Effect of Drug and Diet Treatment of Mild
Hypertension on Diastolic Blood Pressure
Herbert G. Langford, Barry R. Davis, Donald Blaufox, Albert Oberman, Sylvia Wassertheil-Smoller, Morton Hawkins,
and Neal Zimbaldi for the TAIM Research Group
The Trial of Antihypertensive Interventions and Management is a multicenter randomized trial designed to examine the diastolic blood pressure response of various combinations of pharmacological and dietary interventions in the treatment of mild hypertension (diastolic blood pressure 90-100 mm Hg). Eight hundred and seventy-eight participants at 110-160% of ideal weight were randomly allocated to nine drug/diet treatment groups receiving either a placebo, chlorthalidone (25 mg), or atenolol (50 mg), combined with a usual, a weight loss, or a low sodium/high potassium diet The primary outcome was diastolic blood pressure change from baseline to 6 months. Seven hundred and eighty-seven participants had follow-up data. The mean baseline diastolic blood pressure was 93.8 mm Hg; 55.9% of the participants were male, and the weight loss diet group lost an average of 4.7 kg. Multiple comparisons were accounted for in the analysis. A significantly greater lowering of diastolic blood pressure (12.4 mm Hg) was achieved in the atenolol group compared with either the low sodium/high potassium diet group (7.9 mm Hg,/?=0.001) or weight loss group (8.8 mm Hg,/>=0.006). Adding weight loss to chlorthalidone significantly enhanced blood pressure lowering (15.1 mm Hg) when compared with the diuretic alone (10.8 mm Hg,/?=0.002), but adding a low sodium/high potassium diet (12.2 mm Hg, /?=0.029) did not In the short-term treatment of mild hypertension where diastolic blood pressure is the sole consideration, drugs outperform diet, and weight loss is beneficial, especially with diuretics. (Hypertension 1991;17:210-217)
M
ild hypertension is a significant risk factorfor cardiovascular mortality and morbidi-ty.1 Treatment of mild hypertension sig-nificantly reduces this mortality and morbidity,2-5 but the cost/benefit ratio has been questioned.6 How best to treat mild hypertension is undecided and remains a controversial point.7'8 The symptomatic and bio-chemical side effects, need for long-term use and cost of pharmacological treatment for all hypertensive
From the Department of Medicine (H.G.L.), University of Mississippi Medical Center, Jackson, Miss.; The Coordinating Center for Clinical Trials, University of Texas School of Public Health, Houston, Tex. (B.R.D., C.M.H., N.Z.); The Departments of Nuclear Medicine (M.D.B.) and Epidemiology and Social Medicine (S.W-S.), Albert Einstein College of Medicine, Bronx, N.Y.; and the Division of General and Preventive Medicine (A.O.), University of Alabama at Birmingham, Birmingham, Ala. Research on which this publication is based was supported by grants HL-30171, HL-24369, HL-30163, and HL-41445 from the National Heart, Lung, and Blood Institute, National Institutes of Health, US Department of Health and Human Services.
Address for correspondence: Barry R. Davis, MD, PhD, The University of Texas, School of Public Health, 1200 Herman Pressler Street, Suite 801, Houston, TX 77030.
Received June 18,1990; accepted in revised form September 20, 1990.
persons have led to suggestions that dietary therapy may be a preferable alternative therapy for the person with mild blood pressure elevation.9 Also, there is continuing controversy as to the most desir-able medication for initiating treatment in the mild hypertensive individual.10
The Trial of Antihypertensive Interventions and Management (TAIM) is a randomized, controlled trial among three clinical centers involving persons between 21 and 65 years of age with diastolic blood pressure between 90 and 100 mm Hg without other substantial disease. TAIM was undertaken to assess the relative value of the most commonly used approaches to drug therapy (thiazides and /3-blockers) and diet therapy (weight loss and a restricted sodium, increased potas-sium intake) applied singly and in combination. This report, based on the first 6 months of intervention, describe changes in blood pressure for the various combinations of therapy. The trial is continuing, and long-term results will be available in a few years.
Methods
Participants were recruited from January 24, 1985, to June 30, 1987, from the communities of three
clinics, which included the Albert Einstein College of Medicine, Bronx, N.Y.; University of Alabama School of Medicine, Birmingham, Ala.; and the Uni-versity of Mississippi School of Medicine, Jackson, Miss. There was a nutrition and behavioral core located at Albert Einstein College of Medicine and a coordinating center located at the University of Texas School of Public Health, Houston, Tex. The protocol was reviewed and approved by the Institu-tional Review Boards at all the centers. The ratio-nale, experimental design, and selected baseline re-sults of TAIM have been reported elsewhere.11 All the participants were male and female volunteers aged 21-65 years.
