title: a mediterranean diet lowers blood pressure …...33 inflammatory response leading to...
TRANSCRIPT
1
Title: A Mediterranean diet lowers blood pressure and improves endothelial function: results
from the MedLey randomized intervention trial.
Authors: Mrs. Courtney R Davis, Professor Jonathan M Hodgson, Professor Richard
Woodman, Associate Professor Janet Bryan, Professor Carlene Wilson, Dr. Karen J Murphy
Affiliations:
Alliance for Research in Exercise, Nutrition and Activity, School of Health Sciences,
University of South Australia (CRD, KJM)
School of Psychology, Social Work and Social Policy, University of South Australia (JB)
School of Medical and Health Sciences, Edith Cowan University and School of Medicine and
Pharmacology, University of Western Australia (JMH)
Flinders Centre for Innovation in Cancer, School of Medicine, Flinders University (CW)
Flinders Centre for Epidemiology and Biostatistics, Flinders University (RW)
Authors’ last names: Davis, Hodgson, Woodman, Bryan, Wilson, Murphy
Corresponding Author:
Courtney R Davis
Alliance for Research in Exercise, Nutrition and Activity
University of South Australia
GPO Box 2471
Telephone: +61 8 8302 1869
Email: [email protected]
Source of support: National Health and Medical Research Council grant Running head: Mediterranean diet, blood pressure and FMD Abbreviations: CVD = cardiovascular disease, MedDiet = Mediterranean diet, BP = blood
pressure, BMI = body mass index, Med = Mediterranean diet group, Hab = habitual diet
group, FMD = flow mediated dilatation, WFRs = weighed food records
2
Clinical Trial Registry: The trial was registered with the Australia New Zealand
Clinical Trials Registry, ACTRN12613000602729,
http://www.anzctr.org.au/TrialSearch.aspx?searchTxt=ACTRN12613000602729&isBasic=Tr
ue
3
Abstract 1
Background: Consumption of a Mediterranean diet (MedDiet) is associated with lower risk 2
of cardiovascular disease. However, its impact on blood pressure and endothelial function is 3
not clear. 4
Objective: To determine the effect of adhering to a MedDiet for 6 months on blood pressure 5
and endothelial function in older, healthy Australians. 6
Design: 166 men and women aged >64 y were allocated by minimization to consume either a 7
MedDiet (n=85) or their habitual diet (HabDiet; control, n=81) for 6 months. The MedDiet 8
comprised mainly plant foods, abundant extra virgin olive oil and minimal red meat and 9
processed foods. 152 participants commenced the study, and 137 completed the study. Home 10
blood pressure was measured on 5 consecutive days (morning, afternoon and evening) at 11
baseline (n=149), and at 3 and 6 months. In a subset of participants (n=82), endothelial 12
function was assessed by using flow-meditated dilatation at baseline and 6 months. Dietary 13
intake was monitored using 3-day weighed food records completed at baseline, 2 and 4 14
months. Data were analyzed using linear mixed effects models to determine adjusted 15
between-group differences. 16
Results: The MedDiet adherence score increased significantly in the MedDiet group, but not 17
in the HabDiet group (P<0.001). A MedDiet, compared with the HabDiet, resulted in lower 18
systolic blood pressure (P=0.02, for diet*time interaction): -1.3 (95%CI -2.2, -0.3) mmHg, 19
P=0.008 at 3 months, and -1.1 (95%CI -2.0, -0.1) mmHg, P=0.03 at 6 months. Diastolic 20
blood pressure was not significantly different between groups (P=0.14, for diet*time 21
interaction). At 6 months the percentage flow-meditated dilatation was higher by 1.3% (95% 22
CI 0.2, 2.4 P=0.026) in the MedDiet group compared with the HabDiet group. 23
4
Conclusion: Australian men and women were able to follow a MedDiet for 6 months. This 24
resulted in a small but significantly lower systolic blood pressure, and improved endothelial 25
function. 26
Key words: Mediterranean diet, blood pressure, flow mediated dilatation, Australia, 27
cardiovascular disease28
5
Introduction 29
Cardiovascular disease (CVD) is the leading cause of death worldwide (1), and hypertension 30
is the primary risk factor for CVD and all-cause mortality (2). Endothelial function is 31
affected early in the development of vascular disease and hypertension (3), impaired by an 32
inflammatory response leading to atherosclerosis (4). Adopting a more prudent dietary 33
pattern may significantly reduce the burden of CVD via benefits on blood pressure (BP) and 34
endothelial function (5). 35
Evidence is accumulating that a Mediterranean dietary pattern may provide substantial 36
benefit to the risk of CVD (5-12). The principal components of the Mediterranean diet 37
(MedDiet) include vegetables, fresh fruits, fish and seafood, legumes, nuts, extra virgin olive 38
oil and red wine (6). The diet is low in discretionary foods and red meat, and moderate in 39
dairy foods, including mostly cheese and yoghurt. Similar dietary patterns have been 40
associated with lower risk of CVD in other parts of the world (7-9). 41
The MedDiet may mediate its effects partly though maintenance of BP and endothelial 42
function (10). Consumption of a diet high in fruits, vegetables, nuts and unsaturated oils, and 43
low in sodium can lower BP (7, 11). In addition, a number of components of a Mediterranean 44
dietary pattern have been shown to improve endothelial function (12-14). However, few 45
studies have explored the impact of increased adherence to a Mediterranean dietary pattern 46
on BP (15-17) or endothelial function (18-21). 47
Therefore, we have conducted a randomized, controlled intervention trial to assess the impact 48
of increased adherence to a Mediterranean dietary pattern for 6 months on BP and endothelial 49
function in an older Australian population. 50
6
Subjects and Methods 51
The MedDiet for cognitive and cardiovascular health in the elderly (MedLey) study protocol 52
has been described elsewhere (22, 23). Briefly, the study was a randomized controlled 53
intervention trial, with two parallel groups. Volunteers were stratified by gender, body mass 54
index (BMI) and age by the process of minimization [19] into either the MedDiet group 55
(MedDiet) or the habitual diet (HabDiet) control group. Minimization was done by an 56
investigator not involved in data collection after enrolment. Australian adults aged >64 years 57
were recruited from metropolitan Adelaide, who were free of any cardiovascular, liver, 58
kidney, respiratory or gastrointestinal disease, cognitive impairment, type I or II diabetes, 59
malignancy in the past 6 months, significant recent head trauma or significant psychiatric 60
disorder. Participants with BP over 160/100mmHg were excluded. A total of 152 volunteers 61
commenced the study (MedDiet = 80, HabDiet = 72) between August 2013 and September 62
2014. One hundred and thirty seven participants had completed the study by February 2015, 63
resulting in a withdrawal rate of 10% after study commencement. Complete baseline home 64
BP data were available for 149 participants. Complete baseline FMD data were collected for 65
82 participants. Figure 1 shows the flow of participants throughout the study for collection of 66
BP and FMD data. Two participants withdrew from the study because they were unable to 67
comply with the intervention diet; all other withdrawals were unrelated to the study. 68
The study was conducted by means of 6 major clinic visits and 10 minor clinic visits. All 69
visits took place at the Sansom Institute for Health Research, University of South Australia. 70
The 6 major visits comprised 2 separate visits (A and B) spaced 7 days apart, attended after 71
an 8 hour fast at baseline, 3 and 6 months. Primary outcome measures included home-72
