JACC Vol. 12. Xo. 5 November 19RR:127; 8o

1273

EPIDEMIOLOGIC STUDIES

Risk Factors in Siblings of People Wit Heart Disease

reammature Coronary

DIANE M. BECKER, ScD, MPH,* LEWIS C. EECKER, MD, FACC,t THOMAS A. PEARSON, MD, PHD.*$ DAN J. FINTEL, MD, FAX,? DAVID M. LEVINE MD,* PETER 0. KWITEROVICI-I, MD4 Baltimore, Maryland

Prior studies of the contribution of coronary disease risk factors to familial aggregation of premature coronary disease may have underestimated risk factors by relying on self-reported risk factor prevalence levels or, when risk factors have been measured, by using cut points in excessof the9Otb percentile. To determine the actual prevalenceof hyperlipidenda, hypertension and diabetes, and the awareness of these coronary risk factors in unaffected family members, 150 apparently coronary disease-free sibling of 86 people who had documentedcoronary dise@e before 60 years uf age were studii. LUI subjects participated in a I day screetig preceded by a self-administered risk factor questhnmaire and a m interview. Participation of both the b&x patients and sibtings exceeded 86%. With the use of nationally established recommendations for bbod pressure and lipids, which are based on coronary dii risk curves, scmning revealed that 48% of brothers and 41% of sisters were hypertensive,45% of brothers and 22% of sbters bad a lipid abnormality, 38% of siblings

Coronary heart disease has long been shown to cluster significantly within families. First degree relatives of probands who had coronary heart disease events before 60 years of age exhibit a risk ofpremature coronary disease that is 2 to 12 times that of the general population (I-6).Considerable controversy exists concerning the relative indepen-

.

Fmm the Divisions of ?? lntemaI Medicine and tCardioloav. I~. Department of Medicine; flhe Lipid Research Clinics. Department of Pediatrics; and Uhe Deoartment of Euidcmioloav. The Johns Hookins Medical Institutions, Baltin&e. Maryland. This study was supported in part by Preventive CardioIagy Academ;: Award No. KO7HLOI 13 and Grant No. RRllUOJ5 CLINFO (for Computational assWnce) from the National Institutes of Health, Bethesda, from the DuPont Corporation. North Bilkrica, MassaMaryland and a @nant chusetts. Manuscript received January 8,1988; revised manuscript received May 5. 1988, accepted June 3, 1988. A&&&r@& Diane M. Reeker, ScD. MPH, Preventive Cardiology Promms, The Johns Hopkins Hospital, 1830 East Monument Street. Baltimore, Maryland. 01988 by the Americnn College of Cardiology

were current cigarette smokers and 4,?‘% were d&&tic. Two or more risk factors were present in 42% of bro&ers and 25% of sisters. More than 75% of siblings had ooe or more risk factors that woukl require iatervent&L Whea compared with a race-, gender- and age-matched referem population from the Lipid ResearchClinks Prevateace

Study, distributiowsfor b&d presses and for tetal &HI low density lipoproteb~cb&steral were Mgber for the Gbliags in every gender SHI age group. !Sibli&gs were generally unawareoithepresenceofriskfaetars,oZdy43% of hypertensive siblii and 27% of si with @id abnormalities were aware of these risk factors. PreGously u&tected risk factors were highly prevatent in sibiiigs of pe@e witb premature comzary d&ease. These data support cnrreH naGonal recommer&lat&msfor comprehensive risk factor screen& io this fSs@ i&?&i. able high risk fw group. (J An Cdl Cardial19&?;12:1273-80)

dent contribution of unexplained genetic factors and known coronary heart disease risk factors that aggregate within families. Many investigators (7-11) contend that a positive family history remains the most important independent predictor for premature coronary heart disease, whereas a few (12,131 suggest that the major portion of familjaiclustered coronary heart disease can be attributed to in,‘lerited and environmental risk factors. A recent review of methodologic issues in family studies (14) nu!ed that one problem in the majority of prior studies is the reliance on self-reporting or reporting by other relatives of the presence or absence of risk factors in family membersof coronary heart disease probands.Another common problem in prior studies is the use of blood pressure and lipid cut points that exclude values that have been shown fo be associated with increased risk of coronary disease. The existing studies, focusing predominantly on epidemi0735-1097/W13.50