Eligibility criteria for TAIM at a preliminary screen were 1) a diastolic blood pressure of 100 mm Hg or less for participants currently taking antihyper-tensive medication or 2) a diastolic blood pressure between 90 and 104 mm Hg for those on no treat-ment. In addition, participants had to be between 110% and 160% of their ideal weight12 by recall. At the second screen or first clinic visit, all potentially eligible participants had to be on no antihypertensive medication and had to have a diastolic blood pres-sure between 90 and 100 mm Hg and again be between 110% and 160% of their ideal weight by clinic measurement. To become eligible for TAIM by this second screen, those patients on prior antihyper-tensive therapy had their medication reduced, then discontinued over a time period of up to 8 weeks. Determination of blood pressure eligibility was made 2 or more weeks after medication was discontinued. Participants were excluded who had a history or other evidence of myocardial infarction, stroke or bronchial asthma, a creatinine level of 180 /xmol/1 or greater, diabetes requiring insulin therapy, allergy to thiazides or /3-blockers, actual or contemplated preg-nancy, or likelihood of difficulty in complying with the interventions.
There were 10,148 participants screened for TAIM. Of these, 1,949 were seen at the first clinic visit and 881 went on to a second clinic visit. The major reasons for not proceeding to the first or second visit were ineligibility due to blood pressure, age, or weight. At the second visit, 878 people were randomly allocated into the TAIM study.
Randomization
All eligible persons signed an informed consent statement. On completion of the baseline evaluation, designated clinical center personnel contacted the coordinating center, which randomly assigned each participant to one of nine possible drug/diet treatment groups. Random allocation was stratified within clin-ical center and by race. Medications were identified by preassigned code to maintain the double-blind design.
Intervention
Within 1-2 weeks of randomization, most TAIM participants were started on their interventions. The study was a 3x3 factorial design consisting of three
drug and three diet groups. Participants received either a diuretic (chlorthalidone 25 mg), a £-blocker (atenolol 50 mg), or a placebo. All pills were de-signed to appear identical and were administered in double-blind fashion. Two types of active dietary intervention were used: weight loss and low sodium/ high potassium diets. Also, participants could be assigned to a usual (no change) diet. The goal for the weight loss group was a reduction of 10% of baseline weight or 4.54 kg, whichever was greater. Sodium and potassium goals were individualized by weight and ranged from 52 to 100 mmol/day for sodium (average 87 mmol/day), and from 62 to 115 mmol/day for potassium (average 103 mmol/day). To attain these dietary goals, nutritional counseling oriented to be-havioral change was provided in group sessions held weekly for 10 weeks. Thereafter, individual sessions were held every 6-12 weeks. This approach was a modification of the one used in the Dietary Interven-tion Study of Hypertension.1314
Participants randomly assigned to TAIM protocol groups had clinical visits every month for the first 6 months and then quarterly. Long-term follow-up data beyond 6 months are still being acquired. Di-etary intervention was begun for the low sodium/high potassium and weight reduction groups within 3 weeks of the baseline visit. Participants who failed to achieve adequate blood pressure control were stepped up to additional therapy at 6 months or sooner if emergency failure criteria were met. The additional step-up therapy was administered in a double-blind fashion. Either 25 mg chlorthalidone or 50 mg atenolol was given to placebo failures (allocat-ed in a 1:1 random fashion); combin(allocat-ed 25 mg chlorthalidone-50 mg atenolol therapy was given to the chlorthalidone or atenolol failures. Medication was increased during the first 6 months if diastolic blood pressure was equal to or greater than 100 mm Hg for three visits at 2-week intervals, greater than 105 mm Hg at two visits a week apart, or equal to or greater than 115 mm Hg at any visit. If the additional step-up therapy did not adequately control the dia-stolic blood pressure, open-label therapy (known antihypertensive medication) was used.