measured BP and endothelial function measured by flow mediated dilatation (FMD). Blood 73
pressure was assessed between visits A and B at each time point, and FMD was measured at 74
7
baseline and 6 months at clinic visit A. Participants commenced their allocated diets the day 75
after baseline visit B. The 10 minor visits comprised 30 minutes sessions with a study 76
dietitian to ensure adherence was maintained. These were held between major visits (at 77
weeks 3, 5, 7, 9, 11 and at weeks 15, 17, 19, 21, and 23). Dietary intake was assessed by 3-78
day weighed food records (WFR) at baseline between major visits A and B, and between 79
weeks 9-11 and week 19-21, completed between minor visits (22). Body mass (to the nearest 80
0.1 kg) and height (to the nearest 0.1 cm) were assessed with participants wearing light 81
hospital gowns and used to determine BMI (mass/height in meters squared). Demographic 82
information was collected at baseline via survey. 83
Intervention arms 84
At baseline visit B, the participants attended a 45 minute consult with a trained dietitian to 85
receive their dietary instruction. The characteristics of the two dietary arms are shown in 86
Table 1. Those in the HabDiet group were asked to continue consuming their regular diets 87
throughout the trial, while those in the MedDiet group were asked to follow an adapted (for 88
Australia) MedDiet. For those in the MedDiet group, resources were provided including a 89
recipe book, guidelines for eating out, serving sizes and recommended number of servings. 90
Participants also received foods (olive oil, nuts, legumes, tuna and Greek yoghurt) to increase 91
the likelihood of adherence (22). Those in the HabDiet received verbal instructions and a 92
voucher to buy regularly consumed foods from supermarkets. Both groups were required to 93
maintain physical activity, medications and dietary supplements throughout the intervention. 94
Adherence 95
Participants in both dietary groups completed 3-day WFRs at baseline, 2 and 4 months. Data 96
was entered into FoodWorks Professional (Version 7.0.3016, Xyris Software Australia) and 97
the average daily intake of foods was calculated. All foods consumed were characterized into 98
8
one of 15 groups (Supplemental Table 1), and the amount in grams was calculated and 99
adjusted for total energy intake (g/MJ). We then calculated the group-specific mean intake 100
for 15 food groups at baseline. For baseline, 2 and 4 months data, each participant’s intake 101
was compared to their relevant group baseline mean and scored either 1 or 0 points for each 102
group (see Supplemental Table 1). All participants were scored in this way, whether allocated 103
to the control or intervention group. Therefore those in the HabDiet group were not expected 104
to change their adherence scores from baseline, while those in the MedDiet group were 105
expected to increase their adherence to a high level. Low adherence was considered <4.9 106
points, medium from 5-9.9, and high from 10-15 points. 107
Blood pressure 108
Blood pressure and heart rate were monitored at home by the participants for 6 consecutive 109
days using the A&D Company Ltd. digital BP monitor (model UA-767). Data from the first 110
day were not used in the analysis. Participants measured their BP after a 5 minute resting 111
period in the seated position in the morning, during the afternoon and in the evening. 112
Morning readings were taken 15-30 minutes after waking in a fasted state. Afternoon and 113
evening measures were taken at least 1 hour after consumption of caffeine, and at least 30 114
minutes after consumption of food and completing physical activity. At each of these time 115
points, three consecutive measures of BP and heart rate were recorded, each reading spaced 1 116
minute apart. The second and third measures were used in the analysis (22). 117
Flow mediated dilatation (FMD) 118
For analysis of FMD, the right-side brachial artery was assessed using the General Electric 119
Logiq 5 Expert ultrasound machine (22). Participants were rested supine for at least 10 120
minutes before commencement, in a temperature controlled clinic room (22oC). All women 121
were post-menopausal (age >64 years). The right arm was extended horizontally on the same 122
9
vertical plane as the heart and a BP cuff was placed around the forearm. After palpating for 123
the pulse, the brachial artery was located using a two-dimensional B-mode 12MHz transducer 124
at a depth of 2 cm. Three baseline recordings were taken with a 5 second interval in between, 125
totaling 15 seconds. Following this, the cuff was inflated to 200 mmHg for 5 minutes. 126
Immediately at deflation and for 3 minutes after, the image was recorded as a 10 second clip 127
until 5.15, (5.05-5.15), then 15 second clips thereafter until 8:00 minutes (5.15, 5.30, 5.45, 128
6.00, 6.15, 6.30, 6.45, 7.00, 7.15, 7.30, 7.45). The electrocardiogram was recorded with all 129
clips. Analysis of FMD was completed using the Vascular Tools for Research software 130
package, Brachial Analyzer for Research with DICOM module (Medical Imaging 131
Assessment, version 6.1.3) (22). It was not possible to blind the analysis as the image files 132
contained participant ID information. For each of the 15 clips, artery diameter was measured 133
for end-diastole frames only. Between 1 and 5 end-diastole frames were selected, towards the 134
beginning of the recording (with the first 3 cardiac cycles) wherever possible, and averaged 135
together. The equation to determine % FMD expansion for each frame was as follows: 136
% 𝐹𝐹𝑀𝑀𝑀𝑀 =(𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 − 𝑏𝑏𝑑𝑑𝑏𝑏𝑑𝑑𝑏𝑏𝑑𝑑𝑏𝑏𝑑𝑑 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑)
𝑏𝑏𝑑𝑑𝑏𝑏𝑑𝑑𝑏𝑏𝑑𝑑𝑏𝑏𝑑𝑑 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑× 100
To determine baseline diameter the three baseline means were averaged together. Due to the 137
difficulties associated with obtaining high quality FMD data through potential issues with 138
equipment, operator skill or volunteer discomfort, our team developed a system for rating the 139
quality of each FMD scan at baseline. Quality of the scan was primarily determined by 140
visibility of the intima or media layers and consistent location along the brachial artery 141
between baseline and post-deflation measures, with consideration given to volunteer 142
discomfort. A score from 0-5 was given to each baseline scan immediately upon completion. 143
Only those participants with ratings of 3.5/5 or higher at baseline had repeat measures and 144
were included in the analysis (see Figure 1) (22). 145
10
Ethics 146
The study was conducted according to the guidelines laid down in the Declaration of Helsinki 147
and all procedures involving human participants were approved by the University of South 148
Australia Human Research Ethics Committee, 22 June 2013, ethics approval number #31163. 149
Statistics 150
The primary analysis was a modified intent-to-treat analysis. The population was defined as 151
participants for whom there were complete baseline data for home BP who were randomly 152
assigned to the study groups (n=149). The reason for withdrawal or missing data was 153
associated with the assigned treatment in two participants only. Therefore, missing data for 154
participants who were not included in the final analysis were assumed to be missing at 155
random. Baseline characteristics of withdrawals and completers were compared using an 156
independent samples t test. Descriptive statistics are presented as means ± SDs. Categorical 157
variables were summarized by number in each category. A type-1 error rate of P<0.05 was 158
