SUMMARY

1. Both microalbuminuria and left ventricular hypertrophymay reflect target organ damage in essential hypertension. Bothare related to the prevailing level of blood pressure and bothare associated with an increase in morbidity and mortality.

2. The database of the Hypertension Diagnostic Service, amulticentre secondary referral clinic for patients with essentialhypertension, was analysed in order to clarify the level of association between microalbuminuria and left ventricularhypertrophy, which might explain the observed increase in morbidity and mortality in patients with microalbuminuria. Microalbuminuria was measured semiquantitatively by urinedip-stix. After the exclusion of patients with potential secondaryhypertension, renal disease and diabetes mellitus, patients withcomplete data for microalbuminuria, left ventricular mass(LVM) and 24 h blood pressure monitoring were selected.

3. Data were complete for 704 patients (47% male, age 51±12years) and 42% tested positive for microalbuminuria. Micro-albuminuria was positively related to 24 h systolic blood pres-sure and weight and was negatively related to age. Leftventricular mass was higher in patients with microalbuminuria(men, 265±69 g; women, 207±61 g) than in those without (men,250±64 g, P < 0.05; women, 185±50 g, P < 0.001). After correctionfor the effects of gender, body mass index and 24 h systolic bloodpressure, the presence of microalbuminuria was associated withan increase in LVM of 10 g (P < 0.05, 95% confidence interval,2–19 g).

4. Microalbuminuria as measured by urine dip-stix is associated with an increase in LVM in patients with essentialhypertension, independent of other known determinants.

Key words: ambulatory blood pressure monitoring, essentialhypertension, left ventricular structure, microalbuminuria, target organ damage.

INTRODUCTION

The presence of microalbuminuria in patients with essential hyper-tension has been related to cardiovascular morbidity and mortality,independent of other well-known risk factors.1–3 Albumin excretionrate has been shown to be related to the level of blood pressure inessential hypertension.4 The advent of rapid testing by reagent stripshas facilitated the screening of patients with essential hypertensionfor microalbuminuria.5

Left ventricular hypertrophy, however, has long since been estab-lished as a major independent marker of future cardiovascular mor-bidity and mortality in essential hypertension and depends mainlyon gender, body mass and 24 h blood pressure load.6–8

In order to clarify whether there is an association between micro-albuminuria and left ventricular mass, which may explain theobserved increased morbidity and mortality in these patients, weanalysed our database of the Hypertension Diagnostic Service, amulticentre secondary referral clinic for patients with essentialhypertension.

METHODS

Patients were referred to the Hypertension Diagnostic Service by generalpractitioners. The patients were seen at one of the hypertension diagnosticservice centres by an expert physician (vide infra for participating physi-cians and location). At the service centre, the physician took a standardizedhistory, performed a physical examination and measured height, weight,casual blood pressure and heart rate. Body mass index (BMI) was calcu-lated as BMI = weight/height2 and body surface area (BSA) as BSA (m2) =0.007184 3 weight (kg)0.425 3 height (cm)0.725. Blood was drawn for sodium,potassium, creatinine and glucose. A spot urine sample was tested for microal-buminuria with a Micral-Test II dip-stix (Boehringer Mannheim, Mannheim,Germany), which provides a reading of ‘negative’, $ 20, $ 50 and $ 100mg/L urinary albumin, as well as with urine microscopy. For the purposesof the present study microalbuminuria was defined as any reading except‘negative’. Patients underwent echocardiography with a Hewlett-Packard77020 A phased-array echo-Doppler system (Hewlett-Packard, Andover, MA,USA). Subsequent measurements were made on screen with electronic cal-

SHORT COMMUNICATION

LEFT VENTRICULAR MASS AND MICROALBUMINURIA: RELATION TO AMBULATORY BLOOD PRESSURE

Christoph D Gatzka, Chris M Reid, Alan Lux, Anthony M Dart and Garry L Jennings for the

Hypertension Diagnostic Service Investigators*

Hypertension Diagnostic Service, Alfred and Baker Medical Unit, Alfred Hospital and Baker MedicalResearch Institute, Commercial Rd, Prahran, Victoria, Australia

Correspondence: Dr C Gatzka, Alfred and Baker Medical Unit, Com-mercial Road, Prahran, Vic 3181, Australia. Email: <c.gatzka@alfred.org.au>

*Hypertension Diagnostic Service Centres and Investigators: Chris Reid,Alan Lux and Garry Jennings (Alfred Hospital, Prahran, Victoria, Australia),Howard Connor and Rob Ziffer (Gippsland Base Hospital, Sale, Victoria,Australia), John O’Sullivan and Roger Watts (Mersey Community Hospi-tal, La Trobe, Tasmania, Australia), Karen Duggan and Mark Penny(Bankstown-Lidcombe Hospital, Bankstown, New South Wales, Australia),Tony Johnson, David McIntyre and Adam Morton (Mater Private Hospital,South Brisbane, Queensland, Australia) and Shane Carney (John Hunter Hos-pital, Newcastle, New South Wales, Australia).

