38 Hypertension


Published on

Published in: Health & Medicine
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

38 Hypertension

  1. 1. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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. 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