the level of significance used for all hypothesis testing. Log transformation was performed on 159
variables that were not normally distributed as assessed by using normal probability plots. At 160
baseline, characteristics of participants in the 2 groups were compared by using the 161
independent-samples t test on transformed data (natural logarithm, loge) when appropriate 162
and the chi-square test for categorical variables. Baseline, 3 and 6 month values for BP and 163
heart rate, FMD, and adherence, and between-group differences are presented as least-164
squares means and 95% confidence intervals in tables and least-squares means and standard 165
errors in figures. 166
Outcomes were analyzed by using linear mixed models in STATA (Version 13.0, StataCorp, 167
Texas, USA). For home BP, fixed effects in the models included time (baseline, 3 and 6 168
months; modeled as 3 separate binary indicator variables), diet group (MedDiet and 169
11
HabDiet), day (5 days: day 2 to day 6), time of day (morning, afternoon and evening) and 170
reading number (2 readings: reading 2 and reading 3) and a diet*time interaction term. The 171
overall effect of the MedDiet was established by using the global diet*time interaction term. 172
Where a significant diet*time interaction was observed, we evaluated differences between 173
diets at 3 and 6 months (2 and 4 months for adherence scores). For BP and heart rate, the 174
primary analysis used combined data from morning, afternoon and evening measures. In 175
secondary analyses we then further analyzed each time point (morning, afternoon and 176
evening) separately. Previously, we used independent t tests (continuous variables) and chi 177
squared tests (categorical variables) to compare baseline characteristics of those who had 178
baseline FMD scans and those who did not (22). The was no relationship for age, diet group, 179
gender, BMI, waist to hip ratio, HDL cholesterol; only total and LDL cholesterol were 180
highest amongst those without scans. FMD was evaluated at baseline and at 6 months and 181
was assessed using a linear mixed effects model with FMD at 6 months as the outcome 182
variable. The fixed effects in the models included baseline values (at 0 months), diet group 183
(MedDiet and HabDiet), and time post cuff deflation (0 to 165 with 15 second intervals). The 184
overall effect of the MedDiet on FMD was established using the difference in the overall 185
predicted marginal means for the 2 diets (taken across the 165 second recording period). 186
Results 187
Baseline characteristics are shown in Table 2. According to the 15-point score, both groups 188
had medium level adherence to the MedDiet at baseline. Within groups, completers were 189
similar in terms of baseline characteristics to non-completers, with the exception of insulin 190
which was higher in non-completers than completers in the control group (data not shown). 191
BMI decreased significantly in both groups by 6 months (mean change -0.236, 95% CI -192
0.431,-0.041, P=0.012 for the MedDiet, -0.224, 95% CI -0.428,-0.021, P =0.025 for the 193
12
HabDiet), however changes in BMI were not different between groups. Servings of foods and 194
nutrient intakes for each group at baseline and at 4 months are shown in Table 3. Amongst 195
the MedDiet group, olive oil, vegetables, fruit, bread, milk and yogurt, red wine, legumes, 196
fish and nut consumption increased while meat consumption decreased. In the control group, 197
red wine and fish intake decreased, otherwise dietary intake remained similar. Energy from 198
total, monounsaturated and polyunsaturated fats increased, and energy from saturated fat 199
decreased amongst the MedDiet group. There were also increased intakes of vitamin E and 200
long chain omega-3 fatty acids. There were no changes in nutrient intake amongst the 201
HabDiet group. 202
Adherence to a Mediterranean diet 203
Adherence increased by 3.4 points in the MedDiet (P<0.001) and by 0.7 points in the 204
HabDiet (P=0.652) after 4 months (Figure 2). Adherence was not statistically significantly 205
different between groups at baseline, however the differences between groups at 2 (2.1, 95% 206
CI 1.4, 2.8) and 4 months (2.5, 95% CI 1.7, 3.2) were significant (P<0.001), with an overall 207
significant diet*time interaction (P<0.001). 208
Home-measured blood pressure 209
During the 6 month intervention the overall effect of diet (diet*time interaction) was 210
significant for systolic BP (P=0.02) and heart rate (P=0.03), but not for diastolic BP (P=0.14) 211
(Table 4). Systolic BP was significantly lower for MedDiet compared with HabDiet at 3 212
months and 6 months. These BP differences were primarily due to differences in morning and 213
afternoon values, with no effect on systolic BP in the evening (Table 5). The overall effect of 214
diet on heart rate was primarily driven by a higher heart rate at 3 months with the MedDiet 215
compared with HabDiet. 216
Flow mediated dilatation (FMD) 217
13
There was no significant difference in FMD between groups at baseline (P=0.99). In addition 218
to those with poor quality scans, an additional 4 participants who did not complete the study 219
were not included in the analysis (MedDiet=1, HabDiet=3). The effect of diet on FMD 220
overall was significant (P<0.001). Mean FMD at 6 months was 2.5% (95% CI 1.8, 3.2) in the 221
MedDiet group, and 1.2 (95% CI 0.4, 2.0) in the HabDiet group. The mean baseline-adjusted 222
difference between groups at 6 months versus baseline was 1.3% (95% CI: 0.2, 2.4) %, 223
P=0.03). Figure 3 shows the observed FMD values from 0 and 165 seconds post cuff 224
deflation compared for each diet group. 225
Discussion 226
In a randomised, controlled intervention trial involving elderly Australian adults, adhering to 227
a MedDiet for 6 months resulted in lower systolic BP and a higher FMD, compared to the 228
habitual Australian diet. Increased adherence scores and changes in food and nutrient intake 229
suggest successful adoption of Mediterranean dietary principles amongst the MedDiet group. 230
The increased adherence is comparable to the increased adherence observed in the 231
Prevención con dieta Mediterránea (PREDIMED) study, whereby those in the intervention 232
groups improved their scores by 2 points on a 14-point scale compared to no change in the 233
low fat control group (24). Other MedDiet studies generally demonstrate good adherence, 234
indicating that the diet is palatable and satiating (25). Participants in our trial were assisted by 235
fortnightly sessions with a dietitian, which has been shown to enhance adherence in other 236
studies (26). Although there were more withdrawals in the MedDiet group, only two were 237
due to inability to comply with the MedDiet. Other reasons for withdrawal included health 238
issues, (n=4) family issues (n=6) and other commitment (n=3) unrelated to the study. Few 239
studies have assessed whether Western populations can adopt a MedDiet long term, however 240
our results suggest it is feasible (27, 28). 241
14
Our data show following a MedDiet results in a significant but small decrease in systolic BP 242
after 3 and 6 months, and no change in diastolic BP. Results from previous studies 243
investigating the relationship between the MedDiet and BP are ambiguous. Within the 244
PREDIMED study at 4 years follow-up, diastolic BP was lowered by 1.5 mmHg and 0.7 245
mmHg in the olive oil and walnut MedDiet intervention groups, respectively, compared to 246
the low-fat diet control, with no change in systolic BP (15). Two meta-analyses published 247
