More Related Content Similar to 38 Hypertension Similar to 38 Hypertension (20) 38 Hypertension1. 3. Cerebrovascular Accident
Hypertension 1. Relative Risk increases 1.84 for each
10 mmHg DBP
= Htn, Essential Hypertension
2. Midlife hypertension raises
1. See Also longterm CVA risk
1. Hypertension in Children
2. Hypertension in Infants 3. Seshadri (2001) Arch Intern Med
3. Hypertension in Pregnancy 161:2343
4. Hypertension in Athletes
5. Hypertension in the Elderly
4. Alzheimer's Disease
1. Increased SBP in middle age is
predisposing factor
2. Epidemiology
1. Demographics
1. White Adults in US: 20% are hypertensive
2. Black Adults in US: 30% are hypertensive
3. Of all hypertensives, >50% are over age 5. Reference
1. (1995) Lancet 346(8991):1647
65 years 2. Kivipelto (2001) BMJ 322:1447
2. Outcomes
1. BEST PREDICTOR TO OUTCOME
VARIES BY AGE 3. Hypertension Definition
1. Diastolic Blood Pressure best
1. See Hypertension Criteria
predictor <50 years
4. Types
1. Essential Hypertension (Primary
2. SBP and DBP predict outcomes
Hypertension)
equally ages 50-59
1. Stage 1-2 (DBP 90-104) in 80%
of cases
3. Pulse Pressure best predictor age
2. ACCELERATED MALIGNANT
>60 years HYPERTENSION
1. Recent substantial Blood Pressure
increase
2. Associated with retinal vessel damage
2. Coronary Artery Disease
1. Hypertension Causes 35-45% 1. Retinal Hemorrhages
morbidity and mortality 2. Retinal exudates
3. Papilledema
3. Diastolic Blood Pressure over 140
2. Hypertension Resources
3. ISOLATED SYSTOLIC HYPERTENSION
1. Dash Diet and other lifestyle change handouts
1. Systolic Blood Pressure: >160 mmHg 1. http://www.nhlbi.nih.gov/health/public/h
eart/hbp/dash
2. Diastolic Blood Pressure: <90 mmHg 2. NIH Patient Handout Order Forms (free
for single copy)
3. Risks 1. CvHtnNihPubOrders.pdf
1. Coronary Artery Disease
2. Cerebrovascular Accident
4. Onset: 5th decade
5. Affects 11% of those over age 75 years
6. Results from progressive fall in vessel
compliance
5. Causes of secondary
Hypertension
1. See Hypertension Causes
6. Diagnosis
1. See Hypertension Criteria
7. Hypertension Evaluation
1. Hypertension Evaluation History
2. Hypertension Evaluation Exam
3. Hypertension Evaluation Labs
8. Management
1. See Hypertension Management
3. Isolated Systolic
Hypertension = ISH 5. Management
1. USE LOWER ANTIHYPERTENSIVE
DOSAGES
1. See Also
1. Hypertension
2. FIRST CHOICE MEDICATION: Diuretic
1. Even better benefit in Diabetes Mellitus with ISH
2. Epidemiology
1. Most common type of Hypertension in adults
2. Study of 4736 type II diabetics
3. Criteria 1. Lower Incidence of cardiac events
1. Systolic Blood Pressure: >140 mmHg (previously >160)
1. Lower Incidence of
2. Diastolic Blood Pressure: <90 mmHg
Cerebrovascular
Accidents
2. Lower Incidence of
4. Complications: Cardiovascular Risk
1. GENERAL Myocardial Infarction
1. Systolic pressure predicts risk better than
2. Reference
diastolic
1. Curb (1996) JAMA 276:1886
2. Wide Pulse Pressure best predicts cardiovascular
risk
3. OTHER MEDICATIONS
1. Dihydropyridine Calcium Channel Blocker
2. Long Acting Nitroglycerin
3. ACE Inhibitor
2. ADVERSE EFFECTS
4. Labetalol
1. Doubles all cause mortality
2. Triples cardiovascular mortality
3. Increases cardiovascular morbidity 2.5 fold
4. Pulse pressure, widened: Excerpt
from Alarming Signs and
Symptoms: Lippincott Manual of
Nursing Practice Series
LABETALOL
Pulse pressure is the difference between systolic and diastolic
blood pressures. Normally, systolic pressure is about 40 mm
Hg higher than diastolic pressure. Widened pulse pressure — a Pharmacology
difference of more than 50 mm Hg — commonly occurs as a
Metabolism: liver extensively; CYP450: unknown
physiologic response to fever, hot weather, exercise, anxiety,
anemia, or pregnancy. However, it can also result from certain Excretion: urine 50% (<5% unchanged), feces 50%; Half-life:
neurologic disorders — especially life-threatening increased 5-8h
intracranial pressure (ICP) — or from cardiovascular disorders
that cause blood backflow into the heart with each contraction Subclass: Beta Blockers
such as aortic insufficiency. Widened pulse pressure can easily
be identified by monitoring arterial blood pressure and is Mechanism of Action
commonly detected during routine sphygmomanometric selectively antagonizes alpha1-adrenergic receptors;
recordings. antagonizes beta1- and beta2-adrenergic receptors (selective
alpha and non-selective beta blocker)
Act Now: Ifthe patient’s level of consciousness (LOC) is
decreased and you suspect that his widened pulse
pressure results from increased ICP, check his vital signs.
Maintain a patent airway, and prepare to hyperventilate
the patient with a handheld resuscitation bag to help
reduce partial pressure of carbon dioxide levels and,
thus, ICP. Perform a thorough neurologic examination
to serve as a baseline for assessing subsequent changes.
Use the Glasgow Coma Scale to evaluate the patient’s
LOC. (See Glasgow Coma Scale, page 196.) Also, check
cranial nerve (CN) function — especially in CNs III, IV,
and VI — and assess pupillary reactions, reflexes, and
muscle tone. (See Exit points for the cranial nerves.)
The patient may require an ICP monitor. If you don’t
suspect increased ICP, ask about associated symptoms,
such as chest pain, shortness of breath, weakness,
fatigue, or syncope. Check for edema and auscultate for
murmurs.
