PAPERS Comparison of enalapril and nifedipine in treating non-insulin dependent diabetes associated with hypertension: one year analysis J C N Chan, C S Cockram, M G Nicholls, C K Cheung, R Swaminathan Departments of Clinical Pharmacology, Medicine, and Chemical Pathology, Abstract Objectives-To compare the efficacy, safety, and tolerance of enalapril and nifedipine in hypertensive patients with non-insulin dependent diabetes. Design-One year double blind follow up of patients randomly allocated to either enalapril or nifedipine with matching placebos for the alternative drug. Setting-Metabolic Investigation Unit, Hong Kong. Subjects-102 patients were randomised: 52 to nifedipine and 50 to enalapril. At baseline 44 patients had normoalbuminuria, 36 microalbuminuria, and 22 macroalbuminuria. Main outcome measures-Blood pressure, albuminuria, and parameters of renal function and glycaemic control. Results-In patients who completed one year's treatment the median dose required by the nifedipine group (n=49) was 60 mg/day; seven (14%) required additional diuretics. Of 41 patients given enalapril, 37 required the maximum dose (40 mg/day) and 27 (76%) required diuretics. At one year mean arterial blood pressures were similar in both groups. Albuminuria fell by 54% in the enalapril group and /in the nifedipine group (p=0 006). Fractional 11% albumin clearance ratio fell by 47% in the enalapril deterioration in kidney function in patients with insulin dependent diabetes mellitus.' Angiotensin II has potent vasoconstrictive effects on the efferent renal arteriole2; thus angiotensin converting enzyme inhibitors might reduce filtration pressure and preserve renal function more successfully than alternative antihypertensive drugs. Indeed, animal data suggest that this may be so.34 The antiproteinuric and renal protective effects of long term treatment with angiotensin converting enzyme inhibitors have recently been reported in insulin dependent diabetes.) Nevertheless, non-insulin dependent diabetes, which is the more prevalent form of diabetes, particularly among non-whites,6 contributes importantly to the patient population with end stage renal failure.7 To date, only a few clinical trials have studied the effects of angiotensin converting enzyme inhibitors in patients with non-insulin dependent diabetes and the results have been inconclusive.8-'6 We conducted a study to compare the long term effects of the angiotensin converting enzyme inhibitor enalapril with the calcium channel blocker nifedipine in the treatment of non-insulin dependent diabetes associated with hypertension. group and increased by 3% in the nifedipine group (p=0.009). Creatinine clearance fell similarly in both STUDY PROTOCOL The study was approved by the ethics committee of Patients and methods StUDY PROTOCoL TABL.E i-Study protocol for randornised tnial of enalapril anid nzifedipine in hypertensive patients with notzinsulin dependent diabetes 12 i6 24 36 52 4 8 -6 0 Weeks -4 Active treatment Placebo-8 24 Hour urine collection 1 111 1 1 1 1 1 11 Dsrugs (mg/day):. 10 20 40 Enalapril* or 8060 40 Nifedipine (modified release)* 2-5 Indapamide or 40 80 120 Frusemide the Chinese University of Hong Kong. Informed consent was obtained from all participants. Table I outlines the study design. Patients with non-insulin dependent diabetes treated by diet or oral hypoglycaemic drugs, or both, who were either hypertensive or receiving antihypertensive drugs and who were attending the outpatient diabetic clinic at our hospital were invited to participate in the study. All patients were Chinese and aged over 18. Patients were excluded if they were receiving insulin or had a history of non-diabetic renal disease, appreciable renal impairment (plasma creatinine concentration ¢ 200 Vmol/l), a plasma potassium concentration > 5 mmol/l, cardiac failure or any concurrent systemic disease, or were receiving treatment for any concomitant disorder. Previous antihypertensive drugs were discontinued and patients received placebo tablets to match enalapril 10 mg once daily and modified release nifedipine 20 mg twice daily. All patients were maintained on diets of thi otn eia eia previously prescribed as part o hi otn care. At the end of the six week run in period patients