kastarinen ht mott

8/12/2019 Kastarinen Ht Mott

1/8

Non-pharmacological treatment of hypertension in primaryhealth care: a 2-year open randomized controlled trial oflifestyle intervention against hypertension in eastern FinlandMika J. Kastarinena, Pekka M. Puskab, Maarit H. Korhonenc ,Juha N. Mustonen d, Veikko V. Salomaae , Jouko E. Sundvallf,Jaakko O. Tuomilehtoe,g , Matti I. Uusitupac and Aulikki M. Nissinenefor the LIHEF Study Group

Objective To assess whether lifestyle counselling iseffective in non-pharmacological treatment ofhypertension in primary health care.

Design Open randomized controlled trial.

Setting Ten municipal primary health care centres ineastern Finland.

Patients Seven hundred and fteen subjects aged 2574years with systolic blood pressure 140179 mmHg and/ordiastolic blood pressure 90109 mmHg orantihypertensive drug treatment.

Interventions Systematic health counselling given by localpublic health nurses for 2 years.

Main outcome measures Blood pressure, lipids andlifestyle data were collected annually.

Results Among participants with no antihypertensive drugtreatment, the net reductions after 1 year both in systolicblood pressure [ 2 2.6 mmHg; 95% condence interval (CI),2 4.7 to 2 0.5 mmHg] and in diastolic blood pressure(2 2.7 mmHg; 95% CI, 2 4.0 to 2 1.4 mmHg) weresignicant in favour of the intervention group. Thisdifference in blood pressure change was maintainedduring the second year. In participants withantihypertensive drug treatment, no signicant differencein blood pressure reduction was seen between the groupsduring the study.

Conclusions A relatively modest, but systematiccounselling in primary health care can, at least amonguntreated hypertensive subjects, produce reductions inblood pressure levels that are modest for the individual,but very important from the public health point of view. J Hypertens 20:25052512 2002 Lippincott Williams &Wilkins.

Journal of Hypertension 2002, 20 :25052512

Keywords: hypertension, randomized controlled trial, life change events

aDepartment of Public Health and General Practice, University of Kuopio, Kuopio,Finland, b Department of Non-communicable Disease Prevention and HealthPromotion, WHO, Geneva, Switzerland, c Department of Clinical Nutrition,University of Kuopio, Kuopio, Finland, d Department of Internal Medicine, NorthKarelia Central Hospital, Joensuu, Finland, e Department of Epidemiology andHealth Promotion, National Public Health Institute, Helsinki, Finland,fDepartment of Health and Functional Capacity, National Public Health Institute,Helsinki, Finland and g Department of Public Health, University of Helsinki,

Finland.

Sponsorship: This study has been supported by the grants from the followinginstitutions: the Social Insurance Institution of Finland, the Juho VainioFoundation , the National Public Health Institute, the Finnish Ofce for HealthCare Technology Assessment , the Graduate School of Public Health in theUniversity of Kuopio, the Finnish Foundation for Cardiovascular Research, theNorth Savo Regional Fund of the Finnish Cultural Foundation, the DuodecimSociety, the Aarne and Aili Turunen Foundation, the Ida Montin Foundation andthe EVO Funding from the Kuopio University Hospital.

Correspondence and request for reprints to Mika Kastarinen, MD, Department ofPublic Health and General Practice, University of Kuopio, PB 1627, 70211Kuopio, Finland.Tel: +358-17-162912; fax: +358-17-162937; e-mail: mika.kastarinen@uku.

Received 23 May 2002 Revised 26 July 2002Accepted 21 August 2002

IntroductionEvidence indicates that lifestyle measures such asweight reduction, moderation of alcohol consumption,reduction in salt intake and increase in physicalactivity are feasible and effective in lowering bloodpressure (BP), either alone or in combination withantihypertensive drug therapy [16] . Some studieshave suggested that fat-modication of the diet or adiet with low saturated fat intake, but rich in fruits,

vegetab les and bre, have a signicant BP-loweringeffect [4,7]. In addition to its effect on BP, suchmodication of lifestyle can also reduce the levels of other cardiovascular risk factors. According to thelatest international and national hypertension guide-lines, non-pharmacological measures are recommendedas the rst-line therapy for the patients with newlydiagnosed, uncomplicated hypertension and theyshould also be applied in the treatment of every

Original article 2505

0263-6352 & 2002 Lippincott Williams & Wilkins 10.1097/01.hjh.0000042893.24999.db


2/8

hypert ensive patient treated with antihypertensivedrugs [810] .

