 35 40 46 59 61 94 158 180 202 222
Patient's urine, mcg/g cr ≤35 mcg/g cr — 243 777 142 770 81 153 — 109 90 58 45 —
Patient's urine, mcg/L ≤10 mcg/L — 58 101 41 77 13 61 — 12 35 14 13 —
Patient's blood, mcg/L ≤10 mcg/L 26 18 — — 8 — — — — — — — —
Mother's urine, mcg/g cr ≤35 mcg/g cr — — — — — — — 197 — — — — 33
Grandmother's urine, mcg/g cr ≤35 mcg/g cr — — — — — — — 222 — — — — 43

whom she co-sleeps, 53-year-old maternal grandmother, and
uncle. (Note that the uncle declined evaluation and is ex-
cluded from discussion.) There were no other children in the
household. The patient’s maternal grandmother used the cream
for at least 5 months, and several friends outside the house-
hold also used the cream. There were no cross-border trips
within the last several months.
The cream was produced and purchased at a beauty salon
in Mexico and was carried across the border by family and
friends. They all found the cream to be effective in meeting
their cosmetic goals. Several containers of the cream were pro-
vided by the family and were sent to the Arizona State labo-
ratory, where they were found to have between 27 000 and
34 000 mg/kg mercury. Her mother and grandmother did not
complain of symptoms and had no findings on examination
but did have markedly elevated first-void urine mercury levels
of 197 mcg/g creatinine for the mother and 222 mcg/g cre-
atinine for the grandmother. Their serum creatinine levels were
within normal limits.
Due to continued weight loss, the patient had a nasogastric
(NG) tube placed to facilitate feeding on day 24. Chelation was
initiated on day 21 using succimer 10 mg/kg by mouth 3 times
per day (PO TID) for 5 days and 10 mg/kg by mouth 2 times
per day (PO BID) for 14 days. At hospital discharge on day
40, the patient had begun walking and was eating, although
NG tube supplementation was continued. Four additional out-
patient courses of chelation have been performed using
succimer 10 mg/kg PO BID for 14 days. Assessment by a bi-
lingual developmental pediatrician on day 61 noted signifi-
cant delay in receptive language and fine motor skills on a
Bayley Scale. On that evaluation, the patient’s mother had re-
ported a 3- to 4-day period of decreased ambulation about 1
week before evaluation. On her most recent evaluation, on day
222, she was noted to be shy, particularly with other chil-
dren, and to have stereotypical hand-flapping behavior when
stressed. Bayley Scale evaluation has not yet been repeated.
Household Evaluation and Remediation
The family rented a detached single-family residence, which
we felt had a high likelihood of contamination. Federal, state,
and local agencies were contacted to coordinate evaluation of
the home. This was hampered by a lack of resources at the local
and state level. Contacts through the Pediatric Environmen-
tal Health Specialty Units and Environmental Protection Agency
2 Ori, Larsen, and Shirazi
region 9 resulted in dispatch of a commercial survey team from
a neighboring state.
On the initial survey, ambient air mercury vapor levels ranged
from 1900 to 2800 ng/m3 for most areas. Air samples from the
clothes washer were 14 150 ng/m3, with the dryer at 1767 ng/m3.
This was felt to represent trapping of mercury within the
plumbing of the washer and thermal volatility for the dryer.
Wipe samples from the refrigerator and kitchen table were
880 ng/m3 and 25 090 ng/m3, respectively.
The US Agency for Toxic Substance and Disease Registry
recommends immediate remediation for mercury vapor
>1000 ng/m3 and recommends occupancy exclusion for levels
about 10 000 ng/m3.3 The ambient samples in the household
in all cases reached the immediate remediation level. Given that
the patient had clinically significant affects attributable to
mercury exposure, the home was deemed not safe for con-
tinuous occupancy.
Remediation was performed by a private contractor with ex-
pertise in mercury cleanup and was paid for by the landlord.
Afterward, although mercury was still not at background, the
house was deemed safe for occupancy. The family lost per-
sonal and household items, including bedding materials and
the washing machine, which were too contaminated for
remediation. The remediation contractor disposed of these
items in a hazardous waste landfill.
Mexican officials were contacted through local border health
liaisons. The ensuing investigation resulted in seizure of ma-
terial in Mexico. The patient’s mother was reluctant to give
health officials contact information for her friends, but she even-
tually convinced them to discard the cream. The number of
contaminated sites in the community is unknown.
Discussion
Mercury is a toxic heavy metal with elemental, inorganic, and
organic forms. Each form has stereotypical toxicities but all
can cause neurologic dysfunction with sufficient exposure. It
is unique among heavy metals in that it has a relatively high
vapor pressure and significant vapor contamination may occur
from both metallic mercury and mercury-containing salts. In
this case, the patient was exposed to inorganic mercury (prob-
ably mercuric chloride) added to a skin-lightening cream. Ex-
posure was from contact with contaminated people, objects,
and vapor as the cream was not directly applied to her skin.
Of these, vapor may have contributed the most, and her rela-