1274

BECKGR El AL. CORONARY RISK FACTORS

IN FP

‘LIES

ologic issues, may leave the clinical community confused about the importance of screening for remediable risk factors in families. The objectives of this study were I) to determine the true prevalence of modifiable coronar)’ heart disease risk factors including hypertension, hyperlipidcmia, diabetes and cigarette smoking using contemporary guidelines, and 2) to determine the awareness of the presence of potentially modifiabie risk factors in apparently healthy aduk siblings of people with documented coronary heart disease. Further, this study provides the first empirical estimates in the United States population of the yield of multiple risk factor screening in unaffected siblings soon after a premature coronary heart disease event in a young brother or sister. Methods Index coronary heart disease patients. Index patients were eligible if they had a documented coronary disease event before 60 years of age. The majority (93%) of participating index patients were identified between January 1983 and June 1986 during a coronary heart disease hospitalization. An additional 7% were identified in a cardiac rehabilitation program after a recent coronary heart disease event. All but two had angiographicdily documented coronary disease; these two patients had sllstained a documented myocardial infarction. The mean age of index patients was 49 + 1I years; 83% were male. Risk factor distributions were not analyzed in index patients becalrse 6% had experienced coronary disease events before the hospitalization. TLese patients were receiving anti-ischemic medications that po tentially confound risk factor prevalences and awareness. Of 90 eligible index patients, g6 (95%) provided access to siblings. Asymptomatic siblings. The 86 probands yielded 172 asymptomatic biologic siblings 30 to 59 years ofage (mean 45 2 7.5); 51% were male and 99% were white. The majority (78%) were high school graduates and 20% were colleie graduates. All siblings c60 years of age without known coronary heart disease, as determined by telephone interview with verification by the physician of record, were invited for a cardiovascular history, physical examination and laboratory evaluation. Of 172 eligible GbIings, I50 (87%) participated in the screening evaluation, which was performed an average of 2 months (range 2 weeks to 4 monrhs) after hospital discharge of the index patient. The majority (19) of the remaining 22 siblings cited geographic distance from the

screening site as the reason for refusal. Screening protocol. A self-administered questionnaire was mailed to siblings before the screening appointment to

elicit self-reporting of the presence or absence of hypertension, hyperlipidemia, diabetes, cigaret!e smoking and curren: and past medications and illnesses. These items were again elicited in a personal interview by a research nurse just

JACC Vol. 12. No. 5 Novcmbcr 1988: 1271-80

before the screening examination (see Appendix). The sibling was considered aware of the presence of a risk factor if he or she indicated “possibly yes” or “yes” to the elicited risk factor information either on the questionnaire or during the interview. Physical examination and laboratory evaluation. AII siblings received a history and physical examination by a cardiologist according to a standardized protocol. Bloqd

Uo amalehaJd lenw

a41 ueql

:amill

WM sJals!s pue ~~aql0Jq

qloq u! i?!wap!d![Jadhq JO awa[PtaJd *A;![euwuqe

‘uals!b

sJals!s salad

< [owsa[oqa

SAM [p/W B[

pauodaJg[as

aye,

p!dy WapAaJd ]sOw ayi

~(31 SdtloJ% qloq u! ‘(m-0

= d)

c,a!qletmouqe p!dg paiewJowap (%sb) sJaqloJq wql (%ZZ) pue sJaqtoJq JOJ%u!uaaJX uo sma[qoJd p!dy JO WlalEA3Jd

aqa

srCalds!p z

aJll8!d

‘(10’0

=

d)

SJaFjs JOJ lP@U +l $ Z9 PUe sJaqlOJq JOJ lP@U l I T Lb s&n [oJalsa[oqa *f(IH [OJalSa[Oqa l(IH *[p@