Most of the clinical trials assessing the feasibility andeffects of health education targeted to the preventionand control of hypertension have been performed in

academic study centres by expert personnel trained forthe trial. Because these studies were mainly designedto test the efcacy of such interventions, the interven-tion programmes have been intensive. Due to thelimited resources of the public health sector, theimplementation of such intensive lifestyle modicationprogrammes in primary health care, i.e. in the settingwhere most of the hypertensive patients are treated,would be difcult. Therefore, we decided to conduct arandomized, controlled trial to assess the efcacy of arelatively low-intensity patient-counselling programmeplanned for hypertensive subjects in primary healthcare.

MethodsParticipants and randomizationThe Lifestyle Intervention against Hypertension inEastern Finland (LIHEF) study was conducted in 10municipal primary health care centres in eastern Fin-land, mainly in the province of North Karelia. Eight of the health care centres were located in rural municipa-lities with 300012 000 inhabitants and two in thetowns with 50 000 and 90 000 inhabitants (51% of allparticipants). The study protocol was approved by theEthics Committee of the Kuopio University Hospital.The study participants were enrolled between February

1996 and June 1997. Eligible subjects were men andwomen aged 2574 years with systolic blood pressure(SBP) 140 179 mmHg and/or diastolic blood pressure(DBP) 90109 mmHg or on antihypertensive drugtherapy. Exclusion criteria included secondary hyper-tension, mental or physical illness serious enough topotentially inuence the compliance with the studyprocedures, alcoholism, type 1 diabetes, current orplanned pregnancy and history of myocardial infarctionor stroke within the preceding 3 months.

Three screening visits in 1-week intervals were orga-nized in the health centres to measure the BP of the

subjects not using any antihypertensive drugs. Duringeach screening visit, BP was measured twice from theright arm of the subject according to the WHOMONICA p roto col with a standard mercury sphygmo-manometer [11] . The mean of the BP measurementsperformed in the second and third visit (four readings)was used as the screening BP.

The randomization visit was organized within 30 daysafter the third screening visit in a participating healthcentre. A written informed consent was obtained fromevery eligible person agreeing to participate, after

which they were randomized to receive intervention orusual care by the same study physician using a dice(odd numbers intervention group; even numbers control group). Of the 813 subjects originally eligiblefor the study, 715 were eventually assigned to theintervention group or to the usual care group.

MeasurementsBlood pressure, weight, waist and hip circumferenceswere measured annually from every participant. Heightwas measured at the baseline only. Body mass index(BMI) was calculated as kg/m 2 . During the annualvisits, BP was measured twice using the same methodas during the screening and the mean was used in theanalyses. A single trained study nurse, who was blindedto the treatment assignment, performed most of the BPand anthropometrical measurements during the screen-ing and the follow-up. Only in one health centre with176 participants, another trained study nurse carried out

the measurements during the screening and randomiza-tion visits.

Information on socio-economic status, medical history,smoking, alcohol use, physical activity and daily medi-cation were collected annually using questionnaireswith standard questions. A 4-day food record wascollected for the annual visits. The detailed informationof the dietary anal yses used in the study will bepublished elsewhere [12].

All biochemical assays were performed at the Depart-ment of Biochemistry of the National Public Health

Institute in Helsinki. Blood samples were drawn at theparticipating health centres from the study subjectsafter 12-h fasting. Serum total and high-density lipopro-tein (HDL) cholesterol, triglycerides and insulin weredetermined. Low-density lipoprotein (LDL) cholesterolconcentrations were calculated using Friedewalds for-mula [13]. Twenty-four-hour urine specimens werecollected for the determination of the 24-h potassiumand sodium excretion.