who satisfied the inclusion blood pressure criteria and none of the above exclusion criteria were admitted to the study. Inclusion blood pressure criteria were a mean supine systolic blood pressure of 150-220 mm Hg or a diastolic blood pressure (phase V) greater than 100 mm Hg, or both, at three readings during the run in *Patients received a matching placebo for the altemnative period. Patients were then assigned to receive either Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong J C N Chan, lecturer C S Cockram, reader M G Nicholls, professor C K Cheung, scientific officer R Swaminathan, professor Correspondence to: Dr J C N Chan, Department of Clinical Pharmacology, Prince of Wales Hospital, Shatin, NT, Hong Kong. BAI1992;305:981-5 BMJ VOLUME 305 groups but plasma creatinine concentration was increased by 20% in the enalapril group versus 8% in the nifedipine group (p= 000l). Conclusion-Patients taking enalapril often required diuretics to control blood pressure. Enalapril reduced proteinuria significantly more than nifedipine in the microalbuminuric and macroalbuminuric patients but increased plasma creatinine concentrations. Longer follow up is required to clarify the importance of enalapril's antiproteinuric effect. Introduction Early aggressive antihypertensive treatment has been associated with a reduction in the rate of 24 OCTOBER 1992 drug. 981 enalapril or modified release nifedipine with matching placebo tablets for the alternative drug according to a random allocation schedule. A schedule of 102 allocation numbers corresponding to similarly numbered drug supplies was provided for this purpose. Both the patients and the staff measuring blood pressure and biochemical indices were blinded to treatment. Based on the mean value of three measurements of 24 hour urinary albumin excretion during the run in period, 44 patients had normoalbuminuria (< 30 mg/day), 36 microalbuminuria (30-300 mg/day), and 22 macroalbuminuria (> 300 mg/day).', Drug doses were increased over 12 weeks to a maximum of enalapril 40 mg once daily or nifedipine 40 mg twice daily if supine systolic blood pressure remained above 140 mm Hg (table I). Indapamide 2 5 mg/day was added if the blood pressure remained high and, if necessary, this was replaced by frusemide (up to 120 mg/day) to achieve the target supine systolic blood pressure. Follow up visits were scheduled between 8 am and 10 am, at which time the blood pressure was measured, and venous blood was sampled for measurements of renal function, glycated haemoglobin, plasma fructosamine, and fasting plasma glucose concentration, and serum angiotensin converting enzyme activity. Twenty four hour urine samples were collected for estimation of albumin excretion. At the end of one year two 24 hour urine samples were collected within two weeks. week of collection for albumin concentration by immunoturbidimetry using a modification of a published technique.'8 Intra-assay and interassay coefficients of variation were 3.3% and 6-7% respectively within the range 1-2-80 mg/l. The lowest detection limit was 1-2 mg/l. Plasma and urine creatinine concentrations were measured by the Jaffe method on an Astra-8 Chemistry Analyser (Beckman Instrument, Palo Alto, California). Glycated haemoglobin concentration (HbA1) was measured by gel electrophoresis (Ciba Coming Diagnostics, Alto, California) with a reference range of 6-5-8 5%. Plasma glucose concentration was measured by a glucose oxidase method (Diagnostic Chemicals reagent kit) and plasma fructosamine, by published methods."'.Serum angiotensin converting enzyme activity was measured by a modified spectrophotometric method2" and the intra-assay and interassay coefficients of variation were less than 50/n. For presentation of skewed data that were analysed as logarithms, the mean was back transformed (antilogged) to give the geometric mean and the 95% confidence intervals obtained for the mean of the log data were also antilogged. MEASUREMENTS excretion. The mean (SD) or geometric mean (antilog of SD) values for all variables measured during the last two visits of the run in period were taken as baseline values. Doses of drugs administered are given as median. Fractional albumin clearance ratios were calculated as 24 hour