WE


S[SAS[

pI2q SJaqlOJq JO oh pUlZ SJalS!S aq, JO aUON

auou pur! [p@

%uoIue !l[[JA!lXXkaJ @OS put? ()sZ
I60 to 170 mm He systdlic and >I@0 to 110 m-m Hn diastolic level. When both the Lipid Research Clinics Prevatence Study reference population percentile distributions and risk-gssociated i984 guidelines of the Joint National Committee on High Blood Pressure are used. it is clear that brotbnrs and sisters of patients with documented coronary arterv disease before aee 60 have a high prevalence of b&d pressure levels that aYe Figure 2. The prevalence of lipid abnormalities in siblings according to the National Cholesterol Education Program Adult Treatment Gr::‘zlines and the Triglyceride Consensus Conference Guidelines (A, brothers: B, sisters). High risk low density lipoprotein cholesterol = ~160 mg/dl; low high densiiy lipoprotein cholesterol = 535 mgldl;

high triglycerides

= 5250

mgMl.

1277

associated with increased coronary heart &sea% risk and may require medical evaluation and treatment. Brothers had a greater prevalence of hypertension and of “high normal”

Hyperfiprdem,.a cm Scre-nmg” Brothers

BECKER ET AL. RISK FACTORS IN FAMILIES

fied

blood

as having

pressure. “normal”

Only

31%

pressure

of brothers on the

were

average

classiof three

taken several hours apart. of these. 5% were already receiving antihypertensive treatment, yielding only 26% “true normal” subjects among brothers. The majority of brothers and almost hal of the sisters were unaware of blood pressure elevations. A potential problem in analyzing the awareness data in our study is the studies and national initiatives proliferation of rewch targeting blood pressure education for both the general population and physicians between 1983 and the present. There is no evidence from our study to sugest that hypertension detection in the sibling group had markedly improved between 1983 and 1986. it is unlikely that the prevalence of hypertension in siblings was decreasing over the years of the study or that it is significantly lower at this time. However, new initiatives of the National High Blood Pressure Education Program may result in improved detection of high blood pressure in the coming years. Plasma lipids. The LDL cholesterol cut point of 160 mg/ dl was used to define hvpercholesterolemia on the basis of the National Cholesterol Education Program’s Adult Treatment Guidelines. Coronary heart disease risk curves demonstrate a sharp increase in the slope of risk at total cholesterol levels of 240 mg/dl, which correspond approximately to an LDL cholesterol cut point of 160 mg/dl !21,23,24). Although triglyceride elevations have not been independently associated with coronary disease In most studies, levels of 2250 mgldl classified as elevated in this study represent the approximate !Mth percentile for the sibling age groups in this study. Airhough only a small percentage of brothers (6.5%) and sisters (1.4%) were considered to have abnormal lipids solely on the basis of triglyceride levels, including those siblings as hyperlipidemic may be justified because d the increased likelihood that trigiyceride elevation in the sibling sample represents familial combined hyperlipoproteinemia, which is associated with a macke~~ly increased coronary heart disease risk (27). “High risk” L:IL cholesterol was the most frequent abnormality cb+ KQ *II both men and women. As with hypertension, abnormai lipids were more common in brothers than in sisters with 54% and 78%. respectively. exhibiting desirable lipId patterns on screening. twareness of elevated lipid levels was low in both brci5ers and sisters with 73% unaware. The National Chelesterol Education Program physician initiatives tPegan in 1987 with publication of the Adult Treatmen! GuiMines. Improved practice patterns and increased public education efforts may improve awareness of elevated lipid levels in the future. Data from this study suggest. however, that sign& cant awareness changes did not occur between 1983 and

rest measurements

1278

BECKER ET At. CORONARY RISK FACTORS

Table 5. Smoking

Prevalence

JACC Vol. 12. No. 5 Novcmbcr 19llR:127340

IN FAMILIES

end Number

of Cigarettes

in 150 Siblings of 86 Premature

Coronary

Current smokers (5%) Number of

40 IO 49

Heart

Disease

Prabands

Sisters (age. yr)