InterventionThe intervention goals for the study subjects were: (1)normal weight (BMI , 25 kg/m 2); (2) daily sodium

chloride intake less than 5 g; (3) alcohol consumptionfewer than two drinks per day; (4) to exercise atmoderate intensity at least three times per week for30 min; and (5) to stop smoking, if a smoker.

The study physician and a nutritionist trained the localpublic health nurses who participated in the study. Thetraining sessions dealt with simple counselling andbehaviour modication methods targeting weight re-duction, reduction in salt, alcohol and saturated fatconsumption, as well as an increase in leisure-timephysical activity. The nurses were given a folder with

2506 Journal of Hypertension 2002, Vol 20 No 12


3/8

detailed information of the dietary recommendationsand with practical tips to achieve these recommenda-tions in everyday life.

The core of the actual intervention (Fig. 1) consisted of four visits by the participants to local public healthnurses during the rst year of the follow-up (1, 3, 6 and9 months after randomization), and of three visitsduring the second year (15, 18 and 21 months after therandomization). At these visits, the participants weresystematically instructed to change their health behav-iour primarily on the basis of their individual situation.At each visit, BP and weight were measured, and thevalues, as well as the changes in lifestyle factors to bereached before the next study visit, were written downusing a special follow-up card designed for the study. Awritten feedback of the 4-day food record was sent to

the public health nurse to support the intervention. Inaddition, a 2-h group session was organized for theintervention group in every health care centre at 6 and18 months after the randomization. These two groupmeetings concentrated mainly on advice targeting re-duction of salt intake and overweight. During the 2-year follow-up, the participants in the usual care groupwere instructed to visit their own physicians and publichealth nurses according to usual practices.

Statistical analysesStatistical analyses were conducted with SPSS for

Windows version 10.0 (SPSS Inc., Chicago, Illinois,USA). For continuous variables, the t -test was used totest the differences and changes in mean values be-tween the groups. Condence intervals (CI) for the

differences in proportions were calculated using aspecial software package [14]. An intention-to-treat-analysis was used, i.e. all subjects assigned to interven-tion or usual care were included in the analysis. In thecase of missing responses, the last observed responsewas used when calculating the 1- and 2-year changes incontinuous variables (the carry-forward method). Thesame method was used with dichotomous variables. Ina separate analysis of BP changes in subjects withoutprevious antihypertensive drug treatment, the last BPmeasurement without antihypertensive drug treatmentwas used if drug treatment was initiated during thetrial. Accordingly, in subjects already on antihyperten-

sive medication, the last BP measurement with anti-hypertensive drug treatment was used if the treatmentwas discontinued during the trial. In the calculations of BP changes, the changes in doses of antihypertensivemedication were not taken into account. Multiple linearregression analysis was used to examine the associationsof changes in body weight, sodium and potassiumexcretion, alcohol intake and leisure-time physicalactivity (times/week) with the changes in BP, control-ling for the baseline BP levels.

The original target sample size was 800 subjects, which

Follow-up,Months

Randomisation

Measurementsat baseline,12 months and24 months:

PN, FFR

PNPN,GM

PN FFR PN PN,GM

PN

Intervention

0 1 3 6 9 12 15 18 21 24

Usual care*

BPWeightHeight (only at baseline)WaistHipTotal cholesterolHDL cholesterolLDL cholesterolTriglycerides

24-hour urinary sodium excretion24-hour urinary potassium excretionInsulinQuestionnaires4-day Food record

Fig. 1

The design of the study. PN, public health nurse; FFR, feedback from the food record; GM, group meeting. Usual care group visited their ownpublic health nurses as usual. BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

Non-pharmacological treatment of hypertension Kastarinen et al . 2507


4/8

was not reached due to relatively numerous dropoutsbefore the randomization among the already recruitedsubjects. It was estimated that this sample size wouldenable detection of a 3.2 mmHg difference in changeof SBP and a 1.6 mmHg difference in change of DBPbetween the intervention and usual care groups, with

80% power at the 5% signicance level.