5. Hypertension Causes
2. Causes: Secondary Hypertension in age
= Secondary Hypertension Causes, Hypertension Causes in <18 years old
Children, Hypertension Causes in Adolescents 1. SEE HYPERTENSION IN INFANTS
1. Causes: Secondary Hypertension in
Adults
1. MEDICATIONS 2. Renal parenchymal disease
1. See Medication Causes of Hypertension 1. Most common cause in children (up to 70%)
2. PRIMARY ALDOSTERONISM 3. Renal vascular disease
1. Most common treatable secondary cause
of Hypertension
2. Evaluate as cause in Refractory 4. Aortic Coarctation
Hypertension where Hypokalemia or
borderline low potassium
5. Endocrine conditions
1. Metabolic Syndrome
3. RENOVASCULAR OR RENAL 2. Pheochromocytoma
PARENCHYMAL DISEASE
3. Hyperthyroidism
4. PHEOCHROMOCYTOMA
5. CUSHING'S DISEASE
6. Essential Hypertension
6. HYPERPARATHYROIDISM 1. Rare in age <10 years
7. AORTIC COARCTATION
2. Most common cause in adolescents and
8. SLEEP APNEA
adults
9. THYROID DISEASE 7. Medications
1. See Medication Causes of Hypertension
1. Hyperthyroidism causes systolic Hypertension
2. Hypothyroidism causes diastolic Hypertension
1. Dernellis (2002) Am Heart J 143:718
6. Hypertension Evaluation
1. Goals
1. Confirm Hypertension (see Hypertension Criteria)
2. Identify associated RISK FACTORS
3. Identify target organ disease
4. Evaluate for secondary Hypertension
2. Evaluation
1. Hypertension Evaluation History
2. Hypertension Evaluation Exam
3. Hypertension Evaluation Labs
3. Monitoring Protocol
1. MILD BLOOD PRESSURE INCREASE
1. Recheck in 1-2 months
2. MODERATE BLOOD PRESSURE
INCREASE
1. Recheck in 1-2 weeks
3. SEVERE OR ACCELERATED
MALIGNANT HYPERTENSION
1. Immediate Treatment
4. END ORGAN DAMAGE
1. Immediate Treatment
7. Hypertension Evaluation
History 6. History: Symptoms of Urinary tract
Disease
1. Urinary Tract Infection
1. History: Past Medical History
2. Nephrolithiasis (or Hypercalcemia)
1. Onset and severity of Hypertension
3. Benign Prostatic Hypertrophy
2. Average Blood Pressure
2. History: Family History
1. Hypertension
2. Kidney disease 7. Findings: Evidence of Endocrine
Disease
1. Diabetes Mellitus
2. Hyperthyroidism
3. History: Medications 3. Hypothyroidism
1. See Medication Causes of Hypertension
4. HYPERPARATHYROIDISM
4. History: Habits: (HYPERCALCEMIA)
1. Salt intake 1. Confusion
2. Fat intake 2. Major Depression
3. High caloric intake contributing to Obesity 3. Abdominal Pain
4. Alcohol use 4. Nephrolithiasis
5. Tobacco Use 5. Constipation
6. Recreational drugs of abuse
1. Cocaine
2. Methamphetamine
5. CUSHING'S DISEASE
1. Acne Vulgaris
2. Osteoporosis
3. Bone Fractures
5. History: Lead Exposure Risk
4. Glucose Intolerance
1. Lead paints
2. Printer inks
3. Inhalation risks
4. Postmenopausal women
6. ALDOSTERONISM
1. Lead increases due to skeletal lead 1. Hypokalemia
mobilization 2. Muscle Weakness
2. Nash (2003) JAMA 289:1523
3. Paresthesias
4. Tetany
8. 10. Findings: Evidence of Sleep
Disorder
7. 11-HYDROXYLASE DEFICIENCY
1. Premature virilization in males 1. Sleep Apnea
2. Masculinization in females 2. Exaggerated snoring
3. Inappropriate sleep episodes
8. 17-HYDROXYLASE DEFICIENCY _____________
1. Failed sex maturation
Amaurosis
http://www.bing.com/reference/semhtml/Amaurosis?si
9. PHEOCHROMOCYTOMA ds=2&q=Amaurosis+fugax&qpvt=Amaurosis+fugax
1. Sweating
2. Tremor
3. Panic
http://www.bing.com/reference/semhtml/Amaurosis_fu
4. Facial pallor gax?sids=2&q=Amaurosis+fugax&qpvt=Amaurosis+fu
5. Headache gax
Weight loss
6.
8. Findings: Evidence of Neurologic
Disease
1. Previous Neurologic disease or symptoms
2. Headaches
3. Confusion
4. Seizures
9. Findings: Evidence of Cardiovascular
disease
1. Coronary Artery Disease
2. Congestive Heart Failure
3. Amaurosis Fugax
4. Claudication
5. Renal Artery Stenosis
9. Blood Lead, Blood Pressure, and Hypertension in
Perimenopausal and Postmenopausal Women
Denis Nash; Laurence Magder; Mark Lustberg; et al.
Online article and related content
current as of January 3, 2010. JAMA. 2003;289(12):1523-1532 (doi:10.1001/jama.289.12.1523)
http://jama.ama-assn.org/cgi/content/full/289/12/1523
Correction Contact me if this article is corrected.
Citations This article has been cited 60 times.
Contact me when this article is cited.
Topic collections Occupational and Environmental Medicine; Women's Health; Women's Health,
Other; Hypertension
Contact me when new articles are published in these topic areas.
Related Letters Blood Lead Levels and Hypertension
Hans W. Hense. JAMA. 2003;290(4):460.
Robert P. Heaney. JAMA. 2003;290(4):460.
Subscribe Email Alerts
http://jama.com/subscribe http://jamaarchives.com/alerts
Permissions Reprints/E-prints
permissions@ama-assn.org reprints@ama-assn.org
http://pubs.ama-assn.org/misc/permissions.dtl
Downloaded from www.jama.com by guest on January 3, 2010
10. ORIGINAL CONTRIBUTION
Blood Lead, Blood Pressure, and
Hypertension in Perimenopausal
and Postmenopausal Women
Denis Nash, PhD, MPH Context Lead exposures have been shown to be associated with increased blood
Laurence Magder, PhD, MPH pressure and risk of hypertension in older men. In perimenopausal women, skeletal
lead stores are an important source of endogenous lead exposure due to increased
Mark Lustberg, PhD
bone demineralization.
Roger W. Sherwin, MD Objective To examine the relationship of blood lead level with blood pressure and
Robert J. Rubin, PhD hypertension prevalence in a population-based sample of perimenopausal and post-
Rachel B. Kaufmann, PhD menopausal women in the United States.
Design, Setting, and Participants Cross-sectional sample of 2165 women aged 40
Ellen K. Silbergeld, PhD
to 59 years, who participated in a household interview and physical examination, from
the Third National Health and Nutrition Examination Survey conducted from 1988 to 1994.
S
INCE THE 1970S, CONSIDERABLE Main Outcome Measures Associations of blood lead with blood pressure and hy-
attention has been paid to the pertension, with age, race and ethnicity, cigarette smoking status, body mass index,
possibility that low levels of lead alcohol use, and kidney function as covariates.
exposure among adults in the Results A change in blood lead levels from the lowest (quartile 1: range, 0.5-1.6
general population can elevate blood µg/dL) to the highest (quartile 4: range, 4.0-31.1 µg/dL) was associated with small
pressure and increase the risk for hy- statistically significant adjusted changes in systolic and diastolic blood pressures. Women
pertension, a leading risk factor for car- in quartile 4 had increased risks of diastolic ( 90 mm Hg) hypertension (adjusted odds
diovascular disease morbidity and mor- ratio [OR], 3.4; 95% confidence interval [CI], 1.3-8.7), as well as moderately in-
tality.1-3 Evidence for this association creased risks for general hypertension (adjusted OR, 1.4; 95% CI, 0.92-2.0) and sys-
from the epidemiological literature is tolic ( 140 mm Hg) hypertension (adjusted OR, 1.5; 95% CI, 0.72-3.2). This asso-
ciation was strongest in postmenopausal women, in whom adjusted ORs for diastolic
compelling,4 but the exact causal na- hypertension increased with increasing quartile of blood lead level compared with quar-
ture of the relationship remains con- tile 1 (adjusted OR, 4.6; 95% CI, 1.1-19.2 for quartile 2; adjusted OR, 5.9; 95% CI,
troversial. 1.5-23.1 for quartile 3; adjusted OR, 8.1; 95% CI, 2.6-24.7 for quartile 4).