urinary albumin concentrationxplasma creatinine concentration/plasma albumin concentrationx24 hour urinary creatinine concentration. Plasma creatinine concentration, creatinine clearance, fractional albumin clearance ratio, and urinary albumin excretion and serum angiotensin converting enzyme concentration were log transformed before analysis because of skewed distributions. The mean value from two measurements of urinary albumin excretion, fractional albumin clearance ratio, and creatinine clearance estimated at the end of one year were compared with baseline values and differences were compared between treatment groups. Mean (95% confidence interval) differences between the two treatment groups were examined by Student's two tailed t test. The antilog of a difference between two means was transformed into relative changes expressed as ratios compared with baseline 95* confidence intervals. Fisher's exact values with test was used to compare the rates of conversion from Blood pressure was measured by a single research nurse using a Hawksley random zero sphygmomanometer after the patient had rested for five minutes in the supine position and also after two minutes of standing. The mean of two readings for each body position was recorded. Mean arterial blood pressure was calculated as diastolic blood pressure plus one third of the difference between systolic and diastolic blood pressure and was shown as a mean of supine and standing values unless otherwise stated. All urine samples were stored at 4°C and analysed within one TABLE 11-Clinical characteristics of patienits and mean biochenmical data fronm the last two visits of the rnti in period. Values are nmeans (SD) uniless stated otherwise Nifedipine group (n-52) Enalapril group (n=50) No of men No ofwomen 21 31 20 30 Age (years) 56 1 (9 9) 60 1 (9 2) Duration of diabetes (years) 5 6 (4 6) Duration of hypertension (years) (°) with retinopathy No (°) with neuropathy Body mass index (kg/m2) 53 (47) No 5 5 (4 8) 56 (53) 22 (42) 17 (34) 14 (27) 15 (30) 24 8 (3 0) 25 2 (2 9) 174/92 (17/13) Supine blood pressure (mm Hg) 166/91 (16/9) Erect blood pressure (mm Hg) 167/94 (16/12) 171/94 (19/14) Mean arterial blood pressure (mm Hg) Glycatedhaemoglobin (%/) Plasma fructosamine (mmol/l) Fasting plasma glucose(mmol/l) Geometric mean (antilog SD) plasma creatinine (tLmol/l) Geometric mean (antilog SD) urinarw albumin excretion (mg/day) Geometric mean (antilog SD) creatinine clearance (ml/min) 117 (9) 9 8 (1-7) 2 3 (0Q23) 120 (12) 10-4 (1 7) 80 9 (1-4) 83-0 (1 3) 69-5 (6 8) 64 7 (6 6) 70*0 (1 6) 65-5 (1 5) 152 (9 6) 1 5 3 (7 4) 80 (21) 2 3 (0-23) 83 (25) Geometric mean (antilog SD) fractional albumin clearance ratio (x 10-c) T"ABLE ill-Severity of proteinuria at baseline anld the number of patienlts who1 comwpleted onue year of treatment in each categor STATrISTICS The study required 51 patients in each treatment group to have a 90% power at the 5% level (two tailed) to show that one drug was at least twice as effective as the other in reducing urinary albumin normoalbuminuria to abnormal albuminuria between the treatment groups. Repeated measures analysis of variance was used to test for the effects of treatment and its duration. Only data from patients who completed the one year study were included for these comparisons. Statistical analysis was performed with the packages ABstat (Anderson-Bell, Colorado, United States 1989) and SPSS Significance was taken U as p < 0 05 (two tailed). Results I'OPUIATION CHARACTERISTICS Of 123 patients recruited, 102 fulfilled all inclusion Category Normoalbuminuria Microalbuminuria Macroalbuminuria _____ 982 ____ ____ Urinarv albumin excretion(mg/day) WeekO0 Week 52 < 30 24 24 30-300 15 13 > 300 ____ ____ ____ ____ ____ WeekO0 Week 52 20 18 21 9 15 10 ____ criteria and were randomised to active treatment. Enalapril group Nifedipine group ____ ____ ____ 16 7 ____ ____ Table II shows the baseline clinical characteristics and mean biochemical data from the last two visits of the run in period. Table III shows the level of albuminuria a aeieadtenme fptet h opee a aeleadtenme fptet h opee one year of treatment in the subgroups. Forty nine of BMJ VOLUME 305 24 OCTOBER 1992 IAABLE 1V-Changes in urinarv albumlinl excretion, fracictwnal albumiinl