ResultsBaseline characteristics and adherence to treatmentThe mean age of all participants was 54.3 years. Of theparticipants, 52% were on antihypert ensive dru g treat-ment at the beginning of the trial (Table 1) . Therewere no statistically signicant differences between thegroups in any of the baseline variables analysed.Attendance rates at th e 1-year and 2-year study visitswere satisfactory (Fig. 2). The attendance rate at bothgroup meetings organized for the intervention groupafter 6 and 18 months of intervention was 50%. The

subjects who dropped out during the different phasesof the study were, at baseline measurements, signi-cantly younger (50 versus 55 years) and heavier (83.1versus 80.1 kg) compared with the attenders of the 2-year visit. Also, alcohol use (73 versus 42 g/week) andproportion of smokers (13 versus 7%) were signicantlyhigher among the drop-outs at the baseline.

Changes in blood pressure and lifestyle factorsThe changes in BP and othe r continu ous variables inthe two groups are shown in Table 2. Without takinginto account the effect of antihypertensive drug treat-ment, the reduction in DBP during the rst study year

was signicantly greater in the intervention group com-

pared to the usual care group. The reductions in SBP at1-year follow-up and at 2-year follow-up and in DBPfrom baseline to 2 years tended to be greater in the

intervention group, although they did not reach the

Table 1 Baseline characteristics of participants in the intervention and usual care groups. Values are mean6 SD or percentage

Intervention (n 360) Usual care ( n 355)

Age (years) 54.4 10.1 54.2 9.9Female (%) 52 54Antihypertensive drug treatment (%) 52 53Current smoker (%) 9 7History of coronary heart disease (%) 5 3Moderate physical activity at least three times per week (%) 51 51Alcohol consumption (g/week) 47 83 48 74

Body weight (kg) 81.1 15.7 80.0 14.8Body mass index (kg/m 2 ) 28.9 4.6 28.5 4.5Waist circumference (cm) 97.2 13.1 95.8 12.8

Hip circumference (cm) 104.7 10.4 104.1 10.0Systolic blood pressure (mmHg) 149 16 148 16No antihypertensive drug treatment 152 14 150 14Antihypertensive drug treatment 147 18 146 18

Diastolic blood pressure (mmHg) 91 9 91 8No antihypertensive drug treatment 93 8 93 9Antihypertensive drug treatment 89 9 89 8

Total cholesterol (mmol/l) 5.66 0.91 5.59 0.93LDL-cholesterol (mmol/l) 3.64 0.81 3.56 0.79HDL-cholesterol (mmol/l) 1.32 0.33 1.36 0.38Triglycerides (mmol/l) 1.56 1.01 1.49 1.0024-h urinary sodium excretion (mmol) 146 56 142 56

24-h urinary potassium excretion (mmol) 83 27 83 28Serum insulin (IU/l) 12.2 6.8 11.6 6.3

LDL, low-density lipoprotein; HDL, high-density lipoprotein.

813 eligible subjects after screening

98 declined to participate

715 randomized

Interventionn 360

Usual caren 355

Attended 1-year study visit317/360 (88%)




Fig. 2

The numbers of participants involved throughout the trial.



5/8

level of statistical signicance. In the subgroup with noantihypertensive drug treatment, the reductions in both

SBP and DBP were signicantly greater in the inter-vention group compared with the usual care duringboth the rst and the second year of follow-up (Table3). In subjects with antihypertensive drug treatment atbaseline, the BP reductions were of the same magni-tude in both groups. In these subjects, the number of antihypertensive drugs used per patient did not changesignicantly in either of the randomized groups duringthe follow-up. In both groups, 70% of the drug-treatedpatients were on monotherapy at the end of the study.

Among the subjects with antihypertensive drug treat-ment, the self-reported frequency of BP measurements

during the previous year decreased signicantly morein the intervention group during the rst year (netchange 2.3 measurements/year; 95% CI, 4.3 to 0.3;data not shown). Otherwise the number of BP measure-ments did not differ signicantly between the groupsduring any phase of the study.