early in 2016 arrived at opposite conclusions. The first included studies at least 12 months 248
long that had a low fat control group. Although there was a significant reduction found for 249
both systolic and diastolic BP, the authors remained unconvinced the MedDiet lowers BP 250
compared to the low fat diet, mainly due to the heterogeneity and limited number of studies 251
(n=7) (29). Another meta-analysis investigated the effects of different dietary patterns and 252
concluded that a MedDiet lowers systolic BP by 3 mmHg and diastolic BP by 2 mmHg (30). 253
Interestingly, of only 5 MedDiet pattern studies identified for this analysis, two found no 254
significant effect on BP, but were excluded from the analysis, so the result was likely 255
skewed. Further, although both meta-analyses included data from the PREDIMED study, the 256
first (negative conclusion) used data from 24 month follow-up, while the latter used 12 month 257
follow-up data (positive conclusion). Another consideration is whether blood pressure is 258
elevated at baseline among these populations. More than 80% of those recruited for the 259
PREDIMED study had hypertension at baseline (31), but BP was little affected after 4 years 260
(15). Sacks et al. (7) showed a healthy dietary pattern (DASH-diet) reduced BP amongst 261
those with BP above 120/80 mmHg, although this diet had some differences to the MedDiet. 262
Our research suggest a small benefit to systolic BP only, although we did not have a 263
hypertensive or at risk population. Clearly, more research is needed to target BP as a primary 264
outcome in both normotensive and hypertensive individuals to better understand the effects 265
of the MedDiet on BP. 266
15
In accordance with previous studies, our data support the notion that the MedDiet improves 267
endothelial function (14, 18, 19, 32). Rallidis et al. (18) found a 2% improvement in FMD 268
amongst 90 abdominally obese Greek adults (mean age 50.4 y) after 2 months following a 269
MedDiet. Serra-Majem et al. (33) reviewed evidence for the health benefits of the MedDiet 270
and included 8 trials which assessed endothelial function. Two trials found no change, one 271
found a walnut-supplemented diet superior to the MedDiet, and five trials found positive 272
effects of the diet on endothelial function (33). Populations studied were generally at higher 273
risk of CVD, and included secondary CVD, type 2 diabetes and metabolic syndrome patients, 274
as well as persons with hypercholesterolemia. 275
The mechanisms by which endothelial function is improved after consumption of a MedDiet 276
are thought to relate to its favorable fatty acid profile and high antioxidant content. A higher 277
proportion of monounsaturated fatty acids reduce susceptibility of cholesterol molecules to 278
oxidation (34-36). Free radicals can react with NO to produce peroxynitrite, diminishing its 279
vasodilatory effects (37). Antioxidants may inhibit the free radical damage which can initiate 280
and progress atherosclerosis, resulting in progressive loss of endothelial function (38). These 281
functions in turn would impact on the immune-driven cascade response to oxidized-LDL 282
within the sub-endothelial space (4). Our results suggest NO availability can be improved 283
amongst older adults with a MedDiet. The 1.3% FMD difference is likely to be clinically 284
significant. Fathi et al. (39) showed that amongst patients at risk of cardiovascular events, 285
those in the lowest tertile of FMD response (<2%) were significantly more likely to 286
experience an event over 24 months than those with FMD above 2%. In addition, a meta-287
analysis by Inaba et al. (40) found that an FMD higher by 1% at baseline was associated with 288
a 13% lower risk of a cardiovasuclar event after a mean follow-up of 3 years. 289
16
Our study has several limitations which should be considered. Only a subsample of 290
participants had FMD assessed at baseline and 6 months, and we did not include those who 291
withdrew in the FMD analysis (n=4). Endothelial function was only assessed in the fasting 292
state, thus we cannot conclude whether the function of the brachial artery was improved in 293
the post-prandial state. Although care was taken to instruct volunteers how to measure their 294
BP at home, we cannot be certain of the accuracy and conditions of the measurements. 295
However, having multiple BP measurements from across the day strengthens our study. 296
Dietary compliance was self-reported, which can be inherent with reliability issues, such as 297
under-reporting (41). This could have affected the relationship between adherence and 298
outcomes, particularly if those in the control group were altering habits towards a more 299
MedDiet without reporting this change. However WFRs are generally considered the ‘gold 300
standard’ method for collecting dietary intakes, and adherence score changes were consisted 301
at both 2 and 4 months (42, 43). This study has several other key strengths: it was a 302
controlled trial with balanced groups at baseline, with a habitual diet comparator rather than a 303
low fat or alternative dietary control. We had close to complete home-measured BP from 304
three time points during the day. At-home measures have been shown to be more sensitive 305
than measures performed in the clinic/office, and can remove the effects of white coat 306
hypertension (44). To our knowledge, our study is the first to have used BP readings 307
averaged from 5 days of home-monitoring to determine changes after consumption of the 308
MedDiet. 309
In conclusion, healthy Australian adults aged >64 years were able to adhere to a traditional 310
MedDiet, which resulted in reduced systolic BP after 3 and 6 months and improved 311
endothelial function as measured by FMD after 6 months. The demonstrated ability to adhere 312
to the diet for 6 months, and the resulting improvements in CVD risk factors suggests this 313
diet may be a useful strategy to implement in Australia to help reduce the impact of CVD. 314
17
Further study is needed to confirm the impact of the MedDiet on BP and hard CVD endpoints 315
with a controlled intervention. 316
Acknowledgements 317
The authors thank Alissa Knight for contributions to the cognitive study design and data 318
collection; Louise Massie, Sansom Clinical Trials Coordinator; Dr. Catherine Yandell and 319
Mark Cutting for assistance with data collection, trial assistance and sample handling; Crystal 320
Grant, Kathryn Dyer and Nerylee Watson for assistance with data collection and sample 321
handling; dietitians Natalie Blanch and Kristina Petersen who conducted dietetic 322
consultations; Cobram Estate™ for providing Australian extra virgin olive oil; Peanut 323
Company of Australia for providing hi-oleic dry roast peanuts; The Grains & Legumes 324
Nutrition CouncilTM for sourcing legumes for the trial, Simplot Australia Pty. Ltd for 325
providing Edgell canned legumes and John West tuna; Goodman Fielder Ltd for providing 326
canola oil; and The Almond Board of Australia for providing almonds. Jonathan Hodgson is 327
supported by a National Health and Medical Research Council Senior Research fellowship. 328
Conflict of Interest 329
None 330
Author Contributions 331
KJM, JB, CW, JMH and CD designed research. CD conducted research. JMH, RW and CD 332
analyzed data. CD, KJM and JMH wrote the paper. CD had primary responsibility for final 333
content. All authors read and approved the final manuscript. 334
18
References
1. World Health Organisation (WHO). Global status report on non-communicable
diseases 2014. WHO, Geneva, Switzerland, 2014.