Trial End Points
The primary outcome of the study was the 6-month change in diastolic blood pressure. Fourteen pairwise comparisons of this change among the nine treat-ment groups were formulated by the TAIM investi-gators and are presented in detail elsewhere.11 Es-sentially, diet therapy was compared with low dose drug therapy and combined diet and drug treatment was compared with diet alone or drug alone. Other outcomes compared among the treatment groups included change in biochemical variables, changes in Framingham cardiovascular risk scores,115 and changes in quality of life measures.16
TABLE 1. Baseline Characteristics by Treatment Group
Characteristic Sample size Age (mean) (yr) Sex (% male) Race (% white) Prior drug treatment (%) Smokers (%)
Alcohol use (% drinks/wk a l ) Diastolic BP (mean) (mm Hg) Systolic BP (mean) (mm Hg) Pulse rate (mean) (beats/min) Weight (mean) (kg) Ideal weight (%) Subscapular sldnfold (mm) Urinary Na+ (mmol/day) Urinary K+ (mmol/day) P 90 47.4 40.0 67.8 66.7 15.6 36.7 93.7 144.5 76.4 85.6 131.2 31.9 120 55.1 Usual C 87 48.2 62.1 66.7 70.1 12.6 41.4 93.4 141.9 78.5 89.6 133.2 30.8 132 55.4 A 87 47.7 64.4 69.0 64.4 14.9 46.0 93.7 142.8 77.6 88.6 131.7 30.4 137 54.9 P 90 49.1 58.9 67.8 65.6 12.2 35.6 94.0 143.2 79.1 89.8 132.9 31.5 137 59.8 Diet group Weight loss C 87 47.4 47.1 66.7 62.1 16.1 31.9 93.7 141.0 78.1 86.6 133.2 33.6 130 55.2 A 88 47.5 54.5 69.3 64.8 11.4 37.5 93.7 142.8 77.9 86.0 129.8 31.7 136 57.2 Low P 79 50.5 55.7 70.9 64.6 17.7 34.2 94.4 144.9 78.4 87.4 131.2 29.9 130 55.0 Na/high C 89 48.9 55.1 66.3 64.0 13.5 32.6 93.7 143.1 79.2 86.1 130.4 31.7 136 60.5 K A 90 51.0 65.6 64.4 64.4 15.6 44.4 94.0 146.3 76.2 90.2 133.6 31.0 141 59.1 Total 787 48.6 55.9 67.6 65.2 14.4 38.3 93.8 143.4 77.9 87.8 131.9 31.3 133 56.9 P, placebo; C, chlorthalidone; A, atenolol; BP, blood pressure.
Statistical Methods
The sample size was calculated to detect (with an a of 0.05 and a /3 of 0.20) differences of at least 4-5 mm Hg of 6-month diastolic blood pressure changes for the 14 selected pairwise comparisons. The sample size calculations are described elsewhere.11
Comparisons of baseline variables for all nine intervention groups were performed using x2 tests for categorical variables and one-way analysis of variance for continuous variables. The mean differences for the 14 selected questions were compared using the Bonferroni multiple comparison procedures.17 All analyses were by intention to treat. Because of mul-tiple comparisons, a p< 0.0036 (0.05/14) was consid-ered statistically significant.
Multiple linear regression analyses were also per-formed to account for imbalances among the treat-ment groups in baseline factors and to explore pos-sible interaction effects.17
Measurement
Blood pressures were measured, using American Heart Association guidelines, by staff who were trained and certified according to the Hypertension Detection and Follow-up Program criteria.18 The random zero mercury sphygmomanometer was used to minimize observer bias for cutpoints. Blood pres-sure observers were blinded to participants' diet assignments in one out of the three clinical centers. All observers were blinded to drug therapy as previ-ously explained. The reasons for incomplete diet blinding were due to logistical and budgetary consid-erations. Blood pressures were measured after the participant had been seated quietly for at least 5 minutes. The average of two readings of the fifth-phase diastolic blood pressure was used in all analy-ses. Participants normally took their medication in
the morning and made their clinic visits at various times during the day.
Dietary intake was assessed with 3-day food rec-ords at baseline and 6 months. Participants were instructed in keeping an accurate record and com-pleted records were reviewed in detail by nutrition-ists. Additional assessment of sodium and potassium intake was provided by data from 24-hour urine samples collected at baseline and at 6 months.
Participants underwent a standard battery of tests at baseline and 6 months including the measurement of glucose, uric acid, creatinine, cholesterol, high density lipoprotein (HDL) cholesterol, triglycerides, and electrolytes. These analyses were performed at a central laboratory. In addition, at baseline and 6 months serum potassium levels were monitored to evaluate the need for potassium supplements.
Results
At baseline, all nine groups were compared for almost all the major variables measured (Table 1). There were no statistically significant differences between groups with regard to these variables except gender. There were more men than women in each group except the usual diet/placebo and weight loss diet/chlorthalidone groups.