The net reductions (intervention versus usual care) in

weight at 1 and 2 years of follow-up were signicant.Eight per cent of the initially overweight participants

(BMI > 25 kg/m2) assigned to intervention had achie-

ved normal weight at the end of the trial, which wassignicant ly more than that in the usual care group(Table 4). Also, the waist and hip circumferences fellsignicantly more in the intervention group than inusual care throughout the study. The changes in 24-hurinary sodium and potassium excretion were small,with no signicant differences between the groups.Self-reported alcohol consumption fell signicantlymore during the rst study year in the interventiongroup, but this difference disappeared during thesecond year. Compared to the usual care group, asignicantly larger proportion of the participants in the

intervention group had increased their physical activityto the target level at both 1-year and 2-year visits.

The net reduction in weight between the randomizedgroups was signicantly greater in subjects with noantihypertensive drug treatment during the rst yearcompared to the group with antihypertensive drugtreatment ( 1.5 versus 0.8 kg, P for the interaction

Table 2 Changes in continuous variables at 1 and 2 years in the intervention and usual care groups

Change 01 year Change 02 years Difference in change (95% CI) (intervention versus usual care)

Intervention Usual care Intervention Usual care 0 1 year 0 2 years

Systolic blood pressure (mmHg) 4.7 3.4 6.2 4.2 1.3 ( 3.2, 0.6) 2.0 ( 4.3, 0.3)Diastolic blood pressure (mmHg) 4.0 2.4 4.3 3.2 1.6 ( 2.7, 0.6) 1.1 ( 2.4, 0.2)

Body weight (kg)

1.5

0.2

1.5

0.3

1.3 ( 1.7, 0.9)

1.2 ( 1.7, 0.7)Waist circumference (cm) 1.2 0.3 1.2 0.2 1.5 ( 2.1, 1.0) 1.4 ( 2.0, 0.8)Hip circumfererence (cm) 1.4 0.5 1.4 0.4 0.9 ( 1.4, 0.4) 0.9 ( 1.5, 0.4)Alcohol consumption (g/week) 7 1 7 5 8 ( 15, 0) 2 ( 9, 5)Total cholesterol (mmol/l) 0.05 0.03 0.03 0.07 0.02 ( 0.11, 0.06) 0.11 ( 0.20, 0.01)LDL-cholesterol (mmol/l) 0.06 0.01 0.11 0.04 0.05 ( 0.12, 0.03) 0.15 ( 0.23, 0.05)HDL-cholesterol (mmol/l) 0.02 0.01 0.10 0.07 0.01 ( 0.01, 0.04) 0.03 (0.00, 0.07)Triglycerides (mmol/l) 0.03 0.06 0.06 0.06 0.03 ( 0.05, 0.12) 0.00 ( 0.09, 0.10)Urinary sodium excretion (mmol/day) 9 6 7 2 3 ( 10, 5) 5 ( 14, 3)Urinary potassium excretion (mmol/day) 1 1 3 1 0 ( 4, 4) 2 ( 2, 5)Serum insulin (IU/l) 0.8 0.2 1.1 0.5 0.6 ( 1.2, 0.1) 0.6 ( 1.1, 0.1)

CI, condence interval; LDL, low-density lipoprotein; HDL, high-density lipoprotein.

Table 3 Changes (95% CI in parentheses) in blood pressure levels stratied by antihypertensive drug treatmentstatus

No antihypertensive drug treatment Antihypertensive drug treatment

Intervention (n 175) Usual care ( n 166) Intervention (n 185) Usual care ( n 189)

Systolic blood pressure (mmHg)Baseline mean 152 150 147 146Change 0 1 year 3.0 ( 4.6, 1.4) 0.4 ( 1.8, 1.0) 4.8 ( 6.6, 2.9) 4.0 ( 5.9, 2.1)Change 0 2 years 2.0 ( 3.7, 0.3) 0.4 ( 1.3, 2.0) 6.0 ( 8.5, 3.5) 4.7 ( 6.6, 2.6)Net change 0 1 year 2.6 ( 4.7, 0.5) 0.8 ( 3.4, 1.9)Net change 0 2 years 2.4 ( 4.7, 0.0) 1.3 ( 4.5, 1.8)