clearance, endogenous creatinine clearance, plasmiia creatinine concentration, and arterial blood pressuire in1 pati'ents aftetr onze year of treatment zith eaniftedipime or enalapril Urinarv albumin cxcretio;'*: Enalapril Nifedipine p Value Fractional albumin clearance ratio*: Enalapril Nifedipine p Value Creatinine clearance*: Enalapril Nifedipine p Value Plasma creatinine*: Enalapril Nifedipillc pV'aluc Mlean arterial pressuret: Enalapril Nifedipine pValue Wholc group Normoalbuminuria Microalbuminuria Mlacroalbuminuria 0 46 (0Q32 to 0-67) 0(89 (0-66 to 1-2) 0006 0-88 (0 59 to 1 34) 0 81 (0 55 to 12) 0 7151 0Q27 (0 15 to 0-46) 0-85 (0 43 to 0 88) 0(013 0 29 (0 1 I to 0 77) 1 15 (0 85 to 1 54) 0 006 0 53 (0)37 to 0 77) 1 03 (0(75 to 1 43) 0 009 0-95 (0 58 to 1 72) 0 96 (0 63 to 1-45) 0 773 0 28 (0l18 to 0 44) 0-99 (0 43 to 2 3) 0 012 0 38 (0l16 to 088) 1 3 ( to 1 66) 0 005 0 85 (0n75 to 0 97) 0 85 (0 79 to 0 92) 0 989 0 86 (0-72 to 103) 0 85 (0( 77 to 0 93) 0 841 0 89 (0 71 to 11) 0 88 (0 75 to 1.04) 0 989 0-76 (0 52 to 1 12) 0 83 (0 7 to 098) 0 675 1l2 (114to 126) 1 08 (1-05 to 1 12) 0(001 1 15 (109to 123) 1-06 (1 01 to 1 12) 0 034 1l26(117to 136) 1 06 (1 01 to 1 1) 1 17 (0 98to 1 4) 1 19 (1l08 to 1 3) 0 904 -21 2 (-248 to - 16 3) -20 1(-24 1 to -18 4) 0 759 - 18 9 (-24.1 to - 13 8) -17 5 (-20 9 to -14 1) 0 633 <0 001 -21 6 (-292 to - 13 9) -251 (-31l1 to -19 1) 0 481 -22.0 (-34.3 to -9 7) -24-5 (-29S5 to -19 4) 0 685 *iN\ean relative change expressed as ratios compared with baseline values (95% confidence intervals). tMlean difference (95% confidence intervals). p Values are for comparisons bersveen changes in the tsvo treatment groups (Student's two tailed t test). TABLE \-Mean (95% confidence inter-val) chaniges in glycaemic inldices during 52 weeks' treatnmetnt with enalapril or niifedipinle Enalapril group Glycaemic index Glycatedhaemoglobin ('%Io) Plasma fructosamine (mmol/l) Nifedipine group 012 (-0 3 to 0(57) 0-00 (-0 07 to 0 09) 0 48 (-0 8 to 1 7) 063 (-02 to 47) 0( 14 (0-0 to 0 27) 0-42 (-0 6 to 1 5) Plasma glucose (mmoUl) p Value 0253 0 089 0 93 the 52 patients randomised to nifedipine completed one year of treatment compared with 41 of 50 randomised to enalapril. Of the nine patients taking enalapril who did not complete one year, three were withdrawn because of cough, three were withdrawn because of 1 )s ofr blood * * * pressure, one died - 1 ofr inadequate control myocardial infarction, one developed angina, and one defaulted. Of the three patients who received nifedipine, one had inadequate blood pressure control, onehad uberculouslymphaone had tuberculous lymphaonenitishad angina, and one denitis. 140 E E 120 ( oo E 11 \ 100 R l c, N T - 80 , 100 a > t- .: S Q t 10- *b4 E FIG 1-Bloodpressure and an.oeniduringcovrtn activity with treatmnentenym. entalapril () or nifedipine (0). activity wasplotted on logarithnaicscl. en (geometric means) are shown 0 j 0 with SD values drawn as error bars BMJ VOLUME 305 diuretic treatment (20 took indapamide and 1 frusemide). The median dose of nifedipine was 60 mg/ day and 14% of patients (7/49) required additional diuretics (four indapamide and three frusemide). Although at one year the reduction in and the achieved levels of mean arterial pressures were similar in both treatment groups (table IV), blood pressure was significantly higher overall in the enalapril group than Il~~~~~n ference the nifedipine group (p< 0 -001I for drug diffrne p<0 001 for time trend, and p=0 001 for time and drug interaction). The relative change in serum angiot c e a tersin patint eceiving (0 12 getri ainsrciigeaarl(-2(5 confidence interval 01 to 0 15)) than in those receiving nifedipine (0 71(0 63 to 0 81), p<0001; fig 1). Body mass index fell by 0 28 (-0 58 to 0 01) kg/m2 over 12 months in the enalapril group and 0 25 (-0 48 to -0.02) kg/mi in the nifedipine group (p=0 841). The haemoglobin concentration fell by 7 5 (-11-5 to -36) g/l in the enalapril group compared with a reduction of 2 4 (-0 5 to 0 6) g/l in the nifedipine group (p=0w035). At week 4, before the addition of diuretics, the plasma glucose concentration had fallen by 0 98 (-17 to -0 31) mmol/I in the patients receiving enalapril compared with 0 09 (-0 57 to 0 38) mmol/l r < in those receiving nifedipine (p=0033). However, glycaemic indices were similar in the two treatment groups at one year and overall changes were not significantly different between the groups (table V). wenalapril (95%/a ALBUMINURIA AND RENAL FUNCTION *t a) Angiotenain converting enzyme At one year, 90% (37/41) of the patients receiving enalapril required the maximum dose of enalapril (40 mg/day) and 76%/o (31/41) required the addition of I 300 E o T BLOOD PRESSURE, BODY WEIGHT, AND GLYCAEMIC CONTROL 24 OCTOBER 1992 Table IV summarises changes in mean arterial blood pressure, urinary albumin excretion, fractional - . A\ i 9 _ albumin clearance ratio, creatinine clearance, and plasma creatinine concentration in the two treatment _ 1 >-1 a F--groups. Treatment with enalapril reduced proteinuria, the fractional albumin clearance ratio significantly __ _ _ ~~~~~~~~~and more than treatment with nifedipine, in all patients __ _ _ ~~~~~andalso in the microalbuminuric and macroserum ,,-albuminuric groups separately. In normoalbuminuric patients, urinary albumin excretion remained less than 30m/a nalo h 8ptet ie nlpi 0m/a nalo h 8ptet ie nlpi whereas two of the 23 patients receiving nifedipine developed abnormal albuminuria (p=0-62 1). Figure 2 I I l l shows changes in urinary albumin excretion in 12 36 iniiulptesdrngoeya oframn. 48 52 nvdulptet drigoeya oframn. Creatinine clearance fell to a similar extent in both -Weeks E T .983 TABIT V -The cffrcts of additiOnal diuretic on renal fiuction and blood pressure in patients receiving No diurctics (n= 10) Mean (95",, confidence interval) arterial pressure (mmHg) Diuretics (n= 31) -201 (-289to-11 1) -206(-255to-158) Urinarv albumin excretion rate* Fractional albumiin clearance ratio (x 10")* Plasma creatinine (,umoUI)* Creatinine clearance (ml/min)* 0765 (0 38 to 1709) 0 42 (0(27 to 065) 0 77 (0-34 to 7 1) 1 08 (0 98 to 1 18) 0Q90 (0 67 to 1 2) 0-47 (0-31 to 0-71) 124 (118 more effectively than f3 adrenoceptor blocking drugsf prevalent form of diabetes, particularly among non- 0 84 (0-72 to 097) whites,, and accounts for a considerable proportion of 7 pat.ents with end stage renal disease, few trials have *MNean relative change expressed as ratios compared with baseline values (95% confidence intervals). groups of patients but plasma creatinine concentration was increased to a greater extent by treatment with enalapril (table IV). The effects of adding diuretics (indapamide or frusemide) were analysed (table VI). There was no difference in changes in urinary albumin excretion between patients who received enalapril alone and those requiring the addition of a diuretic. The rise in plasma creatinine concentration was significantly greater in patients receiving combined enalapril and diuretic compared with that in those treated with enalapril alone, Normoalbuminuria p=0.751 10 10000 Microalbuminuria p=0.013 E 1000 been reported comparing angiotensin converting enzyme inhibitors with alternative antihypertensive drugs in such patients.5'5 Some of the longer term studies suggest there is little difference between antihypertensive drugs with regard to their effects on urinary protein excretion"' whereas Ferder et al reported that enalapril had a superior antiproteinuric action to nifedipine in hypertensive patients."' We studied a homogeneous population of 102 Chinese patients with hypertension and non-insulin dependent diabetes to compare the effects of an angiotensin converting enzyme inhibitor, enalapril, with a calcium channel blocker, nifedipine, on blood pressure, renal function, and glycaemic control. The antihypertensive effects of treatment with enalapril and nifedipine were similar at one year, but the fall in blood pressure was slower with the angiotensin converting enzyme inhibitor, as has been reported previously in elderly white patients with essential hypertension.2 Furthermore, more patients receiving enalapril required a diuretic to achieve the goal systolic blood pressure (140 mm Hg in the supine position) than patients receiving nifedipine. These apparent differences in antihypertensive effectiveness might reflect the fact that full comparative dose response (drug and blood pressure) curves were not assessed, that the full antihypertensive action of enalapril may require more than 12 weeks to develop under the conditions of the study, and that racial factors might modify the antihypertensive efficacy of the drugs chosen-as is the case for 1 adrenoceptor blockers and diuretics in black, compared with white, patients.22 ALBUMINURIA 0 .j Considerable attention has been directed to the potential renal protective effects of angiotensin converting enzyme inhibitors in insulin dependent diabetes, and recent data suggest that these drugs may slow the rate of decline in glomerular filtration rate Although non-insulin dependent diabetes is the more 1 31)t tp=001. 