2. Ezzati M, Riboli E. Behavioral and dietary risk factors for noncommunicable
diseases. N Engl J Med 2013;369(10):954-64. doi: 10.1056/NEJMra1203528.
3. Charakida M, Masi S, Lüscher TF, Kastelein JJ, Deanfield JE. Assessment of
atherosclerosis: the role of flow-mediated dilatation. Eur Heart J 2010;31:2854-61.
doi: 10.1093/eurheartj/ehq340.
4. Ross R. Mechanisms of disease: Atherosclerosis - An inflammatory disease. N Engl J
Med 1999;340(2):115-26.
5. Yeboah J, Crouse JR, Hsu F-C, Burke GL, Herrington DM. Brachial flow-mediated
dilation predicts incident cardiovascular events in older adults: The cardiovascualr
health study. Circulation 2007;115:2390-7. doi:
10.1161/CIRCULATIONAHA.106.678276.
6. Davis CR, Bryan J, Hodgson J, Murphy KJ. Definition of the Mediterranean diet: A
literature review. Nutrients 2015;7:9139-53. doi: 10.3390/nu7115459.
7. Sacks FM, Svetkey LP, Vollmer WM, Appel LJ, Bray GA, Harsha D, Obarzanek E,
Conlin PR, Miller ER, Simons-Morton DG, et al. Effects on blood pressure of
reduced dietary sodium and the dietary approaches to stop hypertension (DASH) diet.
N Engl J Med 2001;344(1):3-10.
8. Reddy KS, Katan MB. Diet, nutrition and the prevention of hypertension and
cardiovascular diseases. Pub Health Nutr 2004;7(1A):167-86. doi:
10.1079/PHN2003587.
19
9. Appel LJ, Moore TJ, Obarzanek E, Vollmer WM, Svetkey LP, Sacks FM, Bray GA,
Vogt TM, Cutler JA, Windhauser MM, et al. A clinical trial of the effects of dietary
patterns on blood pressure. N Engl J Med 1997;336(16):1117-24.
10. Pérez-López FR, Fernández-Alonso A, Chedraui P, Simoncini T. Mediterranean
lifestyle and diet: deconstructing mechanisms of health benefits 1st. ed. In: Bioactive
Food as Dietary Intervention for the Aging Population. United Kingdom and United
States of America: Academic Press, 2013:129-38.
11. Estruch R, Ros E, Salas-Salvadó J, Covas M-I, Corella D, Arós F, Gómez-Gracia E,
Ruiz-Gutiérrez V, Fiol M, Lapetra J, et al. Primary prevention of cardiovascualr
disease with a Mediterranean diet. New Engl J Med 2013;368:1279-90. doi:
10.1056/NEJMoa1200303.
12. McCall DO, McGartland C, McKinley M, Sharpe P, McCance D, Young I, Woodside
J. The effect of increased dietary fruit and vegetable consumption on endothelial
activation, inflammation and oxidative stress in hypertensive volunteers.
2011;21:658-64.
13. Perona JS, Cabello-Moruno R, Ruiz-Gutierrez V. The role of virgin olive oil
components in the modulation of endothelial function. 2006;17:429-45. doi:
10.1016/j.jnutbio.2005.11.007.
14. Ros E, Núñez I, Pérez-Heras A, Serra M, Gilabert R, Casals E, Deulofeu R. A walnut
diet improves endothelial funciton in hypercholesterolemic subjects. A randomised
crossover trial. Circulation 2004;109:1609-14. doi:
10.1161/01.CIR.0000124477.91474.FF.
15. Toledo E, Hu FB, Estruch R, Buil-Cosiales P, Corella D, Salas-Salvadó J, Covas MI,
Aros F, Gomez-Gracia E, Fiol M, et al. Effect of the Mediterranean diet on blood
20
pressure in the PREDIMED trial: results from a randomized controlled trial. BMC
Med 2013;11(207).
16. Psaltopoulou T, Naska A, Ofanos P, Trichopoulos D, Mountokalakis T, Trichopoulou
A. Olive oil, the Mediterranean diet, and arterial blood pressure: the Greek European
Prospective Investigation into Cancer and Nutrition (EPIC) study. Am J Clin Nutr
2004;80:1012-8.
17. Tzima N, Pitsavos C, Panagiotakos DB, Skoumas J, Zampelas A, Chrysohoou C,
Stefanadis C. Mediterranean diet and insulin sensitivity, lipid profile and blood
pressure levels, in overweight and obese people; The Attica study. Lipids Health Dis
2007;6(22). doi: 10.1186/1476-511X-6-22.
18. Rallidis LS, Lekakis J, Kolomvotsou A, Zampelas A, Vamvakou GD, Efstathiou S,
Dimitriadis G, Raptis SA, Kremastinos DT. Close adherence to a Mediterranean diet
improves endothelial function in subjects with abdominal obesity. Am J Clin Nutr
2009;90:263-8.
