Download Blood pressure change with weight loss is affected by diet type in

Blood pressure change with weight loss is affected by diet
type in men1–3
Caryl A Nowson, Anthony Worsley, Claire Margerison, Michelle K Jorna, Sandra J Godfrey, and Alison Booth
ABSTRACT
Background: Weight loss reduces blood pressure, and the Dietary
Approaches to Stop Hypertension (DASH) diet has also been shown
to lower blood pressure.
Objective: Our goal was to assess the effect on blood pressure of 2
weight-reduction diets: a low-fat diet (LF diet) and a moderate-sodium,
high-potassium, high-calcium, low-fat DASH diet (WELL diet).
Design: After baseline measurements, 63 men were randomly assigned to either the WELL or the LF diet for 12 wk, and both diet
groups undertook 0.5 h of moderate physical activity on most days
of the week.
Results: Fifty-four men completed the study. Their mean (앐SD) age
was 47.9 앐 9.3 y (WELL diet, n ҃ 27; LF diet, n ҃ 27), and their
mean baseline home systolic and diastolic blood pressures were
129.4 앐 11.3 and 80.6 앐 8.6 mm Hg, respectively. Body weight
decreased by 4.9 앐 0.6 kg (앐SEM) in the WELL group and by 4.6 앐
0.6 kg in the LF group (P 쏝 0.001 for both). There was a greater
decrease in blood pressure in the WELL group than in the LF group
[between-group difference (week 12 – baseline) in both SBP (5.5 앐
1.9 mm Hg; P ҃ 0.006) and DBP (4.4 앐 1.2 mm Hg; P ҃ 0.001)].
Conclusions: For a comparable 5-kg weight loss, a diet high in
low-fat dairy products, vegetables, and fruit (the WELL diet) resulted in a greater decrease in blood pressure than did the LF diet.
This dietary approach to achieving weight reduction may confer an
additional benefit in reducing blood pressure in those who are
overweight.
Am J Clin Nutr 2005;81:983–9.
KEY WORDS
calcium
Weight loss, blood pressure, diet, potassium,
INTRODUCTION
Hypertension is an important public health issue and contributes to the incidence of stroke and coronary heart disease (1). The
prevalence of hypertension in Australia was recently shown to be
앒29% (2). Furthermore, hypertension accounts for 6.1% of the
total problems managed in general practice (3). Education pertaining to nutrition and weight accounted for 10% of all nonpharmacologic treatments provided by general practitioners and
was one of the 3 most common forms of advice (3). Around the
world, the incidence of overweight and obesity has increased (4).
The prevalence of obesity in Australia has more than doubled in
the past 20 y, and almost 60% of adults have been estimated to be
overweight or obese (5). There is a direct positive relation between overweight and hypertension, such that it has been estimated that the control of obesity may eliminate 48% of the
hypertension in whites (6). Dietary sodium increases blood pressure (BP), whereas dietary potassium lowers the risk of hypertension and stroke. In a controlled intervention study, a multifaceted dietary approach (DASH: Dietary Approaches to Stop
Hypertension) that included a diet high in fruit, vegetables, and
low-fat dairy products was shown to result in large decreases in
BP (11 mm Hg systolic and 5 mm Hg diastolic pressure in
hypertensive persons and 5 mm Hg systolic and 3 mm Hg diastolic in normotensive persons) (7). Therefore, the aim of the
present study was to determine the effect on BP of a DASH-type
weight-loss diet (WELL diet) and to compare this with usual
low-fat dietary advice (LF diet) in free-living individuals who
selected and prepared their own food.
SUBJECTS AND METHODS
Subjects
Subjects were recruited through newspaper articles advertising the study and at BP measurement sessions provided in workplaces and at the study center. Subjects were eligible if they were
male, aged 쏜 25 y, and had a seated office BP of 욷120 mm Hg
systolic blood pressure (SBP) or 욷80 mm Hg diastolic blood
pressure (DBP) at their first visit (mean of the last 3 of 4 measurements taken at 1-min intervals). Subjects who were taking
antihypertensive medication were included, provided they were
willing to maintain their current medication level. Subjects were
excluded if they had experienced a cardiovascular event in the
past 6 mo, had insulin-dependent diabetes, were taking medications such as warfarin or phenytoin, ate their main meal outside
the home more than twice per week, drank 쏜30 standard (10 g
alcohol) alcoholic drinks per week, were planning to change
smoking habits, or were unwilling to cease taking dietary supplements (including vitamins). Subjects were included if they
had a body mass index (BMI; in kg/m2) between 25 and 35. All
subjects provided written informed consent before starting the
study, which was approved by the Deakin University Human
Research Ethics Committee.
1
From the Centre for Physical Activity and Nutrition, School of Exercise
and Nutrition Sciences, Deakin University, Burwood Highway, Burwood,
Australia.
2
Supported by the Dairy Research and Development Corporation.
3
Address reprint requests to CA Nowson, School of Exercise and Nutrition Sciences, Deakin University, 221 Burwood Highway, Burwood, VIC
3125 Australia. E-mail: [email protected].
Received June 29, 2004.
Accepted for publication December 14, 2004.
