Taking a Family History of Hypertension: Is There Patient Care Benefit?

Authors


Joel Handler, MD, Kaiser Permanente, 411 Lakeview Avenue, Anaheim, CA 92807
E-mail: joel.handler@kp.org

Case 1: A 74-year-old woman was evaluated in hypertension clinic for difficult-to-control hypertension despite documented adherence to losartan, amlodipine, and atenolol. Family history was positive for hypertension in both parents and all siblings (3 sisters and 1 brother). There was no personal history of target organ damage, and no family history of myocardial infarction, stroke, or diabetes. Hydrochlorothiazide was added to her treatment regimen and hypertension was controlled.

Case 2: A 72-year-old man was referred to hypertension clinic because of difficult-to-control hypertension. Antihypertensive medication had been initiated a couple of months prior to evaluation. Most of the time, home systolic pressures were <130 mm Hg, but, taken several times daily, systolic blood pressures (BPs) were occasionally >170 mm Hg, prompting visits to the urgency care clinic and emergency department. BP elevations were associated with facial flushing, a feeling of dizziness in the head, and occasional abdominal bloating. He had been taking losartan 50 mg daily, and was unable to tolerate the addition of metoprolol due to low BPs. Family history was strongly positive. His mother and 7 of 10 siblings had hypertension, and his mother, a brother, and a sister all had strokes.

Discussion revealed that the patient was extremely anxious because of his recent diagnosis of hypertension and positive family history of hypertensive stroke. Counselling led to gradual amelioration of his panic attacks and reduction of home BPs. BP spikes resolved, and hypertension was controlled on losartan monotherapy.

Case 3: A 31-year-old woman was referred to hypertension clinic due to refractory hypertension and hypokalemia associated with hyperreninemia and hyperaldosteronism. Antihypertensive medication was initiated at age 17. She had aldosterone levels of 31 ng/dL, 47 ng/dL, and 80 ng/dL (normal 3–16 ng/dL) and renin levels of 11.3 ng/mL/h, 11.1 ng/mL/h, 15 ng/mL/h, 22.3 ng/mL/h, and 22.8 ng/mL/h (normal 0.25–5.82 ng/mL/h) with potassium replacement, and a transtubular potassium gradient of 11. Results for 24-hour urinary metanephrine, plasma catecholamine fractionation, computed tomography of the abdomen and pelvis, metaiodobenzylguanidine, and a chromosomal assay for glucocorticoid remedial hyperaldosteronism were all normal. Deoxycorticosterone level was 12 ng/dL (normal 3.5–13 ng/dL), and adrenocorticotropin level was 16 pg/mL (normal 7.2–63.3 pg/mL). She was seen in the emergency department on 2 occasions for hypertensive urgency. While taking valsartan 80 mg, hydralazine 25 mg twice a day, and labetalol 50 mg twice a day, BPs were 164/97 mm Hg, 183/11 mm Hg, and 165/98 mm Hg. She had been intolerant to spironolactone and amlodipine.

The patient has an identical twin sister with refractory hypertension, hypokalemia, hyperreninemia, and hyperaldosteronism who had been evaluated at the University of California, Los Angeles, with the same extensive negative workup, including negative bilateral renal vein renin sampling, which the patient refused because the study was nonrevealing in her sister. Renal artery magnetic resonance angiography (MRA) was normal, and the patient refused renal angiography.

She agreed to increase valsartan to 120 mg, hydralazine to 50 mg twice a day, and labetalol to 100 mg twice a day, then 200 mg twice a day following home BPs, but she became hypotensive with this progression and back-titrated the labetalol. Her twin sister’s BP remained uncontrolled on maximal doses of these 3 drugs.