Diastolic blood pressure (mmHg)Baseline mean 93 93 89 89Change 0 1 year 3.3 ( 4.3, 2.3) 0.6 ( 1.4, 0.3) 3.6 ( 4.6, 2.7) 2.8 ( 3.9, 1.7)Change 0 2 years 2.4 ( 3.4, 1.4) 0.4 ( 1.4, 0.6) 3.8 ( 5.1, 2.6) 3.7 ( 4.9, 2.6)Net change 0 1 year 2.7 ( 4.0, 1.4) 0.8 ( 2.3, 0.6)Net change 0 2 years 2.0 ( 3.4, 0.6) 0.1 ( 1.8, 1.6)

CI, condence interval.



6/8

term 0.021), but such a difference was not detected inthe analysis from baseline to 2 years. Similarly, the netreductions in alcohol consumption from baseline to1-year and 2-year visits were signicantly greater in

this group compared to the group with antihypertensivedrug treatment (from baseline to 1-year visit 17 versus1 g/week, P for the interaction term 0.025; from base-line to 2-year visit 9 versus 5 g/week, P for theinteraction term 0.029). The response for the interven-tion in terms of lifestyle changes did not differ sig-nicantly between the sexes or according to thebaseline age. In a multiple linear regression analysis, apositive relation with 2-year change in blood pressurewas found for weight change in both subjects with noantihypertensive drug treatment (SBP, P 0.003; DBP, P , 0.001) and in drug-treated subjects (SBP, P 0.034; DBP, P 0.012), but not for the changes in

other lifestyle variables. In a separate analysis amongthe subjects with no antihypertensive drug treatment,adjusting for both BP and weight at baseline, theestimated effect of the 1 kg weight lost at 2 years was areduction of 0.55 mmHg in SBP and of 0.50 mmHg inDBP

Changes in other cardiovascular risk factorsAfter 2 years of intervention, the net changes in total( 0.10 mmol/l) and in LDL cholesterol ( 0.15 mmol/l)were signicant in favour of the intervention group. Also,in persons without lipid-lowering drug treatment duringthe study, the net reductions in total cholesterol

( 0.11 mmol/l; 95% CI 0.20 to 0.02) and in LDLcholesterol ( 0.13 mmol/l; 95% CI 0.22 to 0.03) werelarger in the intervention group than in the control group(data not shown). Smoking was already rare at baseline,and no signicant changes in that habit were observedduring the study. The serum insulin concentration de-creased more among the persons assigned to interven-tion, but the net change was not statistically signicant.

DiscussionThe lifestyle counselling provided in the LIHEF studycould produce a signicant reduction in BP level of the

subjects with no antihypertensive drug treatment com-pared with usual care for 2 years. Weight loss wassignicantly greater in the intervention group at the 1-year visit, and this difference was maintained during

the second year. The intervention programme couldnot induce any signicant reductions in salt intake.Despite the fact that in 95% of the participants theweekly alcohol consumption was already at the recom-mended level at baseline, a small but signicant reduc-tion in alcohol intake occurred in the interventiongroup during the rst year. The self-reported leisure-time physical activity increased signicantly more inthe intervention group throughout the study.

Large-scale randomized trials reporting the effects of lifestyle intervention on BP and other cardiovascular riskfactors in hypertensive persons in the primary care setting

are rare. The very few previous studies have includedrelatively small numbers of pati ents, wit h the maxim umfollow-up time of 12 months [1517] . Iso e t al . [18]reported that a community-based trial lowered BP amonghypertensive subjects. In their study, the interventionwas based mainly on group sessions instead of theindividualized health counselling used in our study.