Am J Clin Nutr 2005;81:983–9. Printed in USA. © 2005 American Society for Clinical Nutrition
983
984
NOWSON ET AL
Two hundred twenty persons responded to advertisements,
and 165 of these were sent a screening questionnaire and invited
to attend further screening. Ninety-four men attended one screening appointment, and 63 who met the entry criteria and wished to
participate undertook baseline home BP measurements for 2 wk and
were then randomly assigned to either the LF or the WELL diet.
Study design
Subjects were seen twice at baseline, and commenced a 12-wk
intervention study and were seen at weeks 2, 4, 8, and 12. Phone
contact was made with the subjects at weeks 6 and 10. Clinical
BP, height, and weight were measured at baseline. Subjects monitored their home BP daily for 2 wk before being randomly
assigned (stratified by antihypertensive medication use) to 1 of
the 2 diets. Randomization was performed by the chief investigator (CAN) with the use of a random number generator in blocks
of 8 (EXCEL 2000; Microsoft Corporation, Redmond, WA).
Anthropometry and blood pressure measurement
Height was measured with a wall-mounted stadiometer. Body
weight was measured at each visit on a digital scale while the
subjects wore light clothing and no shoes. Waist circumference
was measured with a fiberglass tape measure anteriorly halfway
between the lowest lateral portion of the ribcage and the iliac
crest. Home BP was measured on the left arm with the use of an
automated BP monitor (AND UA-767-PC; A&D Co Ltd, Tokyo,
Japan). Subjects were trained to correctly apply the cuff and were
instructed to take their BP measurements alone, at the same time
of day, after 5 min of rest in a quiet room and to take 3 measurements with 1 min in between (the mean of the last 2 measurements taken each day was used in the analysis). BP measurement
data were downloaded directly to a computer by the study staff at
the end of each fortnight.
Biochemical indexes
Fasting (10 h overnight) blood samples were collected at baseline and at the end of the study. Serum total cholesterol, HDL
cholesterol, and triacylglycerol were measured on the Hitachi
704 analyzer by using enzymatic reagents (Boehringer, Mannheim, Germany). LDL cholesterol was calculated by use of the
Friedewald equation (8).
Dietary assessment
Subjects completed a 24-h dietary record each fortnight on the
day before their visit with study staff. Trained research personnel
checked this record. Dietary information was entered into a dietary analysis program (FOODWORKS, Professional Edition,
version 3.02; Xyris Software, New York, NY) to calculate daily
nutrient intakes. The mean of two 24-h records at baseline and the
mean of four 24-h records at weeks 2, 4, 8, and 12 of the intervention were used in the analysis. A food-frequency questionnaire was completed at baseline and at week 12 to assess usual
intake of fruit, vegetables, and dairy products.
Lifestyle intervention
Subjects were assisted with setting goals for exercise and diet.
At each visit, a trained dietitian set dietary and exercise goals (욷3
goals per visit, including 욷1 for exercise and 욷1 for diet). Recipes, educational materials (diet and exercise), and tips to encourage compliance were provided to all subjects.
Dietary instruction
Dietary counseling was overseen by the coordinating dietitian
(CM) and was provided by trained research staff. The WELL diet
was based on our previous OZDASH diet (9), which had been
modified from the US DASH diet (7). This diet included advice
to consume 욷4 servings of fruit or fruit juice [1 serving ҃ 1
medium piece of fruit (100 g) or fruit juice (200 mL)], 욷4 servings of vegetables [1 serving ҃ 0.5 cup cooked vegetables (50 g),
1 cup salad vegetables, or 1 medium potato] and 욷3 servings
of nonfat dairy products [1 serving ҃ milk (200 mL), yogurt
(200 g), or cottage or ricotta cheese (0.5 cup)] per day. Fish (1
serving ҃ 120 g cooked) was to be consumed 욷3 times per week,
legumes (1 serving ҃ 1 cup cooked) at least once per week, and
unsalted nuts and seeds (1 serving ҃ 30 g) 4 times per week. Red
meat was restricted to no more than 2 servings (1 serving ҃
90 –100 g cooked) per week and fat to a maximum of 4 servings
(4 teaspoons) per day. Subjects were advised to avoid butter,
added salt (table or cooking), and obviously salty foods and to use
lower-salt (쏝380 mg Na per 100 g) mono- or polyunsaturated
margarine. Those in the WELL group received a detailed dietary
information booklet, recipes, and simple advice (tips).
The LF group was advised to limit their intake of high-energy
foods and drinks, reduce their saturated fat intake, choose mainly
plant-based foods, consume nonfat or reduced-fat milk and yogurt, limit their cheese and ice cream intake to twice per week,
select lean meat, and avoid frying foods in fat. No specific targets
were set. The “Healthy Weight Guide” booklet by the National
Heart Foundation of Australia (2002) was provided, together
with the same recipes and tips as received by the WELL group.
A maximum of 4 caffeine-containing drinks per day (eg, cola
drinks, coffee, and tea) and 4 standard (10-g alcohol) alcoholic
drinks per week were permitted for both diet groups.