Presented at the 20th Annual Scientific Meeting of the High Blood Pres-sure Research Council of Australia, Melbourne, December 1998. This paperhas been peer reviewed.

Received 14 December 1998; revision 17 February 1999; accepted 24 February 1999.

Clinical and Experimental Pharmacology and Physiology (1999) 26, 514–516

LV mass, microalbuminuria and blood pressure 515

lipers by operators blind to the other clinical or biochemical measurements.8

Interventricular (IVS) and left ventricular posterior wall thickness (PW) weremeasured from leading edge to leading edge at end diastole. Left ventricu-lar internal diameter (LVID) was measured at end diastole. Left ventricularmass (LVM) was determined as LVM = 0.6 + 0.832 3 ((LVID + PW + IVS)3

– LVID3) g. Ambulatory blood pressure was recorded using a Spacelabs med-ical 90207 or 90217 monitor (SpaceLabs Medical Inc., Redmond, WA, USA).Monitors were set to measure every 30 min from 6 to 22 h and every 60 minfrom 22 to 6 h. Average 24 h blood pressure was calculated as two-third theaverage blood pressure from 6 to 22 h plus one-third the average blood pressure from 22 to 6 h, to compensate for the different frequency of measurements.

In the present study we excluded patients with a history of renal disease,with diabetes mellitus and with clinical suspicion of secondary hypertension(e.g. liquorice intake, renal bruits, cushingoid appearance or history) as wellas patients with a random plasma glucose of ³ 7.8 mmol/L, a plasma crea-tinine > 0.13 mmol/L for men or > 0.12 mmol/L for women, or significanthaematuria or leucocyturia. Only patients in whom all measurements werecomplete for height, weight, plasma creatinine, plasma glucose, 24 h bloodpressure monitoring and LVM were included in further data analysis.

Statistical analysis

Data are expressed as mean±SD except where indicated otherwise. All dataanalyses were done using SPSS for Windows software version 8.0.0 andincluded Student’s t-test, x2 test, logistic and linear regression analysis asappropriate. Statistical significance was taken at a level of P < 0.05.

RESULTS

Complete data were available for 704 patients (47% male, averageage 51±12 years). Sixty-one per cent of patients had not been receiv-ing antihypertensive medication at the time of study. Micro-albuminuria was present in 42% of patients. Univariate differencesbetween patients with or without microalbuminuria are shown inTable 1. A stepwise logistic regression with the presence or absenceof microalbuminuria as the dependent variable and as independentvariables age, gender, height, weight, BMI, BSA, creatinine, 24 hsystolic and diastolic blood pressure, yielded significant positiveeffects of 24 h systolic blood pressure (mmHg, B = 0.03, P < 0.0001)and weight (kg, B = 0.02, P < 0.001) and a negative effect of age(years, B = – 0.02, P < 0.001). No other variable entered had a significant predictive effect on the presence or absence of micro-albuminuria.

Left ventricular mass was higher in patients with micro-albuminuria than in those without (Table 1). Since gender, BMI andaverage systolic blood pressure over 24 h are all independent deter-

minants of LVM, we performed a multiple regression analysis withthese variables, to correct LVM for their effects. Applying this correction, the presence of microalbuminuria was associated withan increase in LVM of 10 g (P < 0.05, 95% confidence interval, 2–19 g).

DISCUSSION

The present study evaluates a population of patients referred to asecondary hypertension clinic by their general practitioners. Wefound a high prevalence of microalbuminuria (42%), which increaseswith the level of 24 h average systolic blood pressure and bodyweightand decreases with age. Due to the multicentric nature, our data setrepresents one of the largest available. It confirms previous smallerstudies that have demonstrated a relationship between albuminexcretion rate and the level of systolic blood pressure.9–11 Our find-ing of a positive relation to bodyweight is corroborated by one survey study,5 while in most other studies this has not been signi-ficant albeit going in the same direction.4,10 Most previous studieshave not been able to demonstrate a significant relationship betweenmicroalbuminuria and age, at least in a middle-aged population likethe one in this study.5 Therefore, the new result of a negative re-lationship between age and the prevalence of microalbuminuriashould be interpreted with caution and requires confirmation. In par-ticular, the study population is a selected subgroup of patientsreferred to a diagnostic centre and is therefore prone to selectionbias.

In addition, it has to be noted that albumin excretion was measured only semiquantitatively in this population. Even thoughthe test has a high sensitivity, the lower specificity might explainthe unusually high prevalence of microalbuminuria in our popu-lation.12,13 The presence of a considerable number of false positiveresults would be expected to have diluted our results; therefore, actu-arial differences between groups with and without albuminuria mightin fact be larger than reported here.