CRP 5.5.0 DTD ■ YMPD9656_proof ■ January 30, 2018

■■ 2017
tive exposure would be expected to be greater than those of
the adults due to age-related increase in minute ventilation.
The role of topical exposure is more difficult to determine.4
Although there was no direct application of cream to her skin,
there would have been some skin-to-skin contact and there
would have been incidental dermal contact through contami-
nated bed linens. She likely had incidental oral contact from
contaminated objects due to toddler propensity to place objects
in their mouths.
She was not exposed to significant oral boluses of inor-
ganic mercury and so did not exhibit caustic gastrointestinal
upset typical of such ingestions. Inorganic mercury also is ex-
pected to cause renal dysfunction, which she did not exhibit.
Over time, central nervous system penetration occurs with re-
sultant neurologic dysfunction, which was the dominant feature
of this patient’s presentation, with debilitating leg pain, an-
orexia, constipation, neurasthenia, and autonomic dysfunc-
tion in the form of hypertension. Acrodynia, meaning extremity
pain, is an idiosyncratic reaction to mercury exposure seen in
childhood. Symptoms include irritability, weakness, paresthesias,
a pink papular rash, and desquamation of the palms and soles.5
Our patient did not have the characteristic dermal findings but
did have other symptoms consistent with acrodynia.
This case highlights the fact that significant individual varia-
tion in symptoms results with similar mercury exposure. The
patient was the only child in the household and suffered
significant neurologic symptoms, whereas the mother and
grandmother were asymptomatic, even though they too had
markedly elevated mercury levels. Therefore, we cannot rely
on identifying a cluster of symptomatic individuals to signal
a contaminated environment.
There have been many previous well-documented cases of
toxicity from mercury-containing cosmetics manufactured in
many parts of the word.6-8 Despite being prohibited, skin-
lightening creams from Mexico have been noted to cause
mercury toxicity in US border states. In 1995, the Texas De-
partment of Health (now the Texas Department of State Health
Service) traced a case of mercury toxicity in a 17-year-old boy
to a specific facial cream called “Crema de Belleza-Manning.”
A public health campaign was initiated to warn users of this
cream in Texas, California, Arizona, and New Mexico. In ad-
dition, users of the cream were invited to participate in a study
and undergo urine mercury testing. A total of 330 cream users
contacted the health departments of these 4 states and were
found to have a mean urine mercury concentration of
147.6 mcg/L.2
Previous reports within the US also demonstrate signifi-
cant household contamination from mercury-containing
cosmetics.9 Copan et al describe 3 households in California that
required remediation due to contamination from facial cream.
The most similar in age and circumstances to the case pre-
sented here was a 20-month-old child whose mother used a
skin-lightening cream.9 The child was hospitalized with symp-
toms similar to those in the case presented here, including hy-
pertension, refusal to walk, and poor appetite requiring an NG
feeding tube. The child’s mercury urine level was 52 mcg/g cre-
atinine, which is elevated but lower than the initial level of
Mercury Poisoning in a Toddler from Home Contamination due to Skin-Lightening Cream
CRP 5.5.0 DTD ■ YMPD9656_proof ■ January 30, 2018
CLINICAL AND LABORATORY OBSERVATIONS
243 mcg/g creatinine in our patient. The cream was found to
contain 38 000 mg/kg mercury, which is similar to the case pre-