The net changes in BP and lipid levels detected in ourstudy are in accordance with trials of multiple riskfactor intervention or dietary intervention with peopleat high risk but not necessarily hypertensive [1921] .The reduction achieved in body weight after 2 years of

intervention was almost the sa me as in some trials withmore intensive intervention [1]. In contrast, the BPreduction observed in the subjects without antihyper-tensive drug treatment was smaller compared to someclinical trials of non-pharmacological treatment of hypertension [4,22] . These trials used very intensiveintervention compared to our study. Thus it seems thatat least some of the results obtained in high-intensityintervention trials can be translated successfully toprimary health care.

According to the separate analysis of the dietary data of

Table 4 The percentage of participants reaching the pre-dened goals for the intervention during the trial among the persons out of goalsat baseline

Year 1 Year 2

Interventiona Usual care aDifference in

change (95% CI) Intervention Usual careDifference in

change (95% CI)

BMI,

25 kg/m2

7.8 (294) 3.9 (279) 3.9 (0.1, 7.7) 8.2 3.6 4.6 (1.0, 8.4)> 5% reduction in BMI 18.7 9.3 9.4 (3.8, 15.0) 22.1 10.4 11.7 (5.8, 17.7)24-hour U-Na excretion , 85 mmol 10.6 (301) 8.7 (286) 1.9 ( 2.9, 6.7) 7.0 7.0 0.0 ( 4.1, 4.1)

> 10% reduction in 24-h U-Na excretion 41.5 33.6 7.9 (0.2, 15.8) 41.5 33.6 7.9 (0.2, 15.8), 2 alcohol drinks per day 27.8 (18) 9.5 (21) 18.3 ( 6.0, 42.5) 16.7 9.5 7.1 ( 14.2, 28.5)Recommended level of physical activity b 34.7 (173) 24.0 (171) 10.7 (1.2, 20.3) 34.1 22.8 11.3 (1.8, 20.8)Not smoking 6.5 (31) 12.0 (25) 5.6 ( 20.9, 9.9) 12.9 12.0 0.9 ( 16.5, 18.3)

CI, condence interval; BMI, body mass index; U-Na, urinary sodium. a Number of subjects out of each goal at baseline in parentheses. b Leisure-time physical activity atleast three times per week and at least to 30 minutes.



7/8

this study, the proportion of fat, and especially of saturated fats, in total energy intake decreased signi-cantly mor e in the intervention group compared tousual care [12]. Also, the total energy intake tended todecrease more in the intervention group, although notreaching the level of statistical signicance. In addition

to the increase in physical activity, these changes indiet may have contributed to the observed differencesin changes of body weight, lipid levels and BP betweenthe randomized groups. The dietary data were inaccordance with the results of the 24-h urinary sodiumexcretion, showing no signicant changes in sodiumintake. These results repeat the ndings of the manyother studies that have demonstrated the difculties inachieving the re comm ended level of salt intake in free-living subjects [23,24] . It has been suggested that themain reason for this difculty in salt restriction seen inall Western countries is the still relat ively high concen-tration of salt in processed foods [25].

The differences in BP reduction observed between thegroups could not be explained by accustomization withBP measurement, because there was not any differencein self-reported frequency of BP measurements be-tween the groups during the study. One explanation forthe greater fall in blood pressure among the participantswho continued antihypertensive drug treatment com-pared with the participants without antihypertensivedrugs could be a more regular use of antihypertensivedrugs during the trial than before.

As usual in volunteer-based intervention studies, the

study sample is seldom fully representative of the back-ground population. Highly motivated volunteers areusually more susceptible to accept the recommendedintervention than the population at large. On the otherhand, many volunteers in lifestyle intervention studieshave already previously changed their lifestyle, whichcould reduce the power of the intervention. In our study,the mean BMI, total cholesterol and the prevalence of smoking were lower than in Finnish hypertensive subjects in the population-based FINRISK study in 1997[2628] . The study participants also came from ageographical area with a long history of cardiovasculardisease prevention activities, and thus many of them

already had a relatively good knowledg e ab out lifestylefactors affecting the cardiovascular risk [29]. In addition,the contamination of the control group, due to the factthat their follow-up visits were done by the same nursesas with the intervention group, might have reduced thedifference in the lifestyle changes between the groups.Also, the fact that they were under systematic observa-tion in an interesting study likely inuenced them.Thus, our observed effects of the intervention, as usualin this kind of studies, are likely to be conservative.