19. Esposito K, Marfella R, Ciotola M, Di Palo C, Giugliano F, Guigliano G, D'Armiento
M, D'Andrea F, Giugliano D. Effect of a Mediterranean-style diet on endothelial
dysfunction and markers of vascular inflammation in the metabolic syndrome. JAMA
2004;292(12):1440-6.
20. Marin C, Ramirez R, Delgado-Lista J, Yubero-Serrano EM, Perez-Martinez P,
Carracedo J, Garcia-Rios A, Rodriguez F, Gutierrez-Mariscal FM, Gomez P, et al.
Mediterranean diet reduces endothelial damage and improves the regenerative
capacity of endothelium. Am J Clin Nutr 2011;93:267-74.
21. Fuentes F, Lopez-Miranda J, Sanchez E, Sanchea F, Paez J, Paz-Rojas E, Marin C,
Gomez P, Jimenez-Pereperez J, Ordovas JM, et al. Mediterranean and low-fat diets
21
improve endothelial function in hypercholesterolemic men. Ann Intern Med
2001;134:1115-9.
22. Davis CR, Bryan J, Hodgson J, Wilson C, Dhillon V, Murphy KJ. A randomised
controlled intervention trial evaluating the efficacy of an Australianised
Mediterranean diet compared to the habitual Australian diet on cognitive function,
psychological wellbeing and cardiovascular health in healthy older adults (MedLey
study): protocol paper. BMC Nutr 2015;1:35. doi: 10.1186/s40795-015-0033-7.
23. Knight A, Bryan J, Wilson C, Hodgson J, Murphy KJ. A randomised controlled
intervention trial evaluating the efficacy of a Mediterranean dietary pattern on
cognitive function and psychological wellbeing in healthy older adults: the MedLey
study. BMC Geriatr 2015;15(55). doi: 10.1186/s12877-015-0054-8.
24. Martínez-González MA, Salas-Salvadó J, Estruch R, Corella D, Fito M, Ros E.
Benefits of the Mediterranean diet: Insights from the PREDIMED study. Prog
Cardiovasc Dis 2015;58:50-60.
25. Willett WC, Sacks F, Trichopoulou A, Drescher G, Ferro-Luzzi A, Helsing E,
Trichopoulos D. Mediterranean diet pyramid: a cultural model for healthy eating. Am
J Clin Nutr 1995;61(supplementary):1402S-6S.
26. Zazpe I, Sanchez-Tainta A, Martinez-Gonzalez MA, Lamuela-Raventos RM,
Schröder H, Estruch R, Salas-Salvado J, Corella D, Fiol M, Gomez-Gracia E, et al. A
large randomized individual and group intervention conduction by registered
dietitians increased adherence to Mediterranean-type diets: the PREDIMED study. J
Am Diet Assoc 2008;108(7):1134-44. doi: 10.1016/j.jada.2008.04.011.
27. Itsiopoulos C, Brazionis L, Kaimakamis M, Cameron M, Best J, O'Dea K, Rowley K.
Can the Mediterranean diet lower HbA1c in type 2 diabetes? Results from a
22
randomized cross-over study. Nutr Metab Cardiovasc Dis 2011;21:740-7. doi:
10.1016/j.numecd.2010.03.005.
28. Davis CR, Bryan J, Hodgson JM, Wilson C, Murphy KJ. Older Australians can
adhere to a traditional Mediterranean style diet over two weeks: a pilot dietary
intervention study. BMC Nutr 2015;1(28). doi: 10.1186/s40795-015-0021-y.
29. Nissensohn M, Román-Vinas B, Sánchez-Villegas A, Piscopo S, Serra-Majem L. The
effect of the Mediterranean diet on hypertension: a systematic review and meta-
analysis. J Nutr Educ Behav 2016;48(1):42-53. doi:
doi.org/10.1016/j.jneb.2015.08.023.
30. Ndanuko RN, Tapsell LC, Charlton KE, Neale EP, Batterham MJ. Dietary patterns
and blood pressure in adults: a systematic review and meta-analysis of randomized
controlled trials. Adv Nutr 2016;7:76-89. doi: 10.3945/an.115.009753.
31. Martínez-Gonzáelz MÁ, Corella D, Salas-Salvadó J, Ros E, Covas MI, Fiol M,
Wärnberg J, Arós F, Ruíz-Gutiérrez V, Lamuela-Raventós RM, et al. Cohort Profile:
Design and methods of the PREDIMED study. Int J Epidemiol 2010;41:377-85. doi:
10.1093/ije/dyq250.
32. Søndergaard E, Møller JE, Egstrup K. Effect of dietary intervention and lipid-
lowering treatment on brachial vasoreactivity in patients with ischemic heart disease
and hypercholesterolemia. Am Heart J 2003;145(5).
33. Serra-Majem L, Roman B, Estruch R. Scientific evidence of interventions using the
Mediterranean diet: a systematic review. Nutr Rev 2006;64(2):S27-S47. doi:
10.1301/nr.2006.feb.S27–S47.
34. Hargrove RL, Etherton TD, Pearson TA, Harrison EH, Kris-Etherton PM. Low fat
and high monounsaturated fat diets decrease human low density lipoprotein oxidative
susceptibility in vitro. J Nutr 2001;131:1758-63.
23
35. Moreno JA, López-Miranda J, Pérez-Martínez P, Marín C, Moreno R, Gómez P,
Paniagua JA, Pérez-Jiménez F. A monounsaturated fatty acid-rich diet reduces
macrophage uptake of plasma oxidised low-density lipoprotein in healthy young men.
2008;100:569-75. doi: 10.1017/S0007114508911508.
36. Bonanome A, Pagnan A, Biffanti S, Opportuno A, Sorgato F, Dorella M, Maiorino
M, Ursini F. Effect of dietary monounsaturated and polyunsaturated fatty acids on the
susceptibility of plasma low density lipoproteins to oxidative modification.
Arterioscler Thromb 1992;12:529-33.
37. Franzini L, Ardigo D, Valtuena S, Pellegrini N, Del Rio D, Bianchi M, Scazzina F,
Piatta P, Brighenti F, Zavaroni I. Food selection based on high total antioxidant
capacity improves endothelial function in a low cardiovascular risk population.
2012;22:50-7.
38. Fiedor J, Burda K. Potential role of carotenoids as antioxidants in human health and
disease. Nutrients 2014;6:466-88. doi: 10.3390/nu6020466.
39. Fathi R, Haluska B, Isbel N, Short L, Marwick TH. The relative importance of
vascular structure and function in predicting cardiovascular events. J Am Coll Cardiol
2004;43(4):616-23. doi: 10.1016/j.jacc.2003.09.042.