sented here. Although not reported separately for this pa-
tient’s home, high levels of home contamination were found
in this case series, with bedroom air ranging from 500 to
8000 ng/m3 of mercury. The treatment course and clinical
outcome for this child are not reported.
The previous literature demonstrates that skin-lightening
creams are manufactured in multiple regions around the globe
and are distributed even in areas in which they are prohib-
ited by law. Our case highlights that toxicity can occur due to
home contamination even when there is no direct applica-
tion of these creams. Clinicians should consider mercury tox-
icity when patients present with unexplained neurologic
symptoms and autonomic dysregulation and should also en-
courage evaluation of other household members and engage
public health officials when mercury is discovered.
When faced with a case of suspected mercury poisoning, the
clinician must choose from several tests. Mercury levels can
be obtained from whole blood, spot (aka random) urine, first
morning urine, or 24-hour urine. The choice of tests depends
on the clinical scenario. Whole blood is typically used for acute
ingestions, as mercury is rapidly absorbed into erythrocytes10
and then distributes into the tissues through a
multicompartment kinetic model.11 This effect is demon-
strated in our patient, as her blood levels decreased from a
mildly elevated 26 mcg/L to 8 mcg/L within 26 days. Urine levels
are used for ongoing monitoring of both acute and chronic
exposure. The preferred test is 24-hour urine collection but
this is difficult to administer in toddlers and therefore spot urine
was monitored in this case. Our patient’s initial random urine
level was elevated; it then climbed at the onset of her initial
3-week inpatient chelation course, as would be expected, and
then declined rapidly only to increase again at the end of che-
lation (Figure 2).
The most important treatment of metal toxicity is removal
from the source. This was accomplished by hospitalization,
during which time her household was decontaminated. Che-
lation therapy is an option for heavy metal exposure, but it is
not a panacea. Chelating agents are nonspecific and may
decrease other cations like Ca2+ and Mg2+. In this case, we felt
chelation was appropriate, as the patient had significant
symptoms. Succimer was chosen as the chelating agent because
she was tolerant of enteral intake after NG placement and
because it can be easily continued in the outpatient setting.
Succimer is known to cause nausea, vomiting, decreased
appetite, diarrhea, transaminitis, and neutropenia. A total of
5 courses of therapy were given because we felt she had
ongoing symptoms and she was tolerating the therapy well.
There has been a downward trend in her overall mercury
levels (Figure 2); however, this would be expected by removal
from the source alone. Thus, the true effect of chelation on
her improvement is uncertain. In contrast, we did not feel
chelation was warranted in either her mother or grand-
mother despite their elevated mercury levels. Their levels
have declined without intervention, and no symptoms have
developed.
3
THE JOURNAL OF PEDIATRICS • www.jpeds.com Volume ■■

Mercury level
800 40

700

600

Urine Hg levels 500

400

300

200

100

0
0 50 100 Day 150 200

Figure 2. Urine mercury levels in relation to chelation. Solid graph line represents random urine mercury levels (mcg/g creati-
nine). Dashed graph line represents random urine levels (mcg/L). Vertical bars reflect chelation periods days 21, 61, 93, 180,
and 206. Dashed vertical line represents hospital discharge on day 40.