Several incentives were included. Both groups received measuring cups and spoons and individual feedback on their daily
intake of fruit and vegetables. Those in the WELL group also
received individual feedback compared with the specified diet
targets (fruit, vegetables, and dairy). A dairy product of their
choice [eg, 200-mL tub of nonfat yogurt or a packet of reducedfat (쏝15% fat) cheese slices] was offered once during the study
to all subjects. Subjects could also participate in a drawing to win
a double movie pass, and feedback on the group’s progression
regarding targets and weight loss was provided graphically
throughout the study.
The main difference between the LF diet and the WELL diet
was that the WELL diet had specified targets for fruit, vegetable,
and dairy intake, whereas the LF diet provided general guidelines
focusing on increasing fruit and vegetable intake and reducing fat
intake, particularly saturated fat.
Physical activity
All subjects were required to participate in moderate-intensity
exercise for 욷30 min on all or most days of the week. The “Be
Active Every Day” booklet by the National Heart Foundation
(1999) was given to each participant, and individual exercise
goals were set at each visit. Information was provided on calculating maximum heart rate [220 –age (y)], and subjects were
advised to increase their heart rate to 60 –79% of their maximum
heart rate to reach a moderate level of exercise intensity and to
maintain this for 욷30 min for each session. The amount of
walking was monitored by using the CHAMPS questionnaire
985
DIET, WEIGHT LOSS, AND BLOOD PRESSURE
(10) at baseline (for 4 wk) and the average hours per week
calculated for the intervention period.
TABLE 1
Baseline characteristics of the study groups1
Statistical analysis
Data were analyzed by using SPSS for WINDOWS (version
11.0; SPSS Inc, Chicago, IL) to calculate the descriptive statistics
and perform the regression analysis. Mean home BP readings
were calculated for each 2-wk period. Unpaired Student’s t tests
were used to evaluate the difference between the LF and WELL
diets in the changes between baseline and the last visit. P values
of 0.05 were considered to be significant. Additionally, the effect
of the diet intervention was assessed by using two-factor
repeated-measures analysis of variance (diet ҂ time) with covariates of baseline weight and BP included in specific analysis
where indicated. Baseline values were used as covariates in GLM
univariate analysis of variance to calculate the adjusted mean
changes to test for the difference between groups for the dietary
data only.
RESULTS
Nine subjects dropped out before completing the study (4 in
the LF group and 5 in the WELL group); the subjects who
dropped out did not differ significantly from the rest of the group
with respect to age or BMI. Eight found it too difficult to comply
with the study demands, and one moved interstate.
Of the 54 men who completed the study, 18 were taking antihypertensive medications (9 WELL, 9 LF). Subjects in the
WELL group who were receiving antihypertensive therapy included 4 receiving single therapy [1 taking an angiotensinconverting enzyme (ACE) inhibitor, 1 taking a calcium-channel
blocker, and 2 taking angiotensin II receptor antagonists (AT1)]
and 5 receiving combination therapy (AT1 ѿ ␤-blocker, n ҃ 1;
AT1 ѿ diuretic, n ҃ 2; AT1 ѿ calcium-channel blocker, n ҃ 1;
ACE inhibitor ѿ diuretic, n ҃ 1). Subjects in the LF group who
were receiving antihypertensive therapy included 5 receiving single
therapy (2 taking an ACE inhibitor, 1 taking an AT1, and 2 taking
calcium-channel blockers) and 4 receiving combination therapy
(ACE inhibitor ѿ diuretic, n ҃ 1; AT1 ѿ diuretic, n ҃ 1; ACE
inhibitor ѿ diuretic, n ҃ 1; AT1 ѿ diuretic ѿ ␤-blocker, n ҃ 1).
At baseline, there were no significant differences in dietary
intakes of fruit, vegetables, or calcium-containing dairy products
between groups; however, those randomly assigned to the LF
group were heavier, were taller, and had a greater waist measurement and had a BMI of 31 compared with 30 for the WELL
group (Table 1 and Table 2).
Effects of the intervention
The amount of time spent walking increased in both groups
over the intervention period, with no significant difference between the groups: WELL group increased to 4.4 앐 0.7 (x៮ 앐 SEM)
h/wk and LF group to 4.6 앐 0.5 h/wk (both P 쏝 0.01 for the
change from baseline).
There were no significant differences between the groups at
baseline in fruit, vegetable, and dairy intakes as recorded on the
food-frequency questionnaire. At week 12, the WELL group
reported a higher intake of dairy products, but there was no
significant difference between the groups in fruit and vegetable
intakes. Fruit intake increased significantly during the diet compared with baseline for both groups (both P ҃ 0.001; Table 3).
Age (y)
Age range (y)
Height (cm)
BMI (kg/m2)
Waist circumference (cm)
Office SBP (mm Hg)
Office DBP (mm Hg)
Office pulse (beats/min)
WELL diet group
(n ҃ 27)
LF diet group
(n ҃ 27)
47.1 앐 10.32
30–66
172.5 앐 5.2
29.6 앐 2.3
102.8 앐 7.7
136.4 앐 16.2
88.0 앐 9.7
69.3 앐 12.3
48.8 앐 8.3
28–64
177.4 앐 5.33
31.2 앐 2.44
110.3 앐 6.75
133.6 앐 9.7
88.7 앐 6.0
73.2 앐 9.9
1
WELL, DASH-type weight-loss diet (moderate sodium, high potassium, high calcium, low fat, with less red meat and more fish); LF, low-fat
diet; SBP, systolic blood pressure; DBP, diastolic blood pressure.