The notion that lead exposure may in-
Conclusions At levels well below the current US occupational exposure limit guide-
fluence blood pressure in humans is bio- lines (40 µg/dL), blood lead level is positively associated with both systolic and dias-
logically plausible. Lead induces hyper- tolic blood pressure and risks of both systolic and diastolic hypertension among women
tension in rats,5,6 and other animal data aged 40 to 59 years. The relationship between blood lead level and systolic and dias-
suggest that lead acts at multiple sites tolic hypertension is most pronounced in postmenopausal women. These results pro-
within the cardiovascular system, in- vide support for continued efforts to reduce lead levels in the general population, es-
cluding direct effects on the excitability pecially women.
and contractility of the heart, alteration JAMA. 2003;289:1523-1532 www.jama.com
of the compliance of the vascular smooth Author Affiliations: Department of Epidemiology and (Dr Rubin); and National Center for Environmental
muscle tissue, and direct action on parts Preventive Medicine, University of Maryland School Health, Centers for Disease Control and Prevention,
of Medicine, Baltimore (Drs Nash, Magder, Lust- Atlanta, Ga (Dr Kaufmann). Dr Silbergeld is now with
of the central nervous system respon- berg, Sherwin, and Silbergeld); New York City De- the Department of Environmental Health Sciences, The
sible for blood pressure regulation.2 Evi- partment of Health and Mental Hygiene, HIV/AIDS Johns Hopkins University Bloomberg School of Pub-
Surveillance and Epidemiology Program, New York (Dr lic Health, Baltimore, Md.
dence in animals also suggests that lead Nash); Department of Epidemiology, Tulane Univer- Corresponding Author and Reprints: Denis Nash, PhD,
may affect blood pressure through the re- sity School of Public Health and Tropical Medicine, New MPH, New York City Department of Health and Men-
Orleans, La (Dr Sherwin); Department of Environ- tal Hygiene, HIV/AIDS Surveillance and Epidemiol-
nin-angiotensin system.6 Lead is neph- mental Health Sciences, The Johns Hopkins Univer- ogy Program, 346 Broadway, Room 706, New York,
rotoxic to humans, and alteration of sity Bloomberg School of Public Health, Baltimore, Md NY 10013 (e-mail: dnash@health.nyc.gov).
©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, March 26, 2003—Vol 289, No. 12 1523
Downloaded from www.jama.com by guest on January 3, 2010
11. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
kidney function may precede the devel- miological data associating relatively low using these cutoff values, excluding per-
opment of hypertension.7,8 However, levels of lead in the blood with cardio- sons who reported being treated for hy-
whether lead affects blood pressure vascular outcomes1,2,4,19-38 and because pertension. More details on measure-
through altering kidney function in hu- hypertension is a significant health con- ment of and outcomes related to blood
mans is not known. cern for women after menopause.39 pressure and hypertension in NHANES
A case-control investigation of men III have been published elsewhere.3
from the Normative Aging Study9 re- METHODS Blood Lead. Blood samples were ob-
ported significantly higher levels of lead The study population included women tained by venipuncture during the
in skeletal and blood compartments from the Third National Health and Nu- physical examination. Blood lead con-
among men with hypertension com- trition Examination Survey (NHANES centration was measured by graphite
pared with normotensives. The all- III), a cross-sectional sample obtained furnace atomic absorption spectropho-
male study population had mean base- through a complex survey design, rep- tometry at the laboratories of the Na-
line blood lead levels of 6.3 µg/dL, resenting the US civilian, noninstitu- tional Center for Environmental Health
similar to men in the general popula- tionalized population. During a 6-year at the Centers for Disease Control and
tion.10 An increase from the midpoint period (1988-1994), participants took Prevention in Atlanta, Georgia. The as-
of the lowest quintile to the highest part in a household interview and an say detection limit was 1.0 µg/dL. Each
quintile of bone lead was associated in-depth physical examination with sample analysis was performed in du-
with an adjusted odds ratio (OR) of laboratory tests. Full details of the sur- plicate, and the mean of both measure-
1.5 (95% confidence interval [CI], vey design have been published by the ments was used in these analyses. All
1.1-1.8) for hypertension, suggesting National Center for Health Statistics of blood lead levels less than 1.0 µg/dL
that cumulative lead exposure, repre- the Centers for Disease Control and were assigned a value of 0.5 µg/dL to
sented by bone lead stores, may be an Prevention.40 be consistent with previous analyses of
independent risk factor for hyperten- Our investigation focused on the sub- NHANES III lead data by other inves-
sion in the general population.9 set of 2574 women aged 40 to 59 years tigators.10
Evidence suggests that bone lead who participated in the NHANES III Menopausal Status. Women were cat-
stores contribute to circulating levels survey interview. From this group, 409 egorized as premenopausal (ovarian
of lead in blood.11-13 In particular, blood women were excluded for the follow- function intact), surgically menopausal
lead levels in women appear to increase ing reasons: 211 did not undergo a (both ovaries removed surgically be-
during the menopausal transition, physical examination or blood test- fore cessation of menses), and naturally
because of the mobilization of skeletal ing; 77 did not have information about menopausal (nonsurgical cessation of
lead stores associated with bone demin- blood lead levels; and 121 women of ovarian function). Women without his-
eralization.14-18 The impact of these small ethnicity other than non-Hispanic tories of reproductive surgery were clas-
but significant increases in blood lead black, non-Hispanic white, and Mexi- sified as premenopausal if they re-
in postmenopausal compared with pre- can American were excluded because ported a menstrual period during the
menopausal women is difficult to inter- of small numbers in any single self- previous 12 months and postmeno-
pret, because relatively few studies have reported category. The remaining 2165 pausal if they did not. Women report-
examined the health impacts of lead in women constituted the sample used. ing having undergone hysterectomy
women. A case-control study7 of 297 (without ovariectomy) within a month
women with hypertension who partici- Definitions of the last menstrual period were as-
pated in the Nurses’ Health Study Blood Pressure and Hypertension. We signed a menopausal classification based
showed that increases in bone patella used the mean of 3 systolic and dias- on their age ( 51 years, premeno-
lead levels from the 10th to the 90th per- tolic blood pressure measurements, all pausal; 51 years, naturally meno-
centile were associated with increased of which were taken by a physician at pausal). Women who underwent bilat-
risks of hypertension (OR, 1.86, 95% the end of the 4-hour physical exami- eral ovariectomy within 1 month of the
CI, 1.09-3.19). However, information nation that occurred in the NHANES date of the last menstrual period were
about menopausal status was extremely mobile examination center. Women classified as surgically menopausal.
limited in this study.7 were categorized as hypertensive if any Women who underwent hysterectomy
The objective of our investigation was of the following criteria were met: cur- or ovariectomy more than 1 month af-
to examine the relationship of blood lead rent user of blood pressure medica- ter the reported date of the last men-
levels with blood pressure and hyper- tion (self-report), a systolic blood pres- strual period were classified as natu-
tension in a population-based sample of sure of 140 mm Hg or higher, or a rally menopausal. A total of 101 women
perimenopausal and postmenopausal diastolic blood pressure of 90 mm Hg could not be assigned a menopausal sta-
women in the United States. We se- or higher. We also examined separate tus due to missing information.
lected blood pressure and hyperten- dichotomous variables for systolic hy- Kidney Function. Serum creatinine
sion as outcomes because of the epide- pertension and diastolic hypertension was measured because it is the most
1524 JAMA, March 26, 2003—Vol 289, No. 12 (Reprinted) ©2003 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on January 3, 2010
12. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
specific of the 3 measures of kidney used multiple logistic regression to ex- tile, respectively (TABLE 1). Women in
function available in NHANES III (se- amine the risks of hypertension (gen- the higher quartiles of blood lead tended
rum creatinine, urinary creatinine clear- eral, systolic, and diastolic) by catego- to be older, current smokers, regular
ance, and blood urea nitrogen) and was rizing blood lead in terms of quartiles drinkers, poorer, less educated, and
consistent with other recent studies of and comparing those women in blood more likely to be non-Hispanic black
the effects of lead on the kidney.41 lead quartiles 2, 3, and 4 with those in than those in the lower quartiles. All of
Covariates. Information about race quartile 1; these analyses were strati- these variables were significantly asso-
and ethnicity (non-Hispanic black, fied by menopausal status. ciated with blood lead level.