Brothers (age. yr) 30 to 39 (n = 241

(n = 32)

so to 59 tn = 21)

Total tn = 77)

30 to 39 In = WI)

40 IO 49 tn = 29)

50 IO 59 (n = 28)

(n = 73)

42 17 r 29

44 18 + 27

29 21 238

39 19 + 31

38 8 f I2

41 I6 ? 25

32 12 f 24

37 13 f 22

Told

cigarettes per day

1986. This finding is further substantiatedby data from the National Institutes of Health Cholesterol Awareness Surveys of both the gergeralpublic and physiciansin 1983 and 1986, which dernol;c!mred only modest changes in the number of people who had blood cholesterol levels checked and no increase between 1983 and 1986 in the number of people who were told that cholesterol levels were elevated (28,29). Cigarettesmoking. The prevalence of current smoking in both brothers and sisters was higher than the 27 to 30% of the general population who smoked cigarettes during the sibling screening period, 1983 to I986 (30). Prior studies (26,311 in high risk families have reported a wide range of smoking prevalence levels in sisters (I5 to 53%) and consist-

Figure3. The prevalence

of major risk factors in siblings (A, brothers; 8, sisters). Lipid abnormality = low density lipoprotein cholesterol B 160 mg/dl. or triglycerides 250 mg/dl or on a lipidlowering drug; hypertension = blood pressure 2 140190 or on antihypertensive drug: smoking = current regular cigarette smoker.

FACTORS

!X?U

P

ent levels of approximately45% in brothers. Most of these studies have not been done in the American population. Although in our study the prevalence rates were similar in brothersand sisters, brothers appeared to be heavier smokers. Given that there is evidence of a multiplicative interacties between cigarette smoking and family history of’ early coronary disease (32,33), it is likely that smoking prevalence alone does not reflect the true importance of smoking as a risk factor in the sibling population. The validity of selfreporting was not examined, but prior studies (35) suggest that the bias is in favor of underreporting, possibly related to a perception of social undesirability. Diabetes. The reported diabetes prevalence of 4.7% agrees with prior studies in families with coronary heart disease, which have shown a prevalence rate of I to 5% (26,31). Although this was a low frequency risk factor, a small number of siblings with probable diabetes were unaware of their condition. MuMpIe risk factors. The majority of brothers and sisters had at least one of the major risk factors for coronary heart disease. It is possible that other more recently reported familial risk factors, such as abnormal piatelet activation (32) and apolipoprotein abnormalities (35), may be prevalent in those siblings who appear “normal” on screening. As it is generally agreed that premature coronary heart disease is of multifactorial origin, it is not surprising that such diverse risk factor patterns exist in high risk families. Male siblings clearly have a higher prevalence of all risk factors; 42% exhibited two or more of the three major risk factors. Conclusions. This study demonstrates that the age-, gender- and race-adjusted distributions of blood pressure and plasma lipids are considerably higher in clinically unaffected siblingsof peopie Gth premature coronary heart disease than in a general reference population. The majority of siblings have one or more risk factors as assessed by contemporary nationally established guidelines for determining risk levels of blood pressure and plasma lipids; further, mer;y were uMware of the presence of potentially modifiable risk factors. Because high risk families such as these account for >50% of coronary heart disease before 55 years of age i36.37). intensive efforts should be made to identify and alter modifiable risk factors. This study provides empirical support for recent recommendation of the National Cholesterol Education Program (21) that physicians provide comprehen-

JACC Vol. 12. No. 5

1279

Novcmbcr tY%l273-so

sive risk factor screening to people premature coronary disease.

with a family history of

Appendix

-.-

Items Used

to Elicit Self-Report of the Presence of Risk Factors

1. Have a. b. c. d.

you ever had any of the foilowing

your your your your

blood blood blood btood

8. Nor:? ‘J. !.nflvRcr RH. Spangler RD. Noct AH. Kimberling WI. Generic ~p~denrologic PW+IYof ezy-ooser ~rehemx heart disease. CirculaliL% IYao:hl 503-x.