Furthermore, we found that LVM is increased in patients withmicroalbuminuria, independent of the level of 24 h systolic bloodpressure. This is consistent with previous studies.5,9,10 The observedeffect is considerable, equivalent in its potential to raise LVM to theeffect of an increase of about 11 mmHg in 24 h systolic blood pres-sure in our population. Therefore, this effect might quite well accountfor the observed increase in cardiovascular morbidity and mortal-ity in patients with microalbuminuria.

Table 1 Univariate differences between groups of patients with and without microalbuminuria. Data are expressed as mean6SD

No microalbuminuria (n = 407) Microalbuminuria (n = 297) PGender 42% male 58% male < 0.01Age (years) 52611 49612 < 0.001Height (cm) 16869 170610 < 0.01Weight (kg) 78614 84616 < 0.001BMI (kg/m2) 27.664.5 28.865.1 < 0.01BSA (m2) 1.960.2 2.060.2 < 0.001Plasma creatinine (mmol/L) 0.0860.01 (all) 0.0860.02 (all) NS (all)

0.0960.01 (men) 0.0960.01 (men) NS (men)0.0860.01 (women) 0.0860.01 (women) NS (women)

Average 24 h blood pressure (mmHg) 133/81613/9 137/83614/10 < 0.001/< 0.01Left ventricular mass (g) 250664 (men) 265669 (men) < 0.05 (men)

185650 (women) 207661 (women) < 0.001 (women)

BMI, body mass index; BSA, body surface area.

516 CD Gatzka et al.

In conclusion, in patients with essential hypertension micro-albuminuria is associated with an increase in left ventricular mass,independent of other known determinants. Further studies are neededto elucidate the underlying pathophysiological links betweenmicroalbuminuria and left ventricular structure.

ACKNOWLEDGEMENT

The Hypertension Diagnostic Service including its centralized dataanalysis and reporting facility is sponsored by Servier Laboratories(Aust) Pty Ltd.

REFERENCES

1. Damsgaard EM, Frøland A, Jørgensen OD, Mogensen CE. Micro-albuminuria as predictor of increased mortality in elderly people. BMJ1990; 300: 297–300.

2. Yudkin JS, Forrest RD, Jackson CA. Microalbuminuria as predictor ofvascular disease in non-diabetic subjects. Islington Diabetes Survey.Lancet 1988; 2: 530–3.

3. Kannel WB, Stampfer MJ, Castelli WP, Verter J. The prognostic significance of proteinuria: The Framingham Study. Am. Heart J. 1984;108: 1347–52.

4. Cerasola G, Cottone S, Mule G et al. Microalbuminuria, renal dys-function and cardiovascular complication in essential hypertension. J. Hypertens. 1996; 14: 915–20.

5. Agrawal B, Berger A, Wolf K, Luft FC. Microalbuminuria screeningby reagent strip predicts cardiovascular risk in hypertension. J. Hypertens. 1996; 14: 223–8.

6. Zanchetti A, Sleight P, Birkenhager WH. Evaluation of organ damagein hypertension. J. Hypertens. 1993; 11: 875–82.

7. Gatzka CD, Schmieder RE, Schobel HP, Klingbeil AU, Weihprecht H.Improved prediction of left ventricular mass from ambulatory bloodpressure monitoring using average tension-time-index. J. Hypertens.1993; 11 (Suppl. 5): 98–9.

8. Gatzka CD, Cameron JD, Kingwell BA, Dart AM. Relation betweencoronary artery disease, aortic stiffness and left ventricular structure ina population sample. Hypertension 1998; 32: 575–8.

9. Redon J, Baldo E, Lurbe E et al. Microalbuminuria, left ventricular massand ambulatory blood pressure in essential hypertension. Kidney Int.1996; 55 (Suppl.): S81–4.

10. Redon J, Liao Y, Lozano JV, Miralles A, Baldo E, Cooper RS. Factorsrelated to the presence of microalbuminuria in essential hypertension.Am. J. Hypertens. 1994; 7: 801–7.

11. Palatini P, Graniero GR, Mormino P et al. Prevalence and clinical correlates of microalbuminuria in stage I hypertension. Results fromthe Hypertension and Ambulatory Recording Venetia Study (HARVESTStudy). Am. J. Hypertens. 1996; 9: 334–41.

12. Marshall SM, Shearing PA, Alberti KG. Micral-test strips evaluated forscreening for albuminuria. Clin. Chem. 1992; 38: 588–91.

13. Tiu SC, Lee SS, Cheng MW. Comparison of six commercial techniquesin the measurement of microalbuminuria in diabetic patients. DiabetesCare 1993; 16: 616–20.