2
x៮ 앐 SD (all such values).
3–5
Significantly different from the WELL diet group (unpaired t test):
3
P 쏝 0.01, 4P 쏝 0.05, 5P 쏝 0.001.
For the WELL group only, intakes of dairy products and vegetables were significantly higher during the diet than at baseline
(dairy P ҃ 0.001, vegetables P ҃ 0.001; Table 3).
After adjustment for baseline dietary intake, the 24-h dietary
records indicated that the reductions in dietary fat (g/d), saturated
fat (g/d), percent of energy from fat, percent of energy from
saturated fat, and sodium (mg/d) were greater in the WELL group
than in the LF group, and the increases in the percent of energy
from protein, percent of energy from carbohydrate, potassium
(mg/d), calcium (mg/d), magnesium (mg/d), and phosphorus
(mg/d) were greater in the WELL group than in the LF group
(Table 3).
Weight and blood pressure changes
Weight decreased significantly in both groups by 앒5.0 kg
(P 쏝 0.001 for both), with subjects in the WELL group losing 6%
of body weight (P 쏝 0.001) and those in the LF group losing 5%
(P 쏝 0.001; Table 2). The rate of weight loss was not significantly different between the diet groups throughout the study. In
the first 2 wk, weight loss was 1.2 앐 0.2 kg in the WELL group
and 1.5 앐 0.2 kg in the LF group, and the effect did not differ
significantly between diet groups (two-factor ANOVA: time ҂
diet effect, NS).
The greatest decrease in BP in both groups was seen after 4 wk
of intervention (Figure 1). There was a greater decrease in the
WELL group than in the LF group [between-group difference
(week 12 – baseline) in both SBP (5.5 앐 1.9 mm Hg; P ҃ 0.006)
and DBP (4.4 앐 1.2 mm Hg; P ҃ 0.001)]. Pulse rate also fell by
3.8 앐 1.6 beats/min more in the WELL group (P ҃ 0.023; Table 2).
The percentage decrease in SBP was 5.5 앐 1.0% in the WELL
group compared with 1.4 앐 0.9% in the LF group. The percentage decrease in DBP was 6.4 앐 1.1% in the WELL group compared with 1.0 앐 1.0% in the LF group. The significance of the
difference in the percentage change between groups was P ҃
0.005 (SBP) and P ҃ 0.001 (DBP).
After adjustment for baseline BP and body weight, the difference in the decrease in SBP and DBP between groups remained
[SBP: 5.2 앐 1.8 mm Hg (P ҃ 0.006); DBP: 4.8 앐 1.3 mm Hg
(P ҃ 0.001)]. Overall, there was a significant effect of diet on BP
(SBP, P ҃ 0.006; DBP, P ҃ 0.001; n ҃ 54; repeated measures,
two-factor ANOVA: time ҂ diet effect). Adding baseline body
986
NOWSON ET AL
TABLE 2
Intervention outcomes at baseline and 12 wk by randomized group
assignment1
Weight (kg)
Baseline
12 wk
Change
Walking (h/wk)
Baseline
12 wk
Change
Home SBP (mm Hg)
Baseline
12 wk
Change
Home DBP (mm Hg)
Baseline
12 wk
Change
Home pulse (beats/min)
Baseline
12 wk
Change
Cholesterol (mmol/L)
Baseline
12 wk
Change
HDL cholesterol (mmol/L)
Baseline
12 wk
Change
LDL cholesterol (mmol/L)
Baseline
12 wk
Change
Triacylglycerol (mmol/L)
Baseline
12 wk
Change
WELL diet group
(n ҃ 27)
LF diet group
(n ҃ 27)
88.2 앐 1.8
83.3 앐 1.8
Ҁ4.9 앐 0.63
98.2 앐 1.9
93.6 앐 1.8
Ҁ4.6 앐 0.63
3.1 앐 0.5
4.4 앐 0.7
ѿ1.4 앐 0.44
2.3 앐 0.4
4.6 앐 0.5
ѿ2.2 앐 0.53
0.249
130.9 앐 2.9
123.3 앐 2.2
Ҁ7.6 앐 1.53
128.4 앐 2.0
126.3 앐 1.7
Ҁ2.1 앐 1.2
0.494
82.3 앐 1.6
76.9 앐 1.6
Ҁ5.4 앐 0.93
84.0 앐 1.6
83.0 앐 1.4
Ҁ1.0 앐 0.8
0.466
67.6 앐 2.1
61.4 앐 1.7
Ҁ6.2 앐 1.33
69.4 앐 2.1
67.1 앐 2.0
Ҁ2.4 앐 1.05
5.3 앐 0.2
4.7 앐 0.2
Ҁ0.7 앐 0.13
5.3 앐 0.2
4.8 앐 0.2
Ҁ0.4 앐 0.13
0.756
1.16 앐 0.04
1.05 앐 0.03
Ҁ0.11 앐 0.033
1.14 앐 0.04
1.09 앐 0.05
Ҁ0.05 앐 0.025
0.697
3.4 앐 0.1
2.9 앐 0.1
Ҁ0.4 앐 0.13
3.2 앐 0.2
2.9 앐 0.2
Ҁ0.3 앐 0.14
0.571
1.9 앐 0.1
1.6 앐 0.1
Ҁ0.3 앐 0.1
2.0 앐 0.2
2.0 앐 0.2
Ҁ0.1 앐 0.2
0.501
P2
0.001
0.75
0.200
0.006
0.001
0.536
0.023
0.099
0.085
0.419
0.346
All values are x៮ 앐 SEM. WELL, DASH-type weight-loss diet (moderate sodium, high potassium, high calcium, low fat, with less red meat and
more fish); LF, low-fat diet; SBP, systolic blood pressure; DBP, diastolic
blood pressure.