100 Discussion Despite a higher overall level of blood pressure, patients receiving enalapril showed the greater 100 D' _ 10 N \ .: Macroalbuminuria 10000 \ p=0.006 ) fall in urinary albumin excretion. A significant antiproteinuric action of enalapril, compared with nifedipine, was observed in patients who initially exhibited macroalbuminuria or microalbuminuria. rates of conversion from normoalbuminuria to ~~~~~~~~~~The microalbuminuria were not different between the two study groups after one year. As diuretics were required by most patients receiving enalapril it is possible that the diuretics, rather than the angiotensin converting enzyme inhibitor, accounted for the fall in urinary albumin excretion, especially since indapamide has recently been shown \ 1000 FIG 2-Urinary albumin excretion, expressed on logarithmic scales, at weeks 0 and 52 in microalbuminuria, and macroalbuminuria. p Values are shown for differences in changes in urinaiy albumin excretion between patients treated with enalapril (a) and niffedipine (0) 984 o < - 100 patients to have an hypertensive ~~~~~~in patients.23 antiproteinuric action non-insulin dependent diabetic Further analysis of our data, however, 2 showed similar antiproteinuric effects of enalapril whether or not a diuretic was added. Our data, \\\ therefore, agree with the findings of Ferder et al, who <reported that fixed dose enalapril (40 mg/day) for one year in 18 hypertensive non-insulin dependent diabetic patients reduced urine excretion significantly whereas withfiedoenfdpn(4mgdyfaldtatruiay normoalbuminuria,fiedoenfdpn(4mgdyfaldtatruiay protein output in 12 patients.'5 -Endogenous creatinine clearance fell to a similar extent in the two treatment groups. By contrast, 10 0 52 0 Weeks 52 plasma creatinine rose to a greater extent in patients treated with enalapril. This discrepancy may be largely due to the greater coefficient of variation inherent in BMJ VOLUME 305 24 OCTOBER 1992 the measurement of endogenous creatinine clearance, as it is dependent not only on determination of plasma creatinine concentrations but also on measurements of urinary creatinine concentration and urine volume.24 Furthe'r analysis of our data showed that pla'sma creatinine concentrations rose more in patients taking enalapril who received concomitant diuretics than in patients taking only enalapril. The higher plasma creatinine concentration in the enalapril group might therefore reflect the greater number of patients requiring a diuretic. In severe cardiac failure the combination of a diuretic and angiotensin converting enzyme inhibitor often induces a rise in plasma creatinine concentration 25-2 which, on prolonged treatment, usually falls again.25 Bjorck et al also reported an early fall in glomerular filtration rate in patients with insulin dependent diabetes after the introduction of enalapril but stabilisation occurred after six months. Whether such a biphasic pattern in glomerular filtration rate occurs in our patients, remains to be seen. EFFECT ON GLYCAEMIC CONTROL AND HAEMOGLOBIN Hypoglycaemia consequent on initiation of treatment with angiotensin converting enzyme inhibitors has been reported in both insulin dependent and noninsulin dependent diabetic patients.28230 Furthermore, the angiotensin converting enzyme inhibitors have been shown to improve insulin sensitivity,"' perhaps as a result of the accumulation of bradykinin, which has an insulin-like action.32 Early in our study, the mean fasting plasma glucose concentration fell significantly in patients receiving enalapril, but not in those receiving nifedipine. Subsequent indices of glycaemic control were similar in the two groups, presumably because