Participation Status
Participation was at a high level throughout the first 6 months of the trial for clinic attendance, completion of food records, procedures completed, and medication taken. The end point blood pressure was obtained after 6 months of intervention (range, 5-8 months; mean, 6.3 months). The mean times from the baseline visit and randomization to the end point blood pressure determination were 7.5 and 6.8 months, respectively. Twenty-four participants had
TABLE 2. Change in Selected Variables From Baseline to 6 Months by Treatment Group Variable Change* in weight (kg) Mean Sample size Total SD=4.67
Change* in urinary sodium (mmol/day) Mean
Sample size Total SD=72.68
Change* in urinary potassium (mmol/day) Mean
Sample size Total SD=30.36
Change* in systolic blood pressure (mm Hg) Mean Sample size Total SD = 15.95 P -0.71 90 9.06 82 0.78 82 -10.34 90 Usual C -1.52 87 12.05 72 12.08 72 -17.41 87 A -0.53 87 -4.56 76 -0.65 76 -15.06 87 P -4.41 89 -5.86 80 -1.98 80 -11.49 90 Diet group Weight loss C -6.85 86 11.97 79 10.15 79 -21.72 87 A -3.02 88 -2.08 78 -1.70 78 -18.11 88 Low P -0.33 79 -34.94 65 12.83 65 -8.66 79 Na/high C -1.09 88 -24.82 69 11.58 69 -19.51 89 K A 0.44 90 -23.59 80 8.77 80 -18.29 90
P, placebo; C, chlorthalidone; A, atenolol.
'Change equals 6-month value minus baseline value.
withdrawn or been lost to follow-up by the first month visit and 67 participants had their blood pressures measured outside the 5-8-month window. Therefore, 787 participants had available blood pres-sure data. The main reason for withdrawal or lost to follow-up was patient refusal (about 90%).
Compliance with drug therapy was assessed by the use of monthly pill counts. By the end of the 6 months, 92% of the participants with pill counts were taking more than 80% of their assigned medication. How-ever, all the 787 participants did not remain on their initial drug assignment over the 6-month study. Among the placebo-assigned participants, 20.0% of those in the usual diet group, 10.0% of those in the weight loss group and 16.5% of those in the low sodium/high potassium group were on additional ther-apy (double-blind step-up or open-label medication) by the end of 6 months. For those participants as-signed to chlorthalidone and atenolol, 2.7% and 3.0%, respectively, were on additional therapy and 3.4% and 2.6%, respectively, were on no active medication. Attainment of Dietary Goals
Table 2 presents the change in weight and 24-hour urinary sodium and potassium for each of the nine subgroups. There was an average loss of 4.7 kg in the total weight loss group with considerable variation among the three drug subgroups. The low sodium/ high potassium group had an average decrease in urinary sodium of 27.4 mmol/day and an average increase of urinary potassium of 10.9 mmol/day. The placebo subgroup within the low sodium/high potas-sium diet group had the greatest reduction in urinary sodium and the greatest increase in urinary
potas-sium, compared with the chlorthalidone and atenolol subgroups within that diet.
Changes in Diastolic Blood Pressure
Table 3 presents the change in diastolic blood pressure from baseline to 6 months for the six main groups (in the margin) and all nine subgroups (cells).
TABLE 3. Change in Diastolic Blood Pressure From Baseline to 6 Months by Treatment Group
Drug group Placebo Chlorthalidone Atenolol Total Usual -7.96 90 8.63 71.1 -10.78 87 7.08 86.2 -12.43 87 6.99 93.1 -10.36 264 7.81 83.3 Diet group Weight loss -8.78 90 10.97 73.1 -15.06 87 8.54 93.1 -14.81 88 9.14 92.0 -12.84 265 10.02 86.0 LowNa/ highK -7.91 79 10.12 64.1 -12.18 89 7.93 86.5 -12.76 90 7.97 87.8 -11.07 258 8.89 80.2 Total -8.23 259 9.92 69.9 -12.67 263 8.04 88.1 -13.33 265 8.13 90.9 -11.43 787 9.01 83.2 Values listed in each cell are mean of diastolic blood pressure change (mm Hg), sample size, standard deviation, and percent of participants with diastolic blood pressure less than 90 mm Hg at 6 months. Change equals 6-month value minus baseline value.