Discussion

Family History of Hypertension Is a Lower-Value Risk Assessment Tool

Taking a family history is a routine feature of the hypertension consultation and is mostly used as a risk assessment tool, although generally it has low utility. Approximately 50% of clinicians routinely collect a family history.1 The value of the family history has been evaluated in a National Institutes of Health (NIH) state-of-the science conference statement where several questions were posed and then investigated.1 Most commonly, basic information is obtained enumerating first-degree relatives with a certain disease or condition because sensitivity/specificity, disease-specific data have been collected for first-degree relatives, and data for second-degree and more-distant relatives lose accuracy. High sensitivity, defined as probability a first-degree relative will have the disorder, and high specificity, defined as the probability a first-degree relative will not have the patient’s disorder, have been associated with mood disorders and depression.

The family history results for hypertension in several studies are conflicting. The Bogalusa Heart Study found that family history of hypertension independently predicted 9-year follow-up systolic BPs in both black and white families of this biracial community.2 Nonetheless, follow-up of Medical Research Council patients in Great Britain concluded that family history data were too weak to justify directed preventative strategies.3 A family history study conducted in Australia showed that positive paternal history was the stronger parental predictor of offspring hypertension,4 but the Minneapolis children’s BP study revealed a stronger parental association between maternal and offspring hypertension.5

Familial aggregation of BP levels is well established.6 There is approximately a doubling of risk of hypertension in children of one parent with hypertension and a quadrupling of risk when both parents have hypertension.7 The level of parental BP is also a predictor of the BP of their children. In a population-based study of 596 children aged 5 to 19 years, systolic BP was significantly higher in patients with parents in the highest tertile of BP compared with those with parents in the 2 lower tertiles. However, the difference between the highest and lowest parental systolic BP tertiles was 2.72 mm Hg, and systolic BP in the offspring only rose 0.09 mm Hg for every millimeter increase in maternal systolic pressure.6

The NIH conference queried direct evidence to evaluate the effect of family history on improving health outcomes or resulting in adverse outcomes. Benefit has primarily been demonstrated, identifying rare genetic syndromes, such as those for breast and ovarian cancers. Detriment is primarily psychological, associated with the burden of poor outcomes in family members with the same condition.

The Predictive Power of Age-Related Hypertension Demographics

Age-related demographics of hypertension overpower contributions from the family history. The 2008 National Health and Nutrition Examination Survey revealed a 29.0% prevalence of hypertension in patients aged 18 years and older.8 Hypertension prevalence is strongly related to age and, irrespective of family history, the residual lifetime risk for developing hypertension for a 55-year-old without hypertension is 90%9 (Figure). In a predictive model using family history, an adolescent in the top BP quintile with a positive family history of hypertension would have a 37% chance of developing hypertension as an adult compared with an adolescent without a family history who would have a 20% risk. Sensitivity of the model was 65% and specificity 73%. Therefore, the demographic risk model has application for a larger number of patients and is more practical because adolescent BPs are unknown when adults are evaluated for hypertension.10

Figure Figure.

 Residual lifetime risk of hypertension in women and men aged 65 years. Reprinted from the American Medical Association.9

What Can We Learn by Comparing Hypertension Prevalence in Monozygotic and Dizygotic Twins?

Several studies of hypertension comparing monozygotic and dizygotic twins are available, and there is heterogeneity in the results. In a twin registry maintained by the National Academy of Sciences National Research Council, genetic factors were responsible for 59% and environmental factors were responsible for 41% of observed hypertension.11 In the Danish Twin Registry, there was a similar prevalence of hypertension in monozygotic and dizygotic twins.12 The Pittsburgh Twin Study showed that covariation of BP and lipid profile status was less important than genetic and environmental effects on the individual risk factors.13 However, the National Heart, Lung, and Blood Institute (NHLBI) Twin Study offered additional insight, with results showing that monozygotic dyslipidemic hypertension was primarily genetic, except when discordant weight gain occurred.14 An NHLBI study of hypertension prevalence in World War II male veteran twins revealed that the clinical manifestation of hypertension was closely dependent on lifestyle behaviors such as weight gain, exercise, and alcohol consumption.15

Have Genomic Studies of Hypertension Had Clinical Impact?