non-Hispanic white, and Mexican Models were constructed based on Of the 2165 women in the sample,
American), age (years), cigarette smok- outcomes known to be biologically as- 604 were classified as hypertensive
ing history (current, former, or never), sociated with blood pressure (age, race based on their systolic and diastolic
family income, and education was ob- and ethnicity, BMI, and serum blood pressures (n=231, untreated), as
tained from the household interview. creatinine), including the study vari- well as whether they self-reported cur-
Information about body mass index able blood lead. Potential confound- rently taking antihypertensive medica-
(BMI, calculated as weight in kilo- ing variables (education, poverty in- tions (n=373). Of those that were un-
grams divided by the square of height come ratio, alcohol use, and cigarette treated (n = 231), 123 had systolic
in meters) and alcohol use (amount smoking status) were included if they hypertension only, 30 had diastolic hy-
consumed per week) was obtained from were found to be significantly associ- pertension only, and 78 had both sys-
the physical examination and exami- ated with blood pressure outcomes in tolic and diastolic hypertension. Of
nation-associated questionnaire, re- any 1 of the models before the inclu- those who were treated for hyperten-
spectively. A 4-level categorical vari- sion of blood lead. Final regression co- sion (n = 373), 202 had neither sys-
able for weekly alcohol intake was variates included age, race and ethnic- tolic nor diastolic hypertension, 102 had
created with the following levels: none, ity, alcohol use, cigarette smoking systolic hypertension only, 14 had di-
less than 1, 1 to 2, or 3 or more drinks status, BMI, and kidney function. astolic hypertension only, 50 had both
per week. The poverty income ratio, a Statistical analyses were conducted us- systolic and diastolic hypertension, and
ratio of family income to the poverty ing SAS version 6 (SAS Institute, Cary, 5 did not have a systolic or diastolic
level income for a given family size ad- NC), incorporating the examination blood pressure measurement during the
justed to the poverty threshold for the sampling weights of NHANES III.40 The examination.
year of the interview, was used to cre- statistical software package SUDAAN In these crude analyses, blood lead
ate a 3-level family income variable. A version 7.0 (Research Triange Insti- quartile was significantly associated
participant was assigned a family in- tute, Research Triangle Park, NC) was with systolic blood pressure (P = .03)
come higher than the poverty level if used to calculate SEs for the estimates, but not diastolic blood pressure
the poverty income ratio was more than accounting for both the weights and the (P =.86) (TABLE 2). A significant dose-
1, at or lower than poverty if the pov- complex survey design. Linear regres- response existed between blood lead
erty income ratio was less than or equal sion coefficients reported are unstand- quartile and general hypertension preva-
to 1, and missing if the survey partici- ardized. The significance of regression lence, with 19.4% of women having gen-
pant did not report a family income coefficients was evaluated using the eral hypertension in the lowest quar-
level. A 4-level education variable was Wald 2 test. Statistical tests for trends tile compared with 28.3% in the highest
created on the basis of the number of of categorical variables were carried out quartile. However, although dose-
years of education reported by the sur- in regression models by coding levels as response trends appeared to exist, blood
vey participant (0-11 years = high integers (scores) and evaluating tests for lead quartile was not significantly asso-
school; 12 years = completed high significance on the slope of the regres- ciated specifically with systolic or dias-
school; 12-15 years=some college; and sion line. Statistical tests with P .05 tolic hypertension prevalence (P=.09
16 years = completion of college or were considered statistically signifi- and P =.25, respectively).
higher). cant. All estimates of proportions, re-
gression coefficients, and ORs are Systolic and Diastolic
Statistical Methods weighted to the 1990 US Census popu- Blood Pressure
We used multiple linear regression mod- lation. In multivariate analyses, blood lead was
els to examine the associations of blood significantly associated with both sys-
lead and menopausal status with sys- RESULTS tolic and diastolic blood pressures
tolic and diastolic blood pressures. Overall, the mean blood lead level for (TABLE 3). In these regression mod-
Analyses that examined systolic and di- women aged 40 to 59 years was 2.9 µg/ els, a difference in blood lead levels be-
astolic blood pressure as continuous out- dL, and the means for the quartiles of tween the lowest and highest quartiles
come variables excluded the 368 women blood lead ranged from 1.0 µg/dL to 6.3 was associated with a difference of 1.7
with hypertension who were treated. We µg/dL in the lowest and highest quar- mm Hg in systolic blood pressure and
©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, March 26, 2003—Vol 289, No. 12 1525
Downloaded from www.jama.com by guest on January 3, 2010
13. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
1.4 mm Hg in diastolic blood pres- diastolic) did not alter the signifi- independently increasing the ORs of
sure, after adjustment for age, race and cance of the blood lead variable. general hypertension in women in-
ethnicity, cigarette smoking, BMI, al- cluded increasing age, being non-
cohol use, and kidney function. Omis- General Hypertension Hispanic black, having an alcohol in-
sion of the serum creatinine variable in Before incorporating blood lead level in take of less than 1 drink per week, and
the multivariable models for blood lead the multiple logistic regression model increasing BMI. Adding blood lead level
and blood pressure (both systolic and (TABLE 4), the most important factors to the model did not greatly alter any of
Table 1. Weighted Descriptive Characteristics of Adult Women Aged 40 to 59 Years Participating in the Third National Health and Nutrition
Examination Survey*
Blood Lead Quartile
Total Quartile 1 Quartile 2 Quartile 3 Quartile 4 P
Characteristic (N = 2165) (n = 568) (n = 498) (n = 556) (n = 543) Value
Blood lead level, mean (range), µg/dL 2.9 (0.50-31.1) 1.0 (0.5-1.6) 2.1 (1.7-2.5) 3.2 (2.6-3.9) 6.4 (4.0-31.1)
Race and ethnicity, %
Non-Hispanic white 83.9 87.4 86.5 83.4 76.3
Non-Hispanic black 11.7 8.1 9.3 12.6 18.5 .001
Mexican American 4.4 4.5 4.2 4.1 5.2
Age, mean (SE), y 48.2 (0.2) 46 (0.32) 48 (0.44) 49 (0.34) 50.4 (0.39) .001
Body mass index, mean (SE) 27.6 (0.25) 28.4 (0.58) 27.5 (0.31) 27.6 (0.34) 26.9 (0.29) .04
Cigarette smoking history, %
Current 25.0 8.1 20.2 35.2 42.8
Former 25.5 30.4 25.2 19.2 26.5 .001
Never 49.5 61.5 54.6 45.6 30.7
Alcohol use, %
None 56.8 62.6 60.1 54.0 47.8
1 per week 15.2 14.5 15.9 15.6 14.6
.001
1-2 per week 17.1 14.5 14.4 21.4 18.6
3 per week 11.0 8.5 9.5 9.0 19.0
Household income, %
At or below poverty 8.8 6.1 7.7 8.0 15.1
Above poverty 84.9 90.3 84.4 84.8 78.0 .001
Missing 6.3 3.6 8.0 7.2 6.9
Education, %
High school 18.9 13.0 16.7 22.6 25.6
Completed high school 40.0 40.0 43.1 38.2 38.5
.001
Some college 19.9 18.4 21.9 21.0 18.5
College or higher 21.1 28.6 18.3 18.2 17.5
2
*Body mass index is calculated as weight in kilograms divided by the square of height in meters. P values obtained from test (categorical variables) or analysis of variance
(continuous variables) based on an overall test across quartiles.
Table 2. Weighted Distributions of Blood Pressure–Related Variables Among Adult Women Aged 40 to 59 Years Participating in the Third
National Health and Nutrition Examination Survey
Blood Lead Quartile
Total Quartile 1 Quartile 2 Quartile 3 Quartile 4 P P for
Characteristic (N = 2165) (n = 568) (n = 498) (n = 556) (n = 543) Value* Trend
Blood lead level, mean (range), µg/dL 2.9 (0.50-31.1) 1.0 (0.5-1.6) 2.1 (1.7-2.5) 3.2 (2.6-3.9) 6.4 (4.0-31.1)
Blood pressure, mean (SE), mm Hg
Systolic 118.7 (0.48) 117.2 (0.95) 117.7 (0.83) 119.3 (1.10) 121.2 (0.92) .03 .001
Diastolic 74.1 (0.29) 73.7 (0.51) 74.2 (0.53) 74.2 (0.62) 74.3 (0.62) .86 .79
Hypertension, %
General† 23.0 19.4 20.6 25.5 28.3 .05 .001
Systolic 140 mm Hg‡ 8.4 6.2 6.6 10.4 11.4 .09 .001
Diastolic 90 mm Hg‡ 4.7 3.1 4.1 5.1 7.1 .25 .001
*P values obtained from 2 test (categorical variables) or analysis of variance (continuous variables) based on an overall test across quartiles.