2
Difference between the WELL and LF diet groups (unpaired t test).
Note that values are not adjusted. Adding baseline blood pressure or weight
as a covariate did not significantly affect the mean changes.
3–5
Significance of change (12 wk Ҁ baseline; paired t test): 3P 쏝 0.001,
4
P 쏝 0.01, 5P 쏝 0.05.
1
weight as a covariate to the model did not significantly affect the
results (SBP, P ҃ 0.001; DBP, P ҃ 0.001; n ҃ 54; repeatedmeasures, two-factor ANOVA: time ҂ diet effect). Additionally,
a model that included baseline BP and body weight as covariates
did not significantly affect the results (SBP, P ҃ 0.003; DBP, P ҃
0.001; n ҃ 54; repeated-measures, two-factor ANOVA: time ҂
diet effect). The pulse rate was lower in the WELL group than in
the LF group (P ҃ 0.031; n ҃ 54; repeated-measures, two-factor
ANOVA: time ҂ diet effect).
Serum total cholesterol, HDL cholesterol, and LDL cholesterol decreased significantly in both groups by the end of the
study; however, there was no significant change in triacylglycerol in the LF group and a tendency for a decrease in the WELL
group of 0.3 앐 0.1 mmol/L (P ҃ 0.051; Table 2).
Regression analysis indicated that initial weight was not related to changes in SBP and DBP. However, the percentage
weight loss was related to the percentage change in SBP and DBP
[SBP: R2 ҃ 0.16, ␤ (앐SE) ҃ 0.65 앐 0.20, P ҃ 0.003; DBP: R2 ҃
0.16, ␤ ҃ 0.73 앐 0.23, P ҃ 0.003]; a 10% change in weight was
associated with a 7% decrease in both SBP and DBP. Univariate
linear regression analysis indicated that the increase in total dairy
product intake was associated with the decrease in DBP (R2 ҃
0.118, ␤ ҃ 0.959 앐 0.364, P ҃ 0.011) and the increase in
vegetable intake was associated with the decrease in DBP (R2 ҃
0.071, ␤ ҃ 0.968 앐 0.487, P ҃ 0.052).
DISCUSSION
The present study investigated the effects on home BP of 2
dietary interventions— one based on the DASH dietary pattern
and the other a usual low-fat diet— combined with increased
physical activity to achieve weight loss. We found that the subjects in both diet groups achieved a weight loss of 앒5– 6% of
body weight over 3 mo. Those in the WELL group, however, had
greater decreases in SBP and DBP of 앒5 and 4 mm Hg, respectively. The groups were well matched at baseline for BP and for
the number of subjects taking antihypertensive medication (33%
in each group), although BMI was initially one unit higher in the
LF group than in the WELL group. This difference, however, is
unlikely to have contributed to the increased effectiveness of the
WELL diet with respect to BP, because there was no significant
difference in percentage weight loss between the groups.
The reason for the greater decrease in BP with the WELL diet
is not clear. We found no significant difference between the 2
groups in the change in blood lipids, although those in the WELL
diet group did appear to have a greater reduction in total fat and
particularly saturated fat intake. Our power to detect a difference
in blood lipids, however, was low because of the limited number
of subjects.
Some of the dietary differences between the WELL and the LF
diet may explain some of the improved BP-lowering effect of the
WELL diet, specifically, the increase in dietary potassium,
which has been shown to lower BP by 앒3 mm Hg systolic and 2
mm Hg diastolic (11). Dietary calcium and magnesium have also
been weakly associated with lower BP in population studies (12,
13), although evidence for a BP-lowering effect in controlled
intervention studies is not consistent (14). It appears, however,
that a diet combining these nutrient changes— eg, lower sodium
and saturated fat and higher potassium, calcium, magnesium, and
phosphorus—within a diet and physical activity pattern that induces negative energy balance achieves a greater reduction in BP
than does a low-fat diet.
The food-frequency questionnaire indicated a difference between groups at the end of the study in dairy intake only and not
in fruit and vegetable intakes. This likely reflects the insensitivity
of a food-frequency questionnaire in picking up relatively small
changes in food intake. It may also indicate that the increase in
dairy products, when combined with more vegetables, is a significant factor with respect to BP reduction, particularly because
the linear regression analysis indicated that the change in total
dairy intake together with the change in vegetable intake was
univariately associated with the reduction in DBP.