40. Inaba Y, Chen JA, Bergmann SR. Prediction of future cardiovascular outcomes by
flow mediated vasodilation of brachial artery: a meta-analysis. Int J Cardiovasc
Imaging 2010;26:631-40. doi: 10.1007/s10554-010-9616-1.
41. Livingstone MBE, Prentice A, Strain J, Coward W, Black A, Barker M, McKenna P,
Whitehead R. Accuracy of weighed dietary records in studies of diet and health. BMJ
1990;300:708-12.
24
42. Bonifacj C, Gerber M, Scali J, Daures J. Comparison of dietary assessment methods
in a Southern French population: use of weighed food records, estimated-diet records
and a food-frequency questionnaire. 1997;51:217-31.
43. Hodge AM, Patterson AJ, Brown WJ, Ireland P, Giles GG. The Anti Cancer Council
of Victoria FFQ: relative validity of nutrient intakes compared with weighed food
records in young to middle-aged women in a study of iron supplementation.
2000;24(6):576-83.
44. Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, Jones DW, Kurtz T,
Sheps SG, Roccella EJ. Recommendations for blood pressure measurement in
humans and experimental animals. Part 1: Blood pressure measurement in humans: a
statement for professionals from the subcommittee of professional and public
education of the American Heart Association council on high blood pressure research.
Hypertension 2005;45:142-61. doi: 10.1161/01.CIR.0000154900.76284.F6.
25
TABLE 1.
Characteristics of the habitual diet (control) and Mediterranean diet intervention for the
Mediterranean diet for cognitive and cardiovascular health in the elderly study
Habitual diet Mediterranean diet
Duration 6 months (24 weeks) 6 months (24 weeks)
Characteristics Consume regular diet without change
Consume abundant extra virgin olive oil (≥1 Tbsp/day)
Seasonal variation permitted
Consume 5-6 serves vegetables/day
Deviations for special occasions/holidays permitted
Consume ≥2 serves fresh fruits/day
Do not begin regularly consuming a new food, or cease consumption of normally consumed food during the 6 months (e.g. swap type of oil used in regular cooking)
Consume 4-6 serves wholegrain cereals/day
Consume 4-6 serves nuts/week
Consume 3 serves legumes/week
Consume 3 serves fish (1 oily)/week
Consume <1 serve red meat/week
Consume Limit discretionary foods1 to ≤3 times per week
1Discretionary foods includes confectionary, chocolate, crisps, pastries, cakes, pies and other
bakery products, biscuits, deep-dried foods, non-red wine alcoholic beverages, and oils and
fats other than olive oil
26
TABLE 2.
Baseline sociodemographic and clinical characteristics of commencing participants in the
MedLey trial, by diet group1
MedDiet (n = 80) HabDiet (n = 69)
Age (years) 71.0 ± 4.9 70.9 ± 4.9
Gender 57.5% female 55.1% female
Country of birth2
Australia 63.8% (51) 43.1% (31)
Europe 28.8% (23) 47.2% (34)
Other 5.0% (4) 2.8% (2)
Level of education3
Tertiary 58.8% (47) 50.0% (36)
Other 38.8% (31) 43.1% (31)
BMI (kg/m2) 26.7 ± 3.7 27.1 ± 4.2
Systolic blood pressure (mmHg) 122.7 ± 16.0 125.9 ± 16.8
Diastolic blood pressure (mmHg) 69.2 ± 9.54 73.1 ± 11.5
Heart rate (beats/minute) 65.5 ± 9.5 67.6 ± 10.5
Flow-mediated dilatation (%)4 1.0 ± 2.6 1.1 ± 2.2
MedDiet adherence score5 7.1 ± 1.9 7.3 ± 2.4
1Mean±SD for continuous variables, % (count) for categorical variables
27
2Options included Australia, New Zealand, Asia, Middle East, Europe, Nth America,
Central America and Sub-Saharan Africa
3Completed education to one of the following: primary school, secondary school,
diploma, undergraduate, post graduate or other. Tertiary education included diploma,
undergraduate and post graduate completion.
4Mean of per cent flow mediated dilatation between 15 and 165 seconds post cuff
deflation. Based on subsample of population (MedDiet = 48, HabDiet = 34)
5Mediterranean diet adherence based on mean intake of 15 food groups (vegetables,
fruits, nuts, legumes, potatoes, bread, cereals, milk/yoghurt, cheese, eggs, meat and
meat alternatives, sugars, miscellaneous products, olive oil and red wine). Maximum
adherence = 15, minimum adherence = 0.
MedDiet, Mediterranean diet; HabDiet, Habitual diet group.
28
TABLE 3.
Daily servings of food and nutrient intake1 of older Australians following either a
Mediterranean diet or their habitual diets at baseline and after 4 months
MedDiet (n=80) HabDiet (n=70)
Baseline 4 months Baseline 4 months
Olive oil 0.2±0.4 1.9±1.0 0.3±0.4 0.3±0.6
Vegetables 2.9±2.5 3.5±1.7 2.8±1.6 2.7±1.6
Fruit 1.8±1.0 2.6±1.3 1.8±1.0 1.9±1.0
Bread 1.8±1.0 2.1±1.2 2.0±1.1 1.8±1.1
Cereal 1.2±1.1 1.1±0.8 1.1±0.9 1.1±1.1
Cheese 0.4±0.6 0.4±0.4 0.6±0.6 0.5±0.7
Milk/yoghurt 1.0±0.7 1.4±0.6 1.0±0.9 0.9±0.8
Red wine 0.8±1.4 1.0±1.1 0.9±1.5 0.5±1.0
Nuts 0.6±1.1 1.7±1.8 0.5±0.6 0.5±0.7
Legumes 0.3±0.6 0.6±0.8 0.2±0.4 0.2±0.5
Meat2 1.1±0.8 0.8±0.6 1.3±0.9 1.4±0.8
Fish3 0.4±0.5 0.8±0.5 0.5±0.5 0.4±0.5
Energy (kcal) 2088.6±498.3 2122.1±479.3 2104.3±492.6 1994.0±499.9
CHO (% en) 42.5±6.9 38.2±5.7 41.6±6.7 40.9±7.0
Protein (% en) 19.2±3.5 19.2±3.0 19.2±3.2 18.9±3.6
Total fat (% en) 33.6±6.0 38.5±7.4 34.5±5.4 35.7±5.6
MUFA (% en) 13.3±3.5 19.5±4.7 13.2±3.3 14.3±3.6
PUFA (% en) 5.7±2.3 7.3±1.9 5.7±1.6 5.6±2.0
SFA (% en) 12.±2.8 9.0±1.8 12.9±2.9 13.1±3.3
LC n-3 FA 449.5±513.7 783.8±606.2 480.6±527.2 300.0±309.7
Fibre 30.4±8.8 34.2±12.6 28.2±8.2 25.7±7.6
29
Vitamin C 147.7±82.3 166.4±78.1 135.5±70.8 119.8±75.0
Vitamin E 11.0±6.3 17.3±5.9 10.9±4.3 10.7±4.9
Total folates 453.8±160.1 484.0±154.9 428.1±168.7 389.9±146.3
Calcium 926.9±338.6 930.2±276.1 930.7±352.7 861.2±344.6
Iron 13.8±4.2 13.7±3.6 12.8±3.7 11.9±3.4
1Data are presented as mean±SD. Standard serving sizes based on the Australian Guide to
Healthy Eating 2013, with the exception of olive oil where 1 serve = 18 ml. Serving sizes for
fruits, bread, cheese, red wine, nuts, meat and fish were 150 g, 40 g, 40 g, 100 ml, 30 g, 80 g
and 100 g respectively. For dairy (225 g) and cereals (60 g) an average of different foods was
used. For legumes a serving size of 75 g was used.