†General hypertension defined as systolic blood pressure of 140 mm Hg or higher, diastolic blood pressure of 90 mm Hg or higher, or self-report of prescription antihypertensive
treatment.
‡Excludes women who reported being currently treated for hypertension.
1526 JAMA, March 26, 2003—Vol 289, No. 12 (Reprinted) ©2003 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on January 3, 2010
14. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
the existing associations between age, strongly associated with general hyper- tile 2: OR, 1.5; 95% CI, 0.61-3.7; quar-
race and ethnicity, alcohol intake, and tension in premenopausal women. tile 3: OR, 2.1; 95% CI, 0.76-5.9; and
BMI. For women in the highest 2 quar- quartile 4: OR, 3.4; 95% CI, 1.3-8.7).
tiles of blood lead level relative to the Systolic and Diastolic Hypertension Stratification by menopausal status re-
lowest quartile, the adjusted ORs of hy- After similar adjustment, a weak asso- vealed a weak dose-response relation-
pertension were elevated but not sig- ciation existed for untreated systolic hy- ship between blood lead level and sys-
nificantly (OR, 1.3; 95% CI, 0.90-2.0 and pertension. For women in the fourth tolic hypertension in premenopausal
OR, 1.4; 95% CI, 0.90-2.0, for quartiles quartile of blood lead, the ORs were the women, and a significantly elevated OR
3 and 4, respectively). Separate models highest (OR, 1.55; 95% CI, 0.72-3.20) of systolic hypertension in postmeno-
of these associations for premeno- (TABLE 5). The adjusted ORs of dias- pausal women in the second and third
pausal women and postmenopausal tolic hypertension relative to women in quartiles of blood lead relative to women
women yielded similar results, with the the lowest quartile of blood lead level in- in the lowest quartile (quartile 2: OR, 3.0;
exception that serum creatinine was creased with a clear dose-response (quar- 95% CI, 1.3-6.9 and quartile 3: OR, 2.7;
Table 3. Unstandardized Regression Coefficients for Blood Lead and Systolic Blood Pressure and Diastolic Blood Pressure in Women Aged 40
to 59 Years Not Treated for Hypertension*
Premenopausal Postmenopausal
All Women Women Women
(N = 1786) P Value (n = 1084) P Value (n = 633) P Value
Systolic Blood Pressure, Regression Coefficients (SE)
R2 0.22 0.22 0.19
Intercept 61.1 (3.88) 57.4 (6.62) 56.1 (7.22)
Blood lead, µg/dL 0.32 (0.16) .03 0.14 (0.26) .59 0.42 (0.21) .29
Age, y 0.70 (0.08) .001 0.77 (0.15) .001 .86 (0.14) .26
Race and ethnicity
Non-Hispanic black −4.01 (1.07) −4.82 (1.4) −3.89 (2.04)
Mexican American −1.25 (0.97) .001 −1.54 (1.31) .002 −1.01 (1.61) .32
Non-Hispanic white 1.0 1.0 1.0
Alcohol use
3 per week −2.10 (1.55) −1.78 (2.12) −2.31 (3.26)
1-2 per week −1.11 (1.39) .32 0.60 (1.37) .77 −5.33 (2.78) .30
1 per week 0.18 (1.12) 0.14 (1.29) 0.86 (2.33)
None 1.0 1.0 1.0
Cigarette smoking status
Current 1.24 (1.09) 0.77 (1.27) 0.72 (1.81)
Former 0.80 (1.43) .32 0.22 (1.89) .83 1.03 (1.43) .20
Never 1.0 1.0 1.0
Body mass index 0.81 (0.08) .001 0.82 (0.10) .001 0.79 (0.17) .001
Serum creatinine 1.10 (1.53) .002 2.64 (0.96) .006 −2.96 (2.50) .01
Diastolic Blood Pressure, Regression Coefficients (SE)
R2 0.14 0.17 0.12
Intercept 56.8 (2.49) 52.2 (4.52) 61.5 (3.91)
Blood lead, µg/dL 0.25 (0.09) .009 0.38 (0.25) .12 0.14 (0.13) .04
Age, y 0.07 (0.05) .13 0.13 (0.11) .25 0.07 (0.06) .001
Race and ethnicity
Non-Hispanic black −1.51 (0.50) −1.93 (0.71) −1.18 (0.89)
Mexican American 0.65 (0.58) .001 0.92 (0.79) .002 0.47 (1.10) .16
Non-Hispanic white 1.0 1.0 1.0
Alcohol use
3 per week −2.04 (0.94) −2.31 (1.62) −1.39 (1.33)
1-2 per week −0.73 (0.77) −0 (0.82) −2.57 (1.43)
1 per week −0.14 (0.73) .18 0.09 (0.97) .54 −0.30 (1.18) .07
None 1.0 1.0 1.0
Cigarette smoking status
Current 2.83 (0.70) 3.25 (0.94) 2.18 (1.21)
Former 1.16 (0.90) .001 1.39 (1.21) .002 0.09 (0.78) .76
Never 1.0 1.0 1.0
Body mass index 0.47 (0.05) .001 0.51 (0.05) .001 0.39 (0.10) .001
Serum creatinine 1.11 (1.02) .28 2.26 (0.58) .001 −2.15 (0.87) .24
*Body mass index is calculated as weight in kilograms divided by the square of height in meters. A total of 69 women could not be assigned a menopausal status due to missing
data.
©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, March 26, 2003—Vol 289, No. 12 1527
Downloaded from www.jama.com by guest on January 3, 2010
15. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
95% CI, 1.2-6.2). A dose-response re- pertension increase markedly in wom- The results are consistent with those
lationship was apparent for blood lead en.39,44 The highest quartile of blood lead of Korrick et al,7 who found an asso-
quartile and diastolic hypertension, (mean, 6.3 µg/dL) was associated with ciation between self-reported hyper-
which was particularly striking for post- a 3.4-fold increase in the risks of dias- tension and bone lead in older women.
menopausal women. tolic hypertension (95% CI, 1.3-8.7) In a study of 45-year-old women liv-
relative to those in the lowest blood lead ing in Copenhagen County, Den-
COMMENT quartile (mean, 1.0 µg/dL). These risks mark, higher blood lead levels were as-
To our knowledge, this is the first study were considerably higher for postmeno- sociated with elevated diastolic blood
to examine the effects of blood lead and pausal women. In addition, blood lead pressure.29 Neither study accounted for
blood pressure in perimenopausal was a significant, positive predictor of menopausal status in either blood lead
women. After accounting for age, race both elevated systolic and diastolic level or hypertension analyses.
and ethnicity, alcohol intake, ciga- blood pressure in these women. A dif- In analyses of systolic and diastolic
rette smoking status, BMI, and kidney ference in blood lead levels between the blood pressures, the relationship be-
function, we found a significant asso- lowest quartile and the highest quar- tween blood lead and blood pressure was
ciation between blood lead and sys- tile was associated with a difference of not stronger for blacks than for whites,
tolic and diastolic hypertension preva- 1.7 mm Hg in systolic blood pressure nor did blood lead levels explain racial
lence among women aged 40 to 59 years and 1.4 mm Hg in diastolic blood pres- differences in hypertension preva-
in the US population. We selected this sure. Blood lead is among the few pre- lence. In fact, the blood lead and hyper-
population to analyze the role of meno- dictors of both systolic and diastolic tension relationships reported ap-
pausal status, which we and others have blood pressures in perimenopausal US peared to be less pronounced among
shown can influence blood lead levels women. Per unit change, blood lead was blacks compared with the cohort as a
in women.14-17,42,43 Furthermore, this is a stronger predictor of diastolic blood whole. However, stratification of the co-
the age range at which the risks for hy- pressure than age. hort by race and ethnicity resulted in
small sample sizes in each blood lead
quartile, limiting precision.