Although mercury toxicity is rare and has variable symp-
toms, it is valuable for clinicians to consider this diagnosis in
cases of unexplained hypertension and neurologic findings and
to consider the potential need for home evaluation and
remediation.
Consent
The patient’s mother graciously provided consent for publi-
cation of this report in the hopes that it will aid in diagnosis
and treatment for other exposed children. ■
Submitted for publication May 1, 2017; last revision received Nov 22, 2017;
accepted Dec 11, 2017
Reprint requests: Michael R. Ori, MD, Arizona Poison and Drug Information
Center, University of Arizona, 1295 N Martin, Tucson, AZ 85721. E-mail:
Mikeori1@gmail.com
References
1. Use of mercury compounds in cosmetics including use as skinbleaching
agents in cosmetic preparations also regarded as drugs. 21 USC § 700.13.
1973.
2. Weldon MM, Smolinski MS, Maroufi A, Hasty BW, Gilliss DL, Boulanger
LL, et al. Mercury poisoning associated with a Mexican beauty cream. West
J Med 2000;173:15.
3. Division of Toxicology and Environmental Medicine Prevention, Re-
sponse and Medical Support Branch Emergency Response Team. Action
levels for elemental mercury spills. Atlanta (GA): Agency for Toxic
Substances Disease Registry; 2012.
4. Clarkson TW, Magos L, Myers GJ. The toxicology of mercury—current
exposures and clinical manifestations. N Engl J Med 2003;2003:1731-7.
5. Tunnessen WW, McMahon KJ, Baser M. Acrodynia: exposure to mercury
from fluorescent light bulbs. Pediatrics 1987;79:786-9.
6. Chan TY. Inorganic mercury poisoning associated with skin-lightening
cosmetic products. Clin Toxicol 2011;49:886-91.
7. Al-Saleh I, Al-Doush I. Mercury content in skin-lightening creams and
potential hazards to the health of Saudi women. J Toxicol Environ Health
1997;51:123-30.
8. Murphy T, Kim S, Chanra P, Lim S, Wilson K, Irvine KN, et al. Mercury
contamination of skin-whitening creams in Phnom Penh, Cambodia. J
Health Pollut 2015;5:33-46.
9. Copan L, Fowles J, Barreau T, McGee N. Mercury toxicity and contami-
nation of households from the use of skin creams adulterated with
mercurous chloride (calomel). Int J Environ Res Public Health
2015;12:10943-54.
10. Barregård L, Sällsten G, Schütz A, Attewell R, Järvholm B. Kinetics of
mercury in blood and urine after brief occupational exposure. Arch Environ
Health 1992;47:176-84.
11. Jonsson F, Sandborgh-Englund G, Johanson G. A compartmental model
for the kinetics of mercury vapor in humans. Toxicol Appl Pharmacol
1999;155:161-8.

4 Ori, Larsen, and Shirazi

CRP 5.5.0 DTD ■ YMPD9656_proof ■ January 30, 2018

■■ 2017 CLINICAL AND LABORATORY OBSERVATIONS

150 Systolic blood pressure

140
130
120

mm Hg
110
100
90
80
70
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Day

Figure 1. Systolic blood pressure. The solid line reflects the patient’s blood pressure as recorded in the chart. Blood pressure
was not recorded on days 22 and 38 due to patient agitation. The dashed line represents 95th percentile for age.

Table I. Patient laboratory values from initial admission

Tests Reference Value
17-Hydroxyprogesterone 4-115 ng/dL 305
Cortisol 1.7-10.8 mcg/dL 21.1
DHEA sulfate ≤19 mcg/dL 14
Epinephrine 36-640 pg/mL 644
Norepinephrine 68-1810 pg/mL 2137
Dopamine 0-20 pg/mL 137
Androstenedione ≤0.149 ng/mL 0.189
Testosterone, total None ng/dL <12
Thyroid-stimulating hormone 0.70-4.17 µIU/mL 1.68
Free T4 0.9-1.4 ng/dL 1.4
T3, free 2.0-4.8 pg/mL 4.7
Plasma renin activity 0.25-5.82 ng/mL/h 5.77
Total free metanephrine ≤205 pg/mL 121

DHEA, dehydroepiandrosterone.

Mercury Poisoning in a Toddler from Home Contamination due to Skin-Lightening Cream 4.e1

CRP 5.5.0 DTD ■ YMPD9656_proof ■ January 30, 2018