the addition of diuretics, which are known to worsen glucose tolerance,3" countered any beneficial effect of the angiotensin converting enzyme inhibitor. We observed a small but significant fall in haemoglobin concentration during treatment with enalapril. Angiotensin converting enzyme inhibitors have also been noted to reduce haemoglobin concentration in patients with insulin dependent diabetes,5 congestive heart failure, chronic renal failure, and kidney transplants."3 This effect may be partly due to a fall in angiotensin II concentration, which is known to stimulate erythropoietin under certain circumstances.33" Chinese hypertensive patients with non-insulin dependent diabetes treated with enalapril often required additional diuretic therapy to control blood pressure. Despite similar blood pressure responses in the two groups at the end of one year, enalapril reduced proteinuria more effectively than nifedipine. The reduction in proteinuria occurred in patients with both microalbuminuria and macroalbuminuria. Plasma creatinine concentration increased in both groups but to a greater extent in the enalapril group, particularly in those requiring diuretics. Withdrawal from the study was more common with enalapril, mostly because of cough or inadequate control of blood pressure. A longer period of treatment is required to establish the significance of the reduction of proteinuria with enalapril and to determine whether this will be translated into a beneficial effect on glomerular filtration rate. Our study will continue for a further two years. Until more information is available it is premature, in our view, to suggest that any one group of antihypertensive drugs is superior to any other in the treatment of non-insulin dependent diabetes associated with hypertension. The study was sponsored by Merck, Sharpe, and Dohme. We thank Ms Margaret Cheung and Ms Angela Wong, our two research nurses for their dedication; Dr Denis Leung from the Centre for Clinical Trials and Epidemiological BMJ VOLUME View publication stats 305 24 OCTOBER 1992 Research, Chinese University of Hong Kong, for help with the statistical analysis; and Dr J A J H Critchley, reader of the department of clinical pharmacology, Chinese University of Hong Kong, for critical appraisal of the manuscript. I Parsing HH, Andersen AR, Smidt UM, Svendsen PAa. Early aggressive antihypertensive treatment reduces rate of decline in kidney function in diabetic nephropathv. Lanscet 1983;li: 1 175-9. 2 Hella J, Horacek V. Angiotensin II: preferential efferent constriction. Renal Phvsiol 1986;9:357-65. 3 Jackson B, Johnston CI. The contribution of systemic hypertension to progression of chronic renal failure in the rat remnant kidney: effect of treatment with an ACE inhibitors or a calcium inhibitor. 7 Hperteoso 1 988;6:495-50 1. 4 vMyers BD, Meyer TW. Angiotensin-converting enzyme inhibitors in the prevention of experimental diabetic glomerulopathv. Am J Kidnevs Dis 1 989;13:20-4. 5 Bj6irck S, Mulec H, Johnsen SA, N6rden G, Aurell M. Renal protective effect of enalapril in diabetic nephropathy. BMJ 1992;304:339-43. 6 Zimmet P. Type 2 (non-insulin-dependent) diabetes-an epidemiological oversiew. Diabetologia 1982;22:399-41 1. 7 Mogensen CE. Angiotensin converting enzyme inhibitors and diabetic nephropathy. BAIJ 1992;304:327-8. 8 Melbourne Diabetic Nephropathy Study Group. Comparison between perindopril and nifedipine in hypertensive and normotensive diabetic patients with microalbuminuria. BAIM 1991;302:210-6. 9 Baba T, Murabavashi S, Takebe K. Comparison of the renal effects of ACE inhibitors and calcium antagonist in hypertensive type 2 (non-insulindependent) diabetic patients with microalbuminuria: a randomized controlled trial. Diabbetologia 1989;32:40-4. 10 Valvo E, Bedogna V, Casagrande P, Antiga L, Zamboni M, Bommartini F, et al. Captopril in patients with type II diabetes and renal insufficiency: systemic and renal hemodynamic alterations. AmJ1fed 1988;85:344-8. 11 Stornello M, Valvo EV, Vasques E, Leone S, Scapellato L. Systemic and renal effects of chronic angiotensin converting enzyme inhibition with captopril in hypertensive diabetic patients. _flvipts)tes 1989;7(suppl): PS65-7. 