Table 4. Adjusted Odds Ratio of General Hypertension, Stratified by Menopausal Status* The associations of blood lead with
Odds Ratio (95% Confidence Interval) systolic and diastolic hypertension were
Adjusted† much more pronounced for postmeno-
pausal women than for premenopausal
Premenopausal Postmenopausal
All Women All Women Women Women
women. The reasons for this associa-
(N = 2165) (N = 2165) (n = 1214) (n = 850) tion are unclear. Postmenopausal
Blood lead quartile women may be more sensitive to the hy-
1 1.0 1.0 1.0 pertensive effects of lead because of loss
2 1.0 (0.63-1.6) 0.78 (0.38-1.6) 0.73 (0.40-1.3) of estrogen at menopause.44 Estrogen has
3 1.3 (0.87-2.0) 1.4 (0.82-2.4) 1.3 (0.75-2.2) been postulated to protect women from
4 1.4 (0.92-2.0) 1.5 (0.78-2.8) 1.3 (0.68-2.3)
age-related increases in blood pres-
Age, y 1.1 (1.1-1.1) 1.1 (1.1-1.1) 1.1 (1.0-1.1) 1.1 (1.0-1.2)
sure,44 although results from a large
Race and ethnicity
Non-Hispanic black 2.3 (1.7-3.1) 2.2 (1.7-2.9) 2.4 (1.5-3.7) 2.2 (1.5-3.2) randomized clinical trial have not sup-
Mexican American 0.90 (0.60-1.4) 0.90 (0.60-1.3) 1.1 (0.60-1.7) 0.80 (0.40-1.5) ported this hypothesis.45 This observa-
Non-Hispanic white 1.0 1.0 1.0 1.0 tion also may reflect complex relation-
Alcohol use ships between bone lead and blood lead,
3 per week 1.0 (0.60-1.7) 1.0 (0.60-1.8) 0.90 (0.40-1.9) 1.0 (0.40-2.7) which are altered by the changes in bone
1-2 per week 0.90 (0.70-1.3) 1.0 (0.70-1.3) 1.2 (0.80-1.6) 0.70 (0.40-1.4) mineral metabolism that accompany the
1 per week 1.9 (1.2-3.0) 1.9 (1.2-3.1) 1.9 (0.90-4.0) 1.8 (0.90-3.7) menopausal transition.
None 1.0 1.0 1.0 1.0 Whether lead affects blood pressure
Cigarette smoking status through altering kidney function in hu-
Former 0.80 (0.50-1.4) 0.90 (0.50-1.4) 0.80 (0.40-1.5) 0.90 (0.50-1.6)
mans is not known. Lead is nephro-
Current 1.0 (0.70-1.4) 1.1 (0.80-1.6) 1.5 (0.80-2.9) 0.90 (0.50-1.4)
toxic to humans, and alteration of
Never 1.0 1.0 1.0 1.0
kidney function may precede the devel-
Body mass index 1.1 (1.1-1.1) 1.1 (1.1-1.2) 1.1 (1.1-1.1) 1.1 (1.1-1.2)
opment of hypertension.7,8 Kidney func-
Serum creatinine 2.5 (0.60-10.1) 2.3 (0.60-9.2) 7.4 (1.7-32.7) 1.1 (0.50-2.4)
tion, as measured by serum creatinine,
*Body mass index is calculated as weight in kilograms divided by the square of height in meters. General hypertension
defined as systolic blood pressure of 140 mm Hg or higher, diastolic blood pressure of 90 mm Hg or higher, or was found to be significantly positively
self-report of prescription antihypertensive treatment. A total of 101 women could not be assigned a menopausal
status due to missing data. For every unit change in each of these variables (age, body mass index, serum creati- associated with both systolic and dias-
nine), the regression coefficient represents the increase in odds of hypertension for each covariate. tolic blood pressures in premeno-
†Adjusted for age, race, alcohol intake, cigarette smoking status, body mass index, and serum creatinine clearance.
pausal women who are untreated for
1528 JAMA, March 26, 2003—Vol 289, No. 12 (Reprinted) ©2003 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on January 3, 2010
16. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
hypertension. Perhaps this reflects that a significant association between lead sion who had reduced renal function had
the kidney can be a common pathway and general hypertension was not found. significantly more chelatable lead than
for blood pressure regulation, and the The mechanisms of lead-induced hy- those with essential hypertension with
effect of lead on the kidney is only part pertension are not well-characterized, normal renal function.
of the relationship between kidney func- even in animal models. One hypoth- In the present study, kidney func-
tion and blood pressure. However, con- esis is that lead induces hypertension tion measured by serum creatinine did
trolling for kidney function did not re- through direct effects on the kidney. A not appear to mediate the associations
duce the association of blood lead with recent retrospective study of 509 healthy between blood lead and blood pres-
blood pressure and hypertension, as participants of the Normative Aging sure. Thus, lead may act on blood pres-
would be expected if kidney function Study found blood lead levels to be sig- sure through effects on the vasculature
were along the causal pathway. In the nificantly positively correlated with se- or central nervous system, or more sen-
present investigation, serum creatinine rum creatinine levels.41 A study of lead- sitive measures of renal function may
was both a sensitive and significant pre- exposed workers, with high blood lead be required to test mechanistic hypoth-
dictor of general hypertension in pre- levels (mean, 37 µg/dL), reported in- eses. However, Staessen47 reported no as-
menopausal women. For every unit in- creases in diastolic blood pressure and sociation between renal markers of lead
crease in serum creatinine, the risks of in levels of urinary biomarkers for re- toxicity and blood pressure in a large
hypertension increased more than 7-fold nal function.7 Batuman et al46 reported cohort study of women. The magni-
(OR, 7.4; 95% CI, 1.7-32.7). However, that patients with essential hyperten- tude of the effects of blood lead on blood
Table 5. Adjusted Odds Ratios for Hypertension, Systolic Hypertension, and Diastolic Hypertension by Blood Lead Quartile*
Blood Lead Quartile
Quartile 1 Quartile 2 Quartile 3 Quartile 4
All Premenopausal and Postmenopausal Women
No. in sample 568 498 556 543
Blood lead, mean (range), µg/dL 1.0 (0.5-1.6) 2.1 (1.7-2.5) 3.2 (2.6-3.9) 6.4 (4.0-31.1)
General hypertension, OR (95% CI)† 1.0 1.0 (0.63-1.6) 1.3 (0.87-2.0) 1.4 (0.92-2.0)
Premenopausal and Postmenopausal Women Untreated for Hypertension
No. in sample 433 476 438 445
Blood lead, mean (range), µg/dL 0.94 (0.5-1.5) 2.0 (1.6-2.5) 3.1 (2.6-3.8) 6.2 (3.9-31.1)
Systolic hypertension 140 mm Hg, OR (95% CI)‡ 1.0 0.89 (0.41-1.9) 1.4 (0.75-2.7) 1.55 (0.72-3.20)
Diastolic hypertension 90 mm Hg, OR (95% CI)‡ 1.0 1.5 (0.61-3.7) 2.1 (0.76-5.9) 3.4 (1.3-8.7)
All Premenopausal Women
No. in sample 304 302 300 308
Blood lead, mean (range), µg/dL 0.8 (0.5-1.4) 1.8 (1.5-2.1) 2.7 (2.2-3.3) 5.4 (3.4-28.7)
General hypertension, OR (95% CI)† 1.0 0.78 (0.38-1.6) 1.4 (0.82-2.4) 1.5 (0.78-2.8)
Premenopausal Women Untreated for Hypertension
No. in sample 279 277 262 266
Blood lead, mean (range), µg/dL 0.8 (0.5-1.4) 1.8 (1.5-2.1) 2.7 (2.2-3.3) 5.4 (3.4-28.7)
Systolic hypertension 140 mm Hg, OR (95% CI)‡ 1.0 0.88 (0.29-2.7) 1.4 (0.49-3.7) 1.6 (0.62-4.2)
Diastolic hypertension 90 mm Hg, OR (95% CI)‡ 1.0 1.1 (0.31-3.6) 1.8 (0.76-4.2) 3.5 (0.89-13.4)
All Postmenopausal Women
No. in sample 206 227 203 214
Blood lead, mean (range), µg/dL 1.3 (0.5-1.9) 2.5 (2.0-3.1) 3.9 (3.2-4.6) 7.4 (4.7-31.1)
General hypertension, OR (95% CI)† 1.0 0.73 (0.40-1.3) 1.3 (0.75-2.2) 1.3 (0.68-2.3)
Postmenopausal Women Untreated for Hypertension
No. in sample 163 148 166 156
Blood lead, mean (range), µg/dL 1.4 (0.5-2.0) 2.6 (2.1-3.0) 3.8 (3.1-4.6) 7.4 (4.7-31.1)
Systolic hypertension 140 mm Hg, OR (95% CI)‡ 1.0 3.0 (1.3-6.9) 2.7 (1.2-6.2) 2.6 (0.89-7.5)
Diastolic hypertension 90 mm Hg, OR (95% CI)‡ 1.0 4.6 (1.1-19.2) 5.9 (1.5-23.1) 8.1 (2.6-24.7)
Abbreviations: OR, odds ratio; CI, confidence interval.