DIET, WEIGHT LOSS, AND BLOOD PRESSURE
987
TABLE 3
Intervention nutrient outcomes at baseline and 12 wk by randomized group assignment1
WELL diet group (n ҃ 27)
Fruit (servings/d)3
Baseline
Intervention
Change
Vegetables (servings/d)3
Baseline
Intervention
Change
Dairy products (servings/d)3,6
Baseline
Intervention
Change
Energy (MJ/d)7
Baseline
Intervention
Change
Protein (g/d)7
Baseline
Intervention
Change
Fat (g/d)7
Baseline
Intervention
Change
Saturated fat (g/d)7
Baseline
Intervention
Change
Carbohydrate (g/d)7
Baseline
Intervention
Change
Cholesterol (mg/d)7
Baseline
Intervention
Change
Protein (% of energy)7
Baseline
Intervention
Change
Fat (% of energy)7
Baseline
Intervention
Change
Saturated fat (% of energy)7
Baseline
Intervention
Change
Carbohydrate (% of energy)7
Baseline
Intervention
Change
Alcohol (g/d)7
Baseline
Intervention
Change
Alcohol (% of energy)7
Baseline
Intervention
Change
Fiber (g/d)7
Baseline
Intervention
Change
Sodium (mmol/d)7
Baseline
Intervention
Change
LF diet group (n ҃ 27)
P2
2.2 앐 0.24
3.8 앐 0.2
1.5 (1.0, 1.9)5
2.7 앐 0.3
3.6 앐 0.3
1.1 (0.6, 1.5)
0.147
2.2 앐 0.2
3.4 앐 0.2
1.1 (0.6, 1.6)
2.6 앐 0.2
3.0 앐 0.3
0.5 (0.0, 0.9)
0.262
2.3 앐 0.2
4.0 앐 0.2
1.5 (1.0, 2.0)
2.9 앐 0.3
2.6 앐 0.3
Ҁ0.0 (Ҁ0.5, 0.5)
0.147
10.4 앐 0.6
8.4 앐 0.3
Ҁ2.4 (Ҁ3.0, Ҁ1.7)
11.3 앐 0.5
8.7 앐 0.4
Ҁ2.3 (Ҁ2.9, Ҁ1.6)
0.262
108.2 앐 6.2
104.1 앐 3.6
Ҁ7.5 (Ҁ14.8, Ҁ0.2)
115.7 앐 5.7
103.7 앐 3.
Ҁ8.6 (Ҁ15.9, Ҁ1.3)
0.373
85.8 앐 6.9
44.0 앐 3.3
Ҁ48.4 (Ҁ56.0, Ҁ40.8)
101.2 앐 8.3
57.4 앐 4.4
Ҁ37.2 (Ҁ44.8, Ҁ29.6)
0.160
31.5 앐 3.2
11.5 앐 1.0
Ҁ23.3 (Ҁ25.8, Ҁ20.7)
38.7 앐 3.2
19.2 앐 1.4
Ҁ16.1 (Ҁ18.7, Ҁ13.6)
0.122
302.4 앐 19.8
289.8 앐 11.4
Ҁ14.6 (Ҁ38.2, 9.0)
307.6 앐 13.1
270.1 앐 13.0
Ҁ35.5 (Ҁ59.1, Ҁ12.0)
0.827
331.7 앐 24.6
237.1 앐 21.0
Ҁ90.5 (Ҁ128.4, Ҁ52.7)
0.840
323.7 앐 31.1
184.3 앐 16.0
Ҁ143.5 (Ҁ181.4, Ҁ105.6)
16.8 앐 0.6
19.8 앐 0.5
3.0 (1.9, 4.1)
6.6 앐 0.7
19.5 앐 0.7
2.8 (1.7, 3.9)
0.210
0.075
0.001
0.810
0.834
0.042
0.001
0.214
0.052
0.814
0.047
29.9 앐 1.4
18.5 앐 1.1
Ҁ12.0 (Ҁ14.3, Ҁ9.8)
31.8 앐 1.6
23.8 앐 1.3
Ҁ7.3 (Ҁ9.6, Ҁ5.1)
0.380
10.8 앐 0.7
4.9 앐 0.4
Ҁ6.5 (Ҁ7.3, Ҁ5.7)
12.1 앐 0.7
8.0 앐 0.5
Ҁ3.6 (Ҁ4.5, Ҁ2.8)
0.184
49.4 앐 1.7
59.1 앐 1.4
10.4 (7.6, 13.3)
12.0 앐 3.3
6.4 앐 2.5
Ҁ6.8 (Ҁ11.5, Ҁ2.2)
3.5 앐 1.0
3.1 앐 1.0
Ҁ0.5 (Ҁ2.1, 1.2)
29.6 앐 1.4
37.5 앐 1.6
8.1 (4.7, 11.5)
2982.1 앐 206.5
2023.9 앐 108.9
Ҁ1144.1 (Ҁ1417.0, Ҁ871.2)
47.3 앐 1.8
53.5 앐 1.6
5.5 (2.6, 8.3)
0.004
0.001
0.396
0.017
17.6 앐 4.5
10.1 앐 3.7
Ҁ6.2 (Ҁ10.9, Ҁ1.6)
0.325
4.1 앐 1.0
4.2 앐 1.1
0.2 (Ҁ1.4, 1.9)
0.631
28.9 앐 1.6
34.3 앐 1.9
5.2 (1.8, 8.6)
3457.7 앐 291.5
2669.2 앐 172.9
Ҁ602.7 (Ҁ875.6, Ҁ329.8)
0.862
0.570
0.735
0.225
0.190
0.007
(Continued)
988
NOWSON ET AL
TABLE 3 (Continued)
Potassium (mmol/d)7
Baseline
Intervention
Change
Calcium (mg/d)7
Baseline
Intervention
Change
Magnesium (mg/d)7
Baseline
Intervention
Change
Phosphorus (mg/d)7
Baseline
Intervention
Change
WELL diet group (n ҃ 27)
LF diet group (n ҃ 27)
P2
3972.9 앐 185.3
5270.1 앐 197.1
1265.7 (894.3, 1637.2)
4056.5 앐 185.1
4344.4 앐 180.1
319.4 (Ҁ52.1, 690.9)
0.751
981.9 앐 68.7
1415.8 앐 67.6
462.8 (349.9, 573.8)
889.6 앐 62.0
986.6 앐 53.0
69.2 (Ҁ42.7, 181.1)
0.324
411.4 앐 21.6
413.4 앐 17.0
Ҁ1.4 (Ҁ36.5, 33.7)
0.751
1845.7 앐 82.6
1756.0 앐 65.1
Ҁ84.1 (Ҁ221.3, 53.2)
0.910
420.9 앐 20.6
474.3 앐 1.49
56.8 (21.7, 91.9)
1831.2 앐 97.3
2025.1 앐 75.9
188.3 (51.1, 325.6)
0.001
0.001
0.022
0.007
1
WELL, DASH-type weight-loss diet (moderate sodium, high potassium, high calcium, low fat, with less red meat and more fish); LF, low-fat diet.
Difference between the WELL and LF diet groups (unpaired t test).
3
Based on food-frequency questionnaire.
4
x៮ 앐 SEM (all such values).
5
x៮ ; 95% CI in parentheses (all such values). Mean changes were adjusted for baseline values.
6