12 Ueda Y, Aoi W, Yamachika S, Shikaya T. Beneficial effects of angiotensinconverting enzyme inhibitor on renal function and glucose homeostasis in diabetics with hypertension. Nephron 1990;55(suppl 1):85-9. 13 Matthews DM, Wathen CG, Bell D, Collier A, Roulston JE, Clarke BF, et al. The use of captopril and captopril plus frusemide as antihypertensive agents in non-insulin dependent diabetes. _Hton Hsperte)ts 1987;1: 19-23. 14 Stomello M, Valvo EV, Scapellato L. Persistent albuminuria in normotensive non-insulin-dependent (type II) diabetic patients: comparative effects of angiotensin-converting enzyme inhibitors and fl-adrenoceptor blockers. Clint Sci 1992;82:19-23. 15 Ferrier C, Ferrari P, Weidmann P, Keller U, Beretta-Piccoli C, Riesen WF. Anihypertensise therapy with calcium antagonist, verapamil and/or ACE inhibitor enalapril in NIDDM patients. Diabetes Care 1991;14:91 1-4. 16 Ferder L, Daccordi H, Martello M, Panzalis M, Inserra F. Angiotensin converting enzyme inhibitors versus calcium antagonists in the treatment of diabetic hypertensive patients. Hyperte?isio?i 1 992;19(suppl II):237-42. 17 Rowe DJF, Dawnay A, Watts GF. Microalbuminuria in diabetes mellitus: review and recommendations for the measurement of albumin in urine. Anini Clii, Biochemi 1990;27:297-312. 18 Cheung CK, Swaminathan R. Automated immunoturbidimetric method for the determination of retinol binding protein, prealbumin and transferrin in urine. CliGu Biochemii 1989;22:425-7. 19 MacDonald D, Pang CP, Cockram CS, Swvaminathan R. Fructosamine measurements in serum and plasma. Clin, Chewt Acta 1987;168:247-52. 20 McGuire GA, Price CP. A continuous monitoring spectophotometric method for the measurement of ACE in human serum. Ami Cli?i Biocheni 1985;22: 204-10. 21 Gilchrist NL, Nicholls MG, Ewer rC, Livesey JH, Sainsbury R. A comparison of long acting nifedipine and enalapril in elderly hypertensives: arandomised, single-blind, cross-over study._ Hun, Hipertens 1988;2:33-9. 22 Veterans Administration Cooperative Study group on Antihypertensive Agents. Comparison of propranolol and hydrochlorthiazide for the initial treatment of hypertension. I. Results of short-term titration with emphasis on racial differences in response.7AMA 1982;248:1996-2003. 23 Gambardella S, Frontoni S, Lala A, Felici MG, Spallone V, Scoppala A, et al. Regression of microalbuminuria in type II diabetic hypertensive patients after long-term indapamide treatment. Aiii HeartJf 1991;122:1232-8. 24 Payne RB. Creatinine clearance: a redundant clinical investigation. A,ill Cliii Biochenii 1986;23:243-50. 25 Cleland JGF, Dargie HJ, Hodsman GP, Ball SG, Robertson JIS, Morton JJ, et al. Captopril in heart failure. A double blind controlled trial. Br Heart J 1 984;52:530-5. 26 Mujais SK, Fouad FM, Textor SC, Tarazi RC, Bravo EL, Hart N, et al. Transient renal dysfunction during initial inhibition of converting enzyme in congestive heart failure. Br HeartJ 1984;52:63-7 1. 27 Nicholls MG. Overview: angiotensin, angiotensin converting enzyme inhibition, and the kidney-congestive heart failure. Kidney lilt 1987;31 (suppl 20):S200-2. 28 Ferriere M, Lachkar H, Richard J-L, Bringer J, Orsetti A, Mirouze J. Captopril and insulin sensitivity.'Ani Intern Med 1985;102:134-5. 29 Arauz-Pacheco C, Ramirez LC, Rios J, Raskin P. Hypoglycemia induced by anglotensin-converting enzyme inhibitors in patients with non-insulindependent diabetes receiving sulfonylurea therapy. Anm J Med 1990;89: 811-3. 30 Chan JCN, Cockram CS. Drug-induced disturbances of carbohydrate metabolism. Adverse Drug Reactions Toxicological Review 1991;10: 1-29. 31 Pollare T, Lithell H, Berne C. A comparison of the effects of hydrochlorothiazide and captopril on glucose and lipid metabolism in patients with hypertension. NEnzglj7Med 1989;321:868-73. 32 Jauch KW, Hartl W, Guenther B, Wicklmayr M, Rett K, Dietze G. Captopril enhances insulin responsiveness of forearm muscle tissue in non-insulin- dependent diabetes mellitus. EurJCliiz Ittvest 1987;17:448-54. 33 Lai KN, Lui SF. Renin and erythropoietin. In: Robertson JIS, Nicholls MG, eds. Renini-angiotensin systemn textbook. London: Gower (in press). 34 Kamper A-L, Nielsen OJ. Effect of enalapril on hemoglobin and serum erythropoietin in chronic nephropathy. Scand 7 Cliii Lab Invest l990;50: 611-8. (Accepted 2O August 1992) 985