*Adjusted for age, race, alcohol intake, cigarette smoking status, body mass index (calculated as weight in kilograms divided by the square of height in meters), and serum cre-
atinine.
†General hypertension defined as systolic blood pressure of 140 mm Hg or higher, diastolic blood pressure of 90 mm Hg or higher, or self-report of prescription antihypertensive
treatment.
‡Excludes women who reported currently receiving antihypertensive treatment.
©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, March 26, 2003—Vol 289, No. 12 1529
Downloaded from www.jama.com by guest on January 3, 2010
17. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
pressure observed in this study are simi- The human skeleton is a dynamic set of perimenopausal bone loss resulting
lar to previous investigations, includ- physiological compartment of min- in increased endogenous lead expo-
ing 1 study of women.29 eral metabolism. Women lose as much sure, followed by a chronic effect of lead
Several cross-sectional21-24,29-31,33,48 and as 50% of trabecular bone and 30% of on blood pressure. A study by Cake et
prospective29,31,35 population-based stud- cortical bone during their lifetime, and al61 suggests that bone lead released into
ies on the association of lead with sys- 30% to 50% of this bone loss occurs in the blood may be more bioavailable than
tolic and diastolic blood pressures have the early postmenopausal years.49-53 Es- lead resulting from environmental ex-
been performed from the mid-1980s. trogen deficiency appears to play a sig- posure. Therefore, if blood lead in peri-
The results of these studies have been nificant role in bone loss.51,54 menopausal women is more driven by
mixed, but there is considerable con- Observational evidence suggests that bone lead levels, it is possible that blood
cordance with the directionality of the lead may be mobilized from the skel- lead levels may be a more sensitive pre-
observed associations, with most con- eton during periods of increased bone dictor of blood pressure outcomes in this
sistently finding a weak-positive asso- demineralization, such as during preg- population, because it represents liber-
ciation between blood lead and both sys- nancy and lactation,12,55-57 very old age,58 ated skeletal lead stores.
tolic and diastolic blood pressure in men, and menopause. 14-16,59 Two cross- Important methodological chal-
women, blacks, and whites.4 A meta- sectional studies14,15 of US women that lenges exist in observational studies of
analysis by Schwartz38 of 15 studies of were performed using data from the sec- lead exposure and blood pressure and
lead and systolic blood pressure in men ond NHANES (NHANES II, 1976- hypertension. First, if an association be-
estimated that a change in blood lead 1980) and the Hispanic HANES (1982- tween lead exposure and blood pres-
from 5 to 10 µg/dL was associated with 1984) documented that postmenopausal sure exists, lead is most likely respon-
an increase of 1.5 mm Hg in systolic women have significantly higher blood sible for a relatively small effect on
blood pressure (95% CI, 0.87-1.63 lead levels than premenopausal women, blood pressure, and thus, this associa-
mm Hg), which compares well with the controlling for age and other factors re- tion may be difficult to consistently as-
corresponding estimate from our study lated to exogenous lead exposure. An- certain in different populations. Sec-
(1.6 mm Hg; 95% CI, 0.97-2.20). The other study59 also identified meno- ond, when examining small effects, the
adjusted ORs from multiple logistic re- pausal status as an independent issue of residual confounding, beyond
gression models performed separately for predictor of blood lead levels in a ran- that which is controlled in the analy-
premenopausal and postmenopausal dom sample of Scandinavian women. sis, becomes extremely important. In
women (Table 5) show a consistent, al- Hu et al60 noted that bone lead may such cases, what may be interpreted as
though not always significant, dose- be a more appropriate marker of lead a small effect of blood lead on blood
response relationship between blood exposure for chronic disease out- pressure may actually be due to inad-
lead quartile and risks of hypertension. comes such as hypertension. The pres- equate control of confounding fac-
These subgroup analyses resulted in ent study is a cross-sectional study in tors. However, the restricted age range
smaller numbers of women in the mod- that the exposures and the outcomes chosen for this investigation helps to
els, and this is reflected in the wide CIs were measured simultaneously. The rel- minimize the effect of confounding by
in some of the estimates. evant exposures affecting blood pres- age, which is strongly related to blood
The conventional predictors of blood sure and hypertension may occur lead, blood pressure, and hyperten-
lead in the current US population have months or years before the observed sion. Third, because the mechanisms by
been published in a previous NHANES effect. For example, the average BMI which lead may act on blood pressure
III analysis by Brody et al.10 Other non- during the 5 years preceding the blood in humans are not well understood, in-
bone density–related exposures that can pressure measurement may have more vestigators may tend to include more
result in elevated blood lead levels in explanatory power than BMI mea- covariates than necessary in their mod-
the United States include residential ex- sured on the same day as the blood pres- els or use mechanical, stepwise ap-
posure to lead paint, residential prox- sure. Likewise, cumulative lead expo- proaches to modeling. The true size of
imity to a lead smeltering facility, oc- sure during the preceding decade, bone the effect may be decreased by over-
cupational exposure (lead smelter, lead burden, or serum creatinine may controlling. This is a particular prob-
battery manufacturing, welding, or be more predictive of blood pressure lem in studies of environmental lead ex-
bridge painting), cigarette smoking, and than blood lead level measured on the posure, because blood lead levels are
alcohol intake.15 Those variables asso- same day as blood pressure. Evidence highly correlated with race and ethnic-
ciated with lead and also known to be suggests that bone lead stores can con- ity, income, and education,10 which also
associated with blood pressure and hy- tribute to circulating levels of lead in may be risk factors for outcomes such
pertension (ie, potential confound- blood.11-13 as hypertension.49
ers) were adjusted for in the blood lead The findings of our study are incon- Whether bone or blood is the appro-
and blood pressure and hypertension sistent with the notion of a latency pe- priate biomarker for lead exposure in
analyses of our study. riod of months to years between the on- studies of chronic disease outcomes is
1530 JAMA, March 26, 2003—Vol 289, No. 12 (Reprinted) ©2003 American Medical Association. All rights reserved.
Downloaded from www.jama.com by guest on January 3, 2010
18. BLOOD LEAD, BLOOD PRESSURE, AND HYPERTENSION IN WOMEN
uncertain.60 Blood lead is a marker of blood lead levels than our data in our hypertension in the US adult population: results from
the Third National Health and Nutrition Examination Sur-
relatively recent exposures to lead. population, the effects observed in our vey, 1988-1991. Hypertension. 1995;25:305-313.