Population-based twin studies might generate expectation that exploration of the human genome would identify specific genotypes responsible for hypertension, but that potential has not been fulfilled. Genomic association studies of single-nucleotide polymorphisms have not found variants with clinical utility for the management of common diseases.16 Rapid and less expensive DNA sequencing will promote additional research, but, at the genomic level, the contribution of heredity in complex conditions may only be 5% to 10%.17 A study showing that a variant of the adducin gene associated with renal sodium reabsorption and salt-sensitive hypertension could predict hypertensive patients with a lower likelihood of myocardial infarction and stroke when treated with diuretics rather than other antihypertensive drug therapy has not been confirmed.18,19

There is a short list of rare monogenetic diseases definitely associated with specific mutations that cause hypertension: glucocorticoid-remediable aldosteronism, “apparent mineralocorticoid excess,” hypertension exacerbated by pregnancy, and Liddle syndrome.20 A chromosomal assay is available to diagnose glucocorticoid-remediable aldosteronism, apparent mineralocorticoid excess occurs in the absence of circulating aldosterone, and Liddle syndrome is associated with low renin and aldosterone levels. Gene mutations are also responsible for hypertensive congenital adrenal hyperplasia syndromes, in which cases deoxycorticosterone excess is associated with high adrenocorticotropin with low renin levels, and hereditary hypertension with hyperkalemia or Gordon’s syndrome, which is associated with sodium/potassium retention and low renin.

Rare inherited syndromes associated with hypertension are multiple endocrine neoplasia (MEN) 1, MEN 2, and VonHippel-Lindau syndrome.

What Is the Relationship Between Hypertension and Anxiety Syndromes?

Careful diagnosis of hypertension is important because the diagnosis, as opposed to the designation of BP elevation, can be associated with psychological distress.21 There appears to be a more frequent association of panic attacks in patients with hypertension as opposed to patients unaware of a diagnosis of hypertension.22–24 That association may be an artifact of detection because the finding of a significant spike in BP is a prominent marker of an anxiety attack missing in nonhypertensive patients who experience less well-recognized panic attacks who are not taking their BP. Severe episodic BP elevations in patients with hypertension who experience panic attacks has been called pseudopheochromocytoma.25

Review of the Value of These Family History Cases

Case 1 is a 74-year-old patient with a strongly positive family history of hypertension, which included both parents and 4 siblings. The diagnosis of essential hypertension would not be in doubt even with a completely negative family history. However, the family connections for disease in this patient could be used to reinforce the need for medication adherence. By taking a family history, the physician is aware of opportunities to build social support for compliance.26

Case 2 is a patient with recently diagnosed hypertension and panic attacks that led to significant BP spikes and a perception of uncontrolled hypertension. The family history enabled the consultant to identify an additional patient concern regarding hypertensive stroke in close family members. Further root cause discussions resulted in sufficient anxiety reduction to halt the panic attacks and demonstrate that his hypertension was adequately controlled.

Positive history of the same unusual hypokalemic hypertensive syndrome in an identical twin sister, both occurring as teenagers, in case 3, led to an extensive metabolic workup that excluded glucocorticoid-remediable aldosteronism and congenital adrenal hyperplasia syndromes. Hyperreninemia was the cause of hyperaldosteronism in this patient with a differential diagnosis including rare renin-secreting tumors and familial renal fibromuscular dysplasia.27,28 Renal artery MRA has better sensitivity for proximal atherosclerotic renal artery occlusion than distal disease more closely associated with fibromuscular dysplasia. Therefore, the negative findings on MRA did not rule out renal arterial dysplasia, but the patient refused more definitive renal angiography. Reninomas, tumors of the renal juxtaglomerular cell apparatus, are diagnosed by renal imaging studies and selective renal vein renin sampling.27 Reninomas are not known to be familial, and both twins had negative imaging studies, with negative selective renal vein renins in the twin sister. Because this patient’s hypertension responded to medical therapy and renal function remained normal, additional workup was not pursued.