Calcium-containing, reduced-fat dairy foods (eg, milk, cheese, and yogurt).
7
Based on 24-h recall.
2
Although dietary records are not a good measure of actual
sodium intake, after adjustment for baseline sodium intake, there
was a significant reduction in dietary sodium in the WELL group
only. It is therefore likely that those following the WELL diet did
have a lower intake of sodium. Subjects were advised to avoid
FIGURE 1. Mean (앐SEM) systolic (SBP) and diastolic (DBP) blood
pressure over 12 wk of intervention in the LF (䉬) and WELL (■) diet groups.
WELL, DASH-type weight-loss diet (moderate sodium, high potassium,
high calcium, low fat, with less red meat and more fish); LF, low-fat diet.
Two-factor repeated-measures ANOVA: time ҂ diet effect, P ҃ 0.006 for
SBP, P ҃ 0.001 for DBP; n ҃ 54 (unadjusted).
added salt and obviously salty foods, and we found in a previous
study (9) that this type of dietary advice can result in a decrease
in 24-h urinary sodium of 앒30 mmol in a weight-stable situation.
In the present study, in which there was a reduction in energy
intake, the reduction in sodium is likely to have been greater, but
without having data on 24-h urine collections we cannot confirm
this finding.
The decreases in BP in the present study were somewhat
greater than in other studies, particularly the 5–mm Hg decrease
in DBP. A meta analysis of weight loss and BP indicated an
average decrease in SBP and BDP of 4 mm Hg in studies with
energy restriction with or without exercise (15), whereas we
found decreases of 8 and 5 mm Hg in SBP and DBP, respectively.
Our results contrast with those of the recent large, multicenter
Premier study (16). In that study, untreated subjects with mild
hypertension who were randomly assigned to the DASH intervention achieved decreases of 11 mm Hg in SBP and 7 mm Hg
in DBP with a 5-kg weight loss over 6 mo, and these decreases
were not significantly different from those seen in the established
care group, who had a similar weight loss.
Ours is the first study to assess the effect of a weight-loss
intervention on home BP measurements rather than on
investigator-measured BP. Some of the difference in BP response could be attributed to the different method of BP assessment. Home BP, measured at the same time of day, under the
same conditions, shows reduced variability. Home BP measurement is now emerging as a preferred method of measuring BP
(17), because it has been shown to share some of the advantages
of ambulatory BP, that is, to have no “white coat” effect (18), to
be more reproducible (19, 20), and to be more predictive of the
presence and progression of organ damage than are office or
clinic values (21). Because all our subjects used BP monitors for
which their data were downloaded directly to a computer by the
study staff, there was no possibility for errors in subject recording.
DIET, WEIGHT LOSS, AND BLOOD PRESSURE
The results of the present study clearly show that targeted
dietary advice (ie, to include 욷4 servings each of fruit and vegetables per day, 3 servings of nonfat dairy products per day, and
3 servings of fish and 1 serving of legumes per week and to avoid
butter and added salt) combined with advice to walk 욷0.5 h on
most days of the week resulted in a 5% weight loss, an 8 –mm Hg
decrease in SBP, and a 5–mm Hg decrease in DBP over 3 mo. In
addition, the study showed that a lifestyle intervention that can be
successfully implemented by obese or overweight free-living
individuals results in a greater decrease in BP than does the usual,
general dietary advice to reduce fat intake. The reason for the
increased efficacy of this diet over the low-fat diet with respect
to BP is not clear, but may be related to the increases in potassium
and calcium intakes.