RESPONSES:YES.NO, POSSIBLY

YES,

NOT

SURE

For each item, if “YES” or “POSSIBLY YES,” responiients were asked to give the level if known. indicate if the ievel was “NORMAL” or “HIGH,” indicate the approximate date and the source of information.

a. b. c. d.

on any occasion

in your

high blood pressure? _ high blood cholesterol? high blood triglycerides? diabetes or a high blood sugar?

RESPONSES:

YES,

NO,

life. had any of the

__

POSSlBLY

_ YES,

NOT

SURE

For each item, if “YES’ or “POSSIBLY YES,” the respondent was asked to name the approximate date, duration, source of

information, level if known and specific treatment. 3. Have

you ever conditions:

a. b. c. d.

taken

any

medicine

for any of the foilowing

9. Snowden CB. McNzmara PM Garrixm RJ. Fei&tb hi. Kannel WB. Epstein Ft!. Predicting sor?nary heart &i:ebxe in shlinp: a mt’lriv~;mte dxe$$mect. Am J Epidemiol I%2:!15:?IF??. 111 ten Kate LP. Boatman H. Da&r 5P. Mocat~ky AG FamdrJ qge@;rrmn of coronary heart d1sea.z and it\ re;alicr! lo krmwr: genettc rirk factors. Am J Cardlol IW?$&94>53. l I. Shea S. Ottman R. Gabrieli C, Srega ?,. Nichoh A. Family hlrrq a=. an mdependent ri?k factor for coronary &III ,%zxe J Am Co11 C+rdiol lY84:1:793-$nl. 12. R~:xmen AM. Nikkill EA. Aggrcp!ilm of coronary risk iactors in hmdler of men wrth fatal and non-fatal coronary hean diseaw. Zr Heail J 1979:4?~37340. 13. Gold~lem JL. Hazzard WR. Schn,.; HG. Bierman EL. Motul&y AG. Lipid level? in 500 rurvlvors rrf myocardial miarcliun. J Clin inwd lY73:52~I53.~3, 14. Perkms KA Family history of coronary heart disease: ir it dn indepndent risk fx:or? Am J Epidcmiol 19&;!24:182-93. 15. Central Patient Registry and Coordinating Center for the Lipid Rerearch Climcx Reference Manual for the Lipid Research Clinics Prevalence Chapel Hill: University of North Carolina. 1974. Sludy. Vol.

I and 2.

high blood pressure’! high blood cholesterol? high blood triglycerides? diabetes or a high blood sugar?

RESPONSES:

.I! diagnouls. Am J Cardiol lvl%.44:6&6.

levels checked:

pressure? cholesterol? triglycerides? sugar?

2. Have you ever, foltowing:

6. Rirsancn AM. FamrhaIoccurrence afc0ron-q

YES,

NO,

16. Repon of The Joint National Committee on Detection. Evatuation and Treatmen of High Blood Pressure. Arch lnlem Med lY84;144:!U45-57. 17. Rose G. Ellackbum H. Cardiovascular Survey Metho&. Orgmlration Monograpn Series No 56 i%.

POSSIBLY

YES,

NOT

SURE

For each item, if “YES” or “POSSIBLY YES,” respondents were asked IO give the name of the drug if known, I!K type of drug if hnown, the duration of use and the medical care source. “NOT SORE” responses were probed and respondents were offered the opportunity to elucidate any potential occasion on which the risk factor may have been present, even if transiently. AWARE = “YES” or “POSSIBLY YES” TV item 2 or a response 10 Item that categorized the number as “HIGH ,” or a response on item 3 that indicated a “YES” or “POSSIBLY YES” to drug treatment for the specific risk factors.