TABLE 4. Change ID Diastolic Blood Pressure From Baseline to 6 Months: Diet and Drug Level Comparisons
Difference of means 95% confidence intervals Comparisons (mm Hg) (mm Hg) p value Diet
Usual as compared with weight loss
Usual as compared with low sodium/high potassium Low sodium/high potassium as compared with weight loss Drug
Placebo as compared with chlorthalidone Placebo as compared with atenolol Chlorthalidone as compared with atenolol
Difference of means is the difference of first and second groups listed in Comparisons column. Positive numbers indicate second group had greater reduction in diastolic blood pressure. Confidence intervals take into account multiple comparisons. n=787. 2.48 0.71 1.77 4.44 5.10 0.67 (0.66, 4.30) (-1.11,2.53) (0.05, 3.59) (2.62, 6.26) (3.28, 6.92) (-1.15,2.49) 0.001 0.347 0.019 <0.001 <0.001 0.38
The differences among marginal changes are dis-played in Table 4. The largest subgroup changes were observed in the weight loss/chlorthalidone and weight loss/atenolol groups. The diet/placebo sub-groups all exhibited about the same drop in blood pressure. Two-way analysis of variance shows that the main effects of drug and diet were highly significant (p<0.001), but there was no significant interaction effect.
TAIM was designed to look at diastolic blood pressure changes. Analysis of the systolic blood pres-sure changes showed similar trends. (See Table 2 for results.) The percent of participants at or below selected levels of diastolic blood pressure by 6 months was also analyzed. The greatest percent below 90 mm Hg were in the weight loss/chlorthali-done groups (93%) and the weight loss/ateriolol groups (92%). The usual diet/placebo group had a response rate of 71% at or below 90 mm Hg by 6 months. If all participants placed on an active anti-hypertensive drug during therapy were removed from this group, the response rate would be 69%.
Table 4 presents the drug and diet comparisons. To achieve statistical significance here, p=0.05/ 3=0.017 must be achieved. In terms of blood pres-sure reduction, the weight loss diet was significantly better than usual diet, and either diuretic or /3-blocker was significantly better than placebo. The weight loss diet lowered blood pressure more than the low sodium/high potassium diet (p=0.019). The usual diet and low sodium/high potassium diet were statistically indistinguishable in terms of blood pres-sure reduction as was the diuretic and /3-blocker.
The first set of comparisons were designed to compare diet with low dose drug therapy. ^-Blocker therapy alone achieved a significantly greater reduc-tion of diastolic blood pressure than did a low sodium diet alone (difference=4.5 mm Hg, /?=0.001) and achieved a substantial reduction of diastolic blood pressure when compared with a weight loss diet alone (difference=3.7 mm Hg,p=0.006), although it was not significant at the prespecified level. Diuretic therapy was statistically equivalent to either a low sodium diet (difference=2.9 mm Hg, /?=0.03), or a weight loss diet (difference=2.0 mm Hg, />=0.13),
although the blood pressure reduction was greater than in either diet.
The next set of comparisons asked whether dietary therapy adds anything to a pharmacological regimen. Weight loss in combination with diuretic therapy added significantly to the lowering of diastolic blood pressure (difference=4.3 mm Hg, p=0.002). Weight loss in combination with /3-blocker therapy also added to blood pressure-lowering (difference=2.4 mm Hg, p=0.07) but less so, and the result was not significant. The low sodium/high potassium diet had little blood pressure-lowering effect when added to diuretic therapy and essentially none when added to ^-blocker therapy.
For participants receiving diuretic therapy, a weight loss diet helped lower blood pressure more than a low sodium/high potassium diet (differ-ence=2.9 mm Hg,p=0.03). Also, for participants on £-blocker therapy, a weight loss diet also helped lower blood pressure more than a low sodium/high potassium diet (difference=2.1 mm Hg, /?=0.12).
For participants on dietary therapy, adding a di-uretic or /3-blocker was equivalent in terms of blood pressure reduction. For participants on both active drug and active diet, weight loss diet plus either diuretic or /3-blocker appeared to be better at lower-ing blood pressure than a low sodium/high potassium diet plus diuretic (difference=2.6 mm Hg,p=0.05) or /3-blocker (differer.ce=2.3 mm Hg,p=0.08), but nei-ther of the results was statistically significant. All the above comparisons were examined by clinical center and by excluding crossovers (see Participation Sta-tus), and no substantial differences from the main results were noted.
To account for possible imbalances in the baseline covariates, a multiple linear regression analysis was performed. This analysis used 6-month change in blood pressure as the dependent variable and the following as independent variables: baseline diastolic blood pressure, age, sex, race, history of prior anti-hypertensive medication use, atenolol group, chlor-thalidone group, weight loss group, and the low sodium/high potassium group.