2Meat included all red and white meats, and excluded small goods such as ham, salami and
bacon
3Fish included all fresh and canned fish and all seafood excluding deep fried.
MedDiet, Mediterranean diet; HabDiet, Habitual diet; en, energy; LC n-3 FA, long chain
omega-3 fatty acids.
30
TABLE 4.
Blood pressure and heart rate at baseline, 3 and 6 months, and baseline-adjusted mean differences between older Australians following a
Mediterranean diet and habitual diet1
MedDiet (n = 80) HabDiet (n = 69) Adjusted difference P-value
Systolic blood pressure (mmHg)
Baseline 122.8 (120.1, 125.5) 125.9 (123.0, 128.8)
3 months 117.8 (115.1, 120.6) 122.2 (119.3, 125.1) -1.3 (-2.2, -0.3) 0.01
6 months 116.6 (113.8, 119.3) 120.7 (117.8, 123.6) -1.1 (-2.0, -0.1) 0.03
Diet*time interaction 0.02
Diastolic blood pressure (mmHg)
Baseline 69.4 (67.7, 71.0) 73.2 (71.4, 75.0)
3 months 67.8 ( 66.1, 69.4) 71.2 (69.5, 73.0) 0.4 (-0.23, 0.96,) 0.23
6 months 67.6 (65.9, 69.2) 70.8 (69.0, 72.6) 0.6 (-0.1, 1.2) 0.05
Diet*time interaction 0.14
Heart rate (beats/minute)
Baseline 65.5 (63.7, 67.2) 67.8 (65.9, 69.6)
3 months 65.7 (64.0, 67.5) 67.3 (65.5, 69.2) 0.7 (0.2, 1.3) 0.01
31
6 months 65.6 (63.8, 67.3) 67.7 (65.9, 69.6) 0.1 (-0.4, 0.7) 0.62
Diet*time interaction 0.03
1Data are presented as mean±95% confidence intervals. Assessed using a mixed effects model with Diet, Time and a Diet x Time interaction
term as fixed effects, and subject ID as a random intercept. Adjusted differences are the predicted marginal means for each visit compared to
baseline.
MedDiet, Mediterranean Diet; HabDiet, Habitual Diet
32
TABLE 5.
Morning, afternoon and evening systolic blood pressure and adjusted mean differences between older Australians following a Mediterranean diet
and habitual diet1
MedDiet (n = 80) HabDiet (n = 69) Adjusted difference P-value
Morning systolic blood pressure (mmHg)
Baseline 123.7 (120.6, 126.8) 126.5 (123.2, 129.9)
3 months 119.2 (116.0, 122.4) 123.5 (120.1, 126.9) -1.5 (-2.8, -0.1) 0.04
6 months 117.6 (114.4, 120.7) 122.6 (119.2, 126.0) -2.2 (-3.6, -0.8) <0.001
Diet*time interaction 0.01
Afternoon systolic blood pressure (mmHg)
Baseline 121.6 (119.0, 124.1) 125.1 (122.4, 127.9)
3 months 116.1 (113.5, 118.6) 121.6 (118.9, 124.4) -2.0 (-3.5, -0.5) 0.01
6 months 115.8 (113.2, 118.4) 119.8 (117.1, 122.6) -0.5 (-2.0, 1.0) 0.52
Diet*time interaction 0.03
Evening systolic blood pressure (mmHg)
Baseline 123.0 (120.1, 126.0) 126.2 (123.0, 129.3)
3 months 118.4 (115.4, 121.4) 121.4 (118.1, 124.6) 0.2 (-1.4, 1.7) 0.82
33
6 months 116.2 (113.2, 119.2) 119.4 (116.2, 122.6) -0.0 (-1.6, 1.5) 0.96
Diet*time interaction 0.96
1Data are presented as mean±95% confidence intervals. Assessed using a mixed effects model with Diet, Time and a Diet*Time interaction term
as fixed effects, and subject ID as a random intercept. Adjusted differences are the predicted marginal means for each visit compared to baseline.
MedDiet, Mediterranean diet; HabDiet, Habitual Diet
34
FIGURE 1. Flow diagram for participation in the MedLey trial, blood pressure data
collection shown in left column and flow mediated dilatation data collection shown in right
column. MedDiet, Mediterranean diet intervention group; HabDiet, habitual diet control
group; BP, blood pressure; FMD, flow mediated dilatation.
FIGURE 2. Mean change in Mediterranean diet adherence score from baseline to 2 and 4
months, for the Mediterranean diet (MedDiet, n = 80) and the Habitual diet (HabDiet, n = 69)
groups. Linear mixed effects model with diet*interaction term used to determine overall diet
effect, within group and between group differences at 2 and 4 months. P=0.001 for
interaction. *P<0.001 for difference from baseline. †P<0.001 for between group difference.
FIGURE 3. Observed values for the flow mediated dilatation (FMD) curves from baseline (0
seconds) to 165 seconds post-cuff deflation, for the Mediterranean diet group (MedDiet,
n=48) and the Habitual diet group, (HabDiet, n=34) and visit (baseline and 6 months). Data
was analysed using a linear mixed effects model with FMD at 6 months as the outcome
variable. The fixed effects in the models included baseline values, diet group, and time post
cuff deflation (0 to 165 with 15 second intervals). The overall effect of the MedDiet on FMD
was established using the difference in the overall predicted marginal means for the 2 diets
(taken across the 165 second recording period) (P<0.001).