CAN and AW were responsible for the conception and overall supervision
of the study. CAN was responsible for the drafting of the manuscript, critical
revision of the manuscript for intellectual content, and final approval of the
manuscript. CM, MKJ, SJG, and AB interviewed all participants, administered dietary counseling, and with AW contributed to drafting of the manuscript. SJG was responsible for biochemical processing and analysis. None of
the authors had any conflicts of interest.
8.
9.
10.
11.
12.
13.
14.
15.
16.
REFERENCES
1. MacMahon S, Cutler JA, Stamler J. Antihypertensive drug treatment.
Potential, expected, and observed effects on stroke and on coronary heart
disease. Hypertension 1989;13:I45–50.
2. Briganti EM, Shaw JE, Chadban SJ, et al. Untreated hypertension among
Australian adults: the 1999 –2000 Australian Diabetes, Obesity and Lifestyle Study (AusDiab). Med J Aust 2003;179:135–9.
3. Britt H, Miller GC, Knox S, et al. General practice activity in Australia
2002-03. AIHW cat. no. GEP 14. Canberra, Australia: Australian Institute of Health and Welfare, 2003 (General Practice Series no. 14).
4. Silventoinen K, Sans S, Tolonen H, et al. Trends in obesity and energy
supply in the WHO MONICA Project. Int J Obes Relat Metab Disord
2004;28:710 – 8.
5. Cameron AJ, Welborn TA, Zimmet PZ, et al. Overweight and obesity in
Australia: the 1999 –2000 Australian Diabetes, Obesity and Lifestyle
Study (AusDiab). Med J Aust 2003;178:427–32.
6. El-Atat F, Aneja A, McFarlane S, Sowers J. Obesity and hypertension.
Endocrinol Metab Clin North Am 2003;32:823–54.
7. Appel LJ, Moore TJ, Obarzanek E, et al. A clinical trial of the effects of
17.
18.
19.
20.
21.
989
dietary patterns on blood pressure. DASH Collaborative Research
Group. N Engl J Med 1997;336:1117–24.
Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low density lipoprotein cholesterol in plasma without use of
preparative ultracentrifuge. Clin Chem 1972;18:499 –502.
Nowson CA, Worsley A, Margerison C, et al. Blood pressure response
to dietary modifications in free-living individuals. J Nutr 2004;134:
2322–9.
Harada ND, Chiu V, King AC, Stewart AL. An evaluation of three
self-reported physical activity instruments for older adults. Med Sci
Sports Exerc 2001;33:962–70.
Whelton PK, He J, Cutler JA, et al. Effects of oral potassium on blood
pressure Meta-analysis of randomized controlled clinical trials. JAMA
1997;277:1624 –32.
Cappuccio FP, Elliott P, Allender PS, Pryer J, Follman DA, Cutler JA.
Epidemiologic association between dietary calcium intake and blood
pressure: a meta-analysis of published data. Am J Epidemiol 1995;142:
935– 45.
Jee SH, Miller ER 3rd, Guallar E, Singh VK, Appel LJ, Klag MJ. The
effect of magnesium supplementation on blood pressure: a meta-analysis
of randomized clinical trials. Am J Hypertens 2002;15:691– 6.
Allender PS, Cutler JA, Follmann D, Cappuccio FP, Pryer J, Elliott P.
Dietary calcium and blood pressure: a meta-analysis of randomized
clinical trials. Ann Intern Med 1996;124:825–31.
Neter JE, Stam BE, Kok FJ, Grobbee DE, Geleijnse JM. Influence of
weight reduction on blood pressure: a meta-analysis of randomized
controlled trials. Hypertension 2003;42:878 – 84.
Appel LJ, Champagne CM, Harsha DW, et al. Effects of comprehensive
lifestyle modification on blood pressure control: main results of the
PREMIER clinical trial. JAMA 2003;289:2083–93.
Guidelines Committee. European Society of Hypertension-European
Society of Cardiology guidelines for the management of arterial hypertension. J Hypertens 2003;21:1011–53.
Mancia G, Zanchetti A, Agabiti-Rosei E, et al. Ambulatory blood pressure is superior to clinic blood pressure in predicting treatment-induced
regression of left ventricular hypertrophy. SAMPLE Study Group. Study
on Ambulatory Monitoring of Blood Pressure and Lisinopril Evaluation
Circulation. 1997;95:1464 –70.
Sakuma M, Imai Y, Nagai K, et al. Reproducibility of home blood
pressure measurements over a one-year period. Am J Hypertens 1997;
10:798 – 803.
Stergiou GS, Baibas NM, Gantzarou AP, et al. Reproducibility of home,
ambulatory, and clinic blood pressure: implications for the design of
trials for the assessment of antihypertensive drug efficacy. Am J Hypertens 2002;15:101– 4.
Mule G, Caimi G, Cottone S, et al. Value of home blood pressures as
predictor of target organ damage in mild arterial hypertension. J Cardiovasc Risk 2002;9:123–9.