Baseline diastolic blood pressure, sex, race, and assignment to either of the two active drug groups or
the weight loss group were significant factors in determining 6-month blood pressure change. Blood pressure decreased in women more than men and in whites more than blacks. Controlling for all other variables in the model, assignment to weight loss group yielded a decrease of 2.3 mm Hg. Again, controlling for all other variables in the model, a person with a baseline diastolic blood pressure of 100 mm Hg would have a 5.9 mm Hg greater decrease in diastolic blood pressure when compared with a per-son with a baseline diastolic blood pressure of 90 mm Hg. A similar regression analysis for systolic blood pressure was performed, and this showed age as well as baseline systolic blood pressure, sex, race, and assignment to either of the two active drug groups or the weight loss group to be significant factors in determining 6-month systolic blood pressure change. Further regression modeling revealed a race-diet-drug interaction and a sex-race-diet-drug interaction. Blacks on the weight loss/chlorthalidone intervention reduced their diastolic blood pressure 5.2 mm Hg more than whites in the same intervention. Women taking chlor-thalidone reduced their diastolic blood pressure by 5.0 mm Hg more than men on chlorthalidone.
Discussion
The desirability of antihypertensive therapy for mild hypertensive patients is emphasized by the recent report from the Tecumseh Blood Pressure Study.19 Abnormalities in target organs and coronary risk factors were significantly more frequent in indi-viduals with borderline hypertension than in nor-motensive individuals. The question of which drug or diet, or combination of the two, was best for treat-ment of mild hypertension was the focus of the TAIM study. It was designed to compare changes in blood pressure among nine programs for the treat-ment of mild hypertension. Both chlorthalidone and atenolol were effective blood pressure-lowering agents in the study. Assignment to the weight reduc-tion group also lowered blood pressure. Assignment to the sodium restriction/increased potassium diet group did not substantially affect blood pressure.
The addition of weight loss to diuretic therapy provided the clearest evidence of benefit from com-bined diet and drug therapy. In addition, the weight loss/atenolol group had greater reduction in blood pressure than the usual diet/atenolol group. The attained 6-month blood pressure change of the par-ticipants on the weight loss/atenolol combination was the same as the 6-month blood pressure change of those on the weight loss/chlorthalidone combination. Yet, atenolol produced a somewhat greater blood pressure drop as sole therapy than chlorthalidone did. A partial explanation for the difference in the additive effects of weight loss between chlorthalidone and atenolol may be found in the differences in weight loss among these groups. Those assigned to weight loss plus chlorthalidone lost an average of 6.9 kg, whereas those assigned to weight loss plus atenolol lost 3.0 kg. Part of the difference in weight
loss between the two groups may be due to the volume reduction effect from the diuretic.
One interesting result in TAIM should be noted. The placebo/usual diet group had a mean drop in diastolic blood pressure of 8.0 mm Hg. This could be due to a combination of the following: 1) 20% of this group was on step-up or open-label therapy by the end of the 6 months, 2) regression to the mean, and 3) habituation to the site of blood pressure determination.
The low sodium/high potassium diet did not lower diastolic blood pressure as much as the weight loss diet. This could be due to the lack of effect of the former diet. Another possible explanation is that the TAIM participants were healthy volunteers with a relatively low baseline urinary sodium of 133 mmol/ day (compared with 180 mmol found in Framing-ham20). In addition, over the 6-month period, those assigned to the low sodium/high potassium diet achieved only a modest decrease in sodium intake and excretion and a minimal change in potassium intake and excretion.
The relatively minimal diastolic blood pressure reduction in the low sodium/high potassium group in TAIM is similar to the effect noted by Logan,21 whose participants lowered their sodium excretion from 164 mmol/day to 121 mmol/day but did not experience a change in blood pressure. However, other studies wherein the sodium excretion rate was reduced to 50-90 mmol/day did demonstrate a blood pressure-lowering effect. The Australian National Health and Medical Research Study22 showed that lowering the sodium excretion rate to 90 mmol/day had a significant lowering effect on diastolic and systolic blood pressure. McGregor et al23 showed that long-term reduction of the urinary sodium to about 50 mmol/day resulted in a significant drop in blood pressure. However, to get to 50 mmol/day requires special dietary reduction such as salt-free products that are not generally available. Therefore, the in-conclusive results of the low sodium/high potassium diet may relate to two factors: 1) present day volun-teers for a hypertension study may already have made major changes in their sodium intake; and 2) it is very difficult to achieve a sodium intake below 100 meq/ day without major modification of the usual Ameri-can diet.