Hypertension in adults that may be asso- study also suggest that lead acts on the 4. Hertz-Picciotto I, Croft J. Review of the relation be-
tween blood lead and blood pressure. Epidemiol Rev.
ciated with past exposures to lead is con- cardiovascular system and much lower 1993;15:352-373.
sistent with a follow-up study of lead- levels in the blood. 5. Nowack R, Wiecek A, Exner B, Gretz N, Ritz E.
poisoned children in whom the risks for From a public health perspective, the Chronic lead exposure in rats: effects on blood pres-
sure. Eur J Clin Invest. 1993;23:433-443.
hypertension were significantly higher most important and troubling implica- 6. Vander AJ. Chronic effects of lead on the renin-
than they were in controls matched by tion of these findings is that lead ap- angiotensin system. Environ Health Perspect. 1988;
78:77-83.
age, sex, race and ethnicity, and neigh- pears to increase blood pressure in 7. Korrick SA, Hunter DJ, Rotnitzky A, Hu H, Speizer
borhood.50 Bone lead is a more appro- women at very small increments above FE. Lead and hypertension in a sample of middle-
aged women. Am J Public Health. 1999;89:330-
priate marker for chronic exposure; 1.0 µg/dL, well below what is consid- 335.
however, its interpretation depends on ered deleterious in adults. The mean 8. Cowley AW Jr, Roman RJ. The role of the kidney
an understanding of bone physiology blood lead level in this sample of in hypertension. JAMA. 1996;275:1581-1589.
9. Hu H, Aro A, Payton M, et al. The relationship of
and events such as pregnancy and women was 2.9 µg/dL. These results bone and blood lead to hypertension: the Normative
menopause.11 Future studies of blood demonstrate effects of lead at levels less Aging Study. JAMA. 1996;275:1171-1176.
10. Brody DJ, Pirkle JL, Kramer RA, et al. Blood lead
pressure and hypertension should con- than the US occupational blood lead ex- levels in the US population: phase 1 of the Third Na-
sider blood lead and bone lead as inde- posure limits (40 µg/dL) and even less tional Health and Nutrition Examination Survey
pendent factors influencing the risk for than the current Centers for Disease (NHANES III, 1988 to 1991). JAMA. 1994;272:277-
283.
hypertension. Control and Prevention level of con- 11. Gulson BL, Mahaffey KR, Mizon KJ, et al. Con-
The R2 values in Table 3 suggest that cern for preventing lead poisoning in tribution of tissue lead to blood lead in adult female
subjects based on stable lead isotope methods. J Lab
the models explain 22% and 14% of the children (10 µg/dL). Finally, the find- Clin Med. 1995;125:703-712.
variation in systolic and diastolic blood ings from our study of associations of 12. Gulson BL, Jameson CW, Mahaffey KR, et al. Preg-
nancy increases mobilization of lead from maternal skel-
pressure, respectively. We interpret this blood lead with systolic and diastolic eton. J Lab Clin Med. 1997;130:51-62.
to mean that much of the variation in hypertension and blood pressure among 13. Smith DR, Osterloh JD, Flegal AR. Use of endog-
blood pressure is random or due to un- women in the general population lend enous, stable lead isotopes to determine release of lead
from the skeleton. Environ Health Perspect. 1996;
known or immeasurable factors. Be- support for further studies on the health 104:60-66.
cause blood pressure has been so well effects of bone lead mobilization dur- 14. Silbergeld EK, Schwartz J, Mahaffey K. Lead and
osteoporosis: mobilization of lead from bone in post-
studied, it is unlikely that there are un- ing the menopausal transition. These menopausal women. Environ Res. 1988;47:79-94.
discovered factors responsible for the results provide support for continued 15. Symanski E, Hertz PI. Blood lead levels in rela-
remaining unexplained variation. efforts to reduce lead levels in the gen- tion to menopause, smoking, and pregnancy history.
Am J Epidemiol. 1995;141:1047-1058.
Other factors that contribute to the eral population, especially women. 16. Muldoon SB, Cauley JA, Allen L. Effect of bone
variation in blood lead levels observed mineral density changes on blood lead levels in peri-
Author Contributions: Study concept and design: menopausal women. Paper presented at: Society for
in our study include measured and un- Nash, Magder, Lustberg, Sherwin, Rubin, Kaufmann, Epidemiologic Research; June 1997; Calgary, Al-
measured aspects of conventional and Silbergeld. berta.
Acquisition of data: Nash. 17. Nash D, Silbergeld E, Magder L, Stolley P. Meno-
bone density–related predictors of blood Analysis and interpretation of data: Nash, Magder, pause, hormone replacement therapy (HRT), and blood
lead, as well as other variables that were Lustberg, Rubin, Kaufmann, Silbergeld. lead levels among adult women from NHANES III,
not measured by NHANES. However, Drafting of the manuscript: Nash, Sherwin, Rubin, 1988-1994 [abstract]. Am J Epidemiol. 1998;147:
Kaufmann, Silbergeld. S93.
we controlled for all of the known fac- Critical revision of the manuscript for important in- 18. Silbergeld E, Nash D. Lead and human health: is
tors associated with blood pressure and tellectual content: Nash, Magder, Lustberg, Sherwin, this mine exhausted. Progress in Environmental Sci-
Rubin, Kaufmann, Silbergeld. ence. 2000;1:53-68.
hypertention,3 including alcohol in- Statistical expertise: Magder. 19. Pirkle JL, Brody DJ, Gunter EW, et al. The decline
take. This approach presumably mini- Obtained funding: Nash, Silbergeld. in blood lead levels in the United States: the National
Administrative, technical, or material support: Health and Nutrition Examination Surveys (NHANES).
mizes residual confounding of our es- Lustberg, Silbergeld. JAMA. 1994;272:284-291.
timate of the associations of blood lead Study supervision: Nash, Sherwin, Rubin, Kaufmann, 20. Harlan WR. The relationship of blood lead levels
Silbergeld. to blood pressure in the US population. Environ Health
with blood pressure and hypertension. Funding/Support: This study was supported with an Perspect. 1988;78:9-13.
Vital status data are not yet avail- award from the Centers for Disease Control and Pre- 21. Harlan WR, Landis JR, Schmouder RL, et al. Blood
able on the NHANES III cohort. How- vention/Association of Teacher’s of Preventive Medi- lead and blood pressure: relationship in the adoles-
cine cooperative agreement TS 288-14/14 and by a cent and adult US population. JAMA. 1985;253:530-
ever, a recent analysis of men and grant from the Heinz Family Foundation. 534.
women from the NHANES II cohort by 22. Pocock SJ, Shaper AG, Ashby D, et al. The rela-
Lustberg and Silbergeld62 found el- tionship between blood lead, blood pressure, stroke,
REFERENCES
and heart attacks in middle-aged British men. Envi-
evated blood lead levels to be associ- 1. Pirkle JL, Schwartz J, Landis JR, Harlan WR. The ron Health Perspect. 1988;78:23-30.
ated with a dose-related increase in relationship between blood lead levels and blood pres- 23. Neri LC, Hewitt D, Orser B. Blood lead and blood
sure and its cardiovascular risk implications. Am J Epi- pressure: analysis of cross-sectional and longitudinal
deaths due to hypertension-related demiol. 1985;121:246-258. data from Canada. Environ Health Perspect. 1988;
coronary heart disease and stroke for 2. Kopp SJ, Barron JT, Tow JP. Cardiovascular ac- 78:123-126.
tions of lead and relationship to hypertension: a re- 24. Elwood PC, Davey-Smith G, Oldham PD, Toothill
both men and women. Although the view. Environ Health Perspect. 1988;78:91-99. C. Two Welsh surveys of blood lead and blood pres-
timing of NHANES II resulted in higher 3. Burt VL, Whelton P, Roccella EJ, et al. Prevalence of sure. Environ Health Perspect. 1988;78:119-121.
©2003 American Medical Association. All rights reserved. (Reprinted) JAMA, March 26, 2003—Vol 289, No. 12 1531
Downloaded from www.jama.com by guest on January 3, 2010