I

World Heabh

18. Manual of Laboratory Opetalioes. Lipid Rerearch Clinic% Progam: ‘Jot. I: Lipid and Lipoprotein Analysis. Belhetda. Maryland: Natiowal Institutes of He&h. NIH Publication 75-628. 1974. 19 Friedewald WT. Levy RI. Frcdickwn DS. Estimation of the concentratmn of low-denshy lipoprobzin cholesterol in plasma without use of the prepanrivc uhrdcemrifuge. Clin Clem 197?;!8:499-502. 20. The Lloid Research Clinics Powlation Sludies Data BoJk: Vol. I: The Preval&ce Study. Bethesda, #arylanJ: National Indllutes of He&h, NIH Publication No. 80-1527, 1980. ?I. Experl Panel. Nalional Cholesterol Educabon Program. Expert panel MI detection. evalualion and treatment of high blood cholesterol in adults. Arch Intern Med 19811;11:3f&9. 22. Conrensus Conference: 1984;?Sl: t 19f+200.

Trealmenl

of hypertriglyceridemia

JAMA

’

References I. Rose G.

23. M&n JJ. Hulley SE. Browner WS. Kuller H. Wenworth 0 Serum choleqterol. blood pressure and mortality: rmplic~lions from a cohort of 361.662 mep. Lance1 19&X:933-6.

heart disease. Br J Prev Sue Med

14. Kannel WB. McGee D. Gordon T. A general cardiovascular risk p-ofile: the Fmmmgham Study. Am J Cardiol 19%x38:4&51.

2. Slack J, Evans KA. The increased rirk of death from ischemic hedr! discabe in first degree relatives of I!1 men acid 96 wmcn \vith irchemic heart disease. J Med &net 19sh:3:?3%S7

25. Gudmund-son S.“hoFgtitsson G. ThorsteinssonT. Sigfusson N. Sigurdsyen G Ri& factor screr~.ing among51 firs1 degree relatives of paIrems with myocardial infarclian. Dan Med Bull lY~?~l0:25%62.

Familial patterns in Ixhemic 1964:18:75-%0.

1280

BECKER ET AL. CORONARY RISK FACTORS

IN FAMILIES

JACC Vol. 12. No. 5 November 198li:12734l

26. Hamsten A. de Faire U. Risk factors for coronary artery disease in families of young men with myocardial infarction. Am J Cardiol 1987;59:14-9.

cigar&e smoking. strong family history and medical therapy. Circulation 1981:63:546-51.

27. Goldstein JL, Schrott HG, Hazzard WR, Bierman EL, Molukcy AG. Hyperlipidemia in coronary heart disease !i. Genetic analysis of lipid levels in 176 families and delineation of a new inherited disorder. combined hyperlipidemia. J Clin Invest 1973;52:1544-68.

33. Hopkins PN. Williams RR, Hun1 SC. Magnified risks from cigarette smoking for coronary prone families in Utah. West J Med 1984:141:196202.

2%. Shucker B, Wittes JT. Cutler JA. el al. Change in physician perspective on cholesterol and heart disease. JAMA 1987:258:3521-6. 29. Shucker B. Bailey K. Heimbach JT, et al. Change in public perspective on cholesterol and heart disease, JAMA 1987;258:3327-35. 30. Centers for Disease Control. Cigarette smoking in the United States 1986. MMWR 1987:36:582-S. 31. Rissanen AM. Familial occurrence of coronary heart disease according to clinical manifestation. Acta Med Stand 1985;218:355-63. 32. Fuster V. Cheseboro JH, Frye RL. Elveback LR. Platelet survival and the development of coronary artery disease in the young adult: effects of

34. Brockway BS. Chemical validation of self-reported smoking rates. Behav Ther 1978;9:685-6. 35. Freedman DS, Srinivasan SR, Shear CL, Franklin FA, Webber LS, Berenson GS. The relationship of apolipoproteins A-l and B in children to parental myocardial infarction. N Engl J Med 1986;315:721-6. 36. Williams RR. Understanding genetic and environmental risk factors in susceptible persons. West J Med 1984:141:799-6& 37. Williams RR, Hasstedt SJ. Wilson DE, et al. Evidenct :!x: xi: oi:h familial hypercholestcrolemia can avoid early coronary death: an anarysis of 77 gclle carriers in four Utah pedigrees. JAMA 1986:255:2l%24.