Previous studies have shown that weight reduction can lower diastolic blood pressure. MacMahon et al24 reported an average weight loss of 6.8 kg to be an effective blood pressure-reducing maneuver. A pooled analysis of five weight loss diet trials25 showed a 3.1 mm Hg drop in diastolic blood pressure with an average weight loss of 9.2 kg. However, participants in the study by Haynes et al26 lost an average of 4.2 kg and had no change in blood pressure compared with control patients at 6 months. TAIM participants assigned to the weight loss group lost an equivalent amount, 4.7 kg, without significant hypotensive ben-efit in the placebo-assigned participants. However, this loss combined with a diuretic or /3-blocker
low-ered blood pressure more than either drug treatment alone or weight reduction alone.
The overall picture that emerges from the TAIM study of diet change and antihypertensive medication is 1) a successful weight loss program provided additional major benefit to thiazide or /3-blocker antihypertensive therapy in terms of blood pressure reduction; 2) a diet that yielded a modest reduction in sodium intake and a minimal increase in potassium intake in a group with a low salt intake resulted in little additional blood pressure reduction compared with the diet plus active drug; 3) in a regression analysis, weight loss alone produced a beneficial effect on blood pressure, although the beneficial effect was less marked than the addition of weight loss to thiazide-type diuretics; and 4) the evaluation of therapy for mild hypertension requires consider-ation of three components—blood pressure, risk factors, quality of life —and, if possible, cost. The long-term implications of these observations are cur-rently being assessed.
Appendix
Participating Institutions and Investigators
Birmingham Clinical Center (University of Alabama at Birmingham): Albert Oberman, MD (Principal Investigator); James M. Raczynski, PhD; James R. Kitts, PA-C; Heidi Hataway, MS, RD; Gary R. Cutter, PhD, Glenn Hughes, PhD.
Jackson Clinical Center (University of Mississippi): Herbert G. Langford, MD (Principal Investigator); Sheila Corrigan, PhD; Lori Danks, BS, RN; Stepha-nie Jennings, MS, RD.
New York Clinical Center (Albert Einstein College of Medicine): M. Donald Blaufox, MD, PhD (Prin-cipal Investigator); Maureen Magnani, RN; Judy Stern, MS, RD; Gail Miller, RN.
Nutrition and Behavioral Core (Albert Einstein College of Medicine): Sylvia Wassertheil-Smoller, PhD (Director); Judith Wylie-Rosett, EdD, RD; Charles Swencionis, PhD; Yvonne P. Raiford.
Coordinating Center (University of Texas Health Science Center at Houston, School of Public Health): C. Morton Hawkins, ScD (Principal Investigator); Barry R. Davis, MD, PhD; Neal Zimbaldi, BS; Maura O'Connell Knerr, MS.
Central Laboratory (SmithKline Bio-Science Labo-ratories, Van Nuys, Calif.): Sue Bird.
24-Hour Urinafysis Laboratory (University of Mis-sissippi School of Medicine): K.T. Holder, CLS(C).
Plasma Renin Activity Laboratory (Albert Einstein College of Medicine): Hyo-Bok Lee, MS.
Trial of Antihypertensive Interventions and Manage-ment Policy Advisory Committee: Lloyd Filer, MD; W. Gordon Walker, MD; Max Halperin, PhD (de-ceased); Elisabeth McSherry, MD; Pat Elmer, MS; Thomas P. Blaszkowski, PhD.
National Heart, Lung, and Blood Institute: Thomas
P. Blaszkowski, PhD; Jeffrey Cutler, MD.
Acknowledgment
We acknowledge the donation of study drugs from ICI Americas, Inc., Wilmington, Del.
References
1. Kannel WB, Gordon T: Coronary heart disease, atherothrom-botic brain infarction, intermittent claudication —A multivari-ate analysis of some factors relmultivari-ated to their incidence: 16 year follow-up, in The Framingham Study, An Epidemiologic Inves-tigation of Cardiovascular Disease. Appendix B. Washington, DC, US Government Printing Office, 1971
2. Hypertension Detection and Follow-up Program Cooperative Group: Five-year findings of the Hypertension Detection and Follow-up Program: I. Reduction in mortality of persons with high blood pressure, including "mild" hypertension. JAMA 1979;242:2562-2571
3. Hypertension Detection and Follow-up Program Cooperative Group: Five-year findings of the Hypertension Detection and Follow-up Program: II. Mortality by race, sex and age. JAMA 1979;242:2572-2577
4. Report by the Management Committee: The Australian Ther-apeutic Trial in Mild Hypertension. Lancet 1980;l:1261-1267 5. Hegleland A: Treatment of mild hypertension: A five-year controlled drug trial. The Oslo Study. Am J Med 1980;69: 725-732
6. Medical Research Council Working Party: MRC Trial of treatment of mild hypertension: Principal results. Br J Med 1985;291:97-104
7. MacMahon SW, Cutler JA, Furberg CD, Payne GH: The effects of drug treatment for hypertension on morbidity and mortality from cardiovascular disease: A review of randomized controlled trials. Prog Cardiovasc Dis 1987;29(suppl 1):99-118 8. The 1988 Joint National Committee: The 1988 report of the
Joint National Committee on Detection, Evaluation, and Treatment of High Blood Pressure. Arch Intern Med 1988;148: 1023-1038
9. In conclusion: The 1986 guidelines for the treatment of mild hypertension, in Strasser T, Ganten D (eds): Mild Hyperten-sion: From Drug Trials to Practice. New York, Raven Press Publishers, 1987, pp 285-292
10. Kaplan NM: Maximally reducing cardiovascular risk in the treatment of hypertension. Ann Intern Med 1988;109:36-40 11. Davis BR, Blaufox MD, Hawkins CM, Langford HG,
Ober-man A, Swencionis C, Wassertheil-Smoller S, Wylie-Rosett J, Zimbaldi N: Trial of Antihypertensive Interventions and Man-agement: Design, methods, and selected baseline results. Controlled Clin Trials 1989;10:ll-30
12. 1983 Metropolitan height and weight tables. Stat Bull NY Metropolitan Life Insurance Co 1983;64:2-9
13. Langford HG, Blaufox MD, Oberman A, Hawkins CM, Curb JD, Cutter GR, Wassertheil-Smoller S, Pressel S, Babcock C, Abemethy JD, Hotchkiss J, Tyler M: Dietary therapy slows the return of hypertension after stopping prolonged medica-tion. JAMA 1985;253:657-664
14. Wassertheil-Smoller S, Langford HG, Blaufox MD, Oberman A, Hawkins CM, Levine B, Cameron M, Babcock C, Pressel S, Caggiula A, Cutter G, Curb D, Wing R: Effective dietary intervention in hypertensives: Sodium restriction and weight reduction. J Am Diet Assoc 1985;85:423-430
15. Oberman A, Wassertheil-Smoller S, Langford H, Blaufox MD, Davis BR, Blaszkowski T, Zimbaldi N, Hawkins CM: Pharma-cologic and nutritional treatment of mild hypertension: Change in cardiovascular risk status. Ann Intern Med 1990;112: 89-95
16. Croog SH, Levine S, Testa MA, Brown B, Bulpitt CJ, Jenkins CD, Herman GL, Williams GH: The effects of antihyperten-sive therapy on the quality of life. N Engl J Med 1986^14: 1657-1664
17. Fleiss JL: The Design and Analysis of Clinical Experiments. New York, John Wiley & Sons, Inc, 1986
18. Curb JD, Labarthe DR, Cooper SP, Cutter GR, Hawkins CM: Training and certification of blood pressure observers. Hyper-tension 1983;5:610-614
19. Julius S, Jamerson K, Mejia A, Krause L, Schork N, Jones K: The association of borderline hypertension with target organ changes and higher coronary risk —Tecumseh Blood Pressure Study. JAMA 1990;264:354-358
20. Dawber TR, Kannel WB, Kagan A, Donabedian RK, McNa-mara PM, Pearson G: Environmental factors in hypertension, in Stanke J, Stamler R, Pullman TN (eds): The Epidemiology of Hypertension. New York, Grune & Stratton, Inc, 1967 21. Logan AG: Sodium manipulation in the management of
hypertension. The view against its general use. Can J Physiol Pharmacol 1986;64:743-807
22. Australian National Health and Medical Research Council Dietary Salt Study Management Committee: Fall in blood pressure with modest reduction in dietary salt intake in mild hypertension. Lancet 1989;l:399-402
23. MacGregor GA, Markandu ND, Sagnella GA, Singer DRJ, Cappuccio FP: Double-blind study of three sodium intakes
and long-term effects of sodium restriction in essential hyper-tension. Lancet 1989;2:1244-1247
24. MacMahon SW, MacDonald GJ, Bernstein L, Andrews G, Blacket RB: A randomized controlled trial of weight reduction and metoprolol in the treatment of hypertension in young overweight patients. Clin Exp Pharmacol Physiol 1985;12: 267-271
25. MacMahon S, Cutler J, Brittain E, Higgins M: Obesity and hypertension: Epidemiological and clinical issues. Ew Health J 1987;8(suppl B):57-80
26. Haynes RB, Harper AC, Costley SR, Johnston M, Logan AG, Flanagan PT, Sackett DL: Failure of weight reduction to reduce mildly elevated blood pressure, a randomized trial. / Hypertcns 1984;2:535-539
KEY WORDS