In conclusion, the favourable changes in BP and other

cardiovascular disease risk factors in hypertensive per-sons participating in our study were smaller than in thetrials with more intensive interventions. However, theprincipal aim of this study was to nd out the extent towhich lifestyle intervention will work in the usualprimary health care setting. From this point of view,

our results were quite satisfactory, considering thelimited requirement for the use of health care re-sources. The potential of the intervention shown in thisstudy can certainly be much improved by furtherdevelopment and systematic dissemination, especiallyconcerning newly detected hypertensive persons. Non-pharmacological treatment of hypertension has beenadvocated for a long time, but so far only limitedevidence and experience has been available as to itseffective implementation within primary care. The taskis not easy, due to the limited time and resources thatthe public health service can allocate for such preven-tive services. However, we have shown that this

approach works.

AcknowledgementsThe statistician Pirjo Halonen, MSc, advised M. Kastar-inen about data analysis. Mr Veli Koistinen was respon-sible for preparing the database. Study nurses AnneliMitrunen and Mari Aalto screened the study subjectsand performed BP and anthropometrical measurements.Registered dietitians Sari Aalto and Sointu Lassila wereresponsible for the training of nurses in dietary issuesand for the group sessions organized for subjectsassigned to intervention. We thank for the staff of theNorth Karelia Project for the help given in coordinationof the study. Antti Jula MD, PhD; Antti Malmivaara,MD, PhD; Matti Romo, MD, PhD; Jyrki Olkinuora,MD, Markku Helio vaara, MD, PhD; Erkki Vartiainen,MD, PhD and Timo Lakka, MD, PhD were membersof the LIHEF study group and helped in designing thestudy. The authors are grateful to all practice staff working in the participating health care centres, as wellas to the hypertensive persons participating the study.

References1 Stevens VJ, Obarzanek E, Cook NR, Lee IM, Appel LJ, Smith West D,

et al . Long-term weight loss and changes in blood pressure: results of theTrials of Hypertension Prevention, phase II. Ann Intern Med 2001; 134 :111.

2 Puddey IB, Beilin LJ, Vandongen R. Regular alcohol use raises bloodpressure in treated hypertensive subjects. A randomised controlled trial.Lancet 1987; 1 :647651.

3 Cutler JA, Follmann D, Allender PS. Randomised trials of sodium reduction:an overview. Am J Clin Nutr 1997; 65 (suppl 2) :643651.

4 Sacks FM, Svetkey LP, Vollmer WM, Appel LJ, Bray GA, Harsha D, et al .Effects on blood pressure of reduced dietary sodium and the DietaryApproaches to Stop Hypertension (DASH) diet. DASHSodium Colla-borative Research Group. N Engl J Med 2001; 344 :310.

5 Halbert JA, Silagy CA, Finucane P, Withers RT, Hamdorf PA, Andrews GR.The effectiveness of exercise training in lowering blood pressure: a meta-analysis of randomised controlled trials of 4 weeks or longer. J HumHypertens 1997; 11 :641649.

6 Neaton JD, Grimm RH Jr, Prineas RJ, Stamler J, Grandits GA, Elmer PJ,et al . Treatment of Mild Hypertension Study. Final results. Treatment ofMild Hypertension Study Research Group. JAMA 1993; 270 :713724.



8/8

7 Puska P, Iacono JM, Nissinen A, Korhonen HJ, Vartiainen E, Pietinen P,et al . Controlled, randomised trial of the effect of dietary fat on bloodpressure. Lancet 1983; 1 :15.

8 The sixth report of the Joint National Committee on prevention, detection,evaluation, and treatment of high blood pressure. Arch Intern Med 1997;157 :24132446.

9 Guidelines Subcommittee.1999 World Health Organization InternationalSociety of Hypertension Guidelines for the Management of Hypertension.

J Hypertens 1999; 17 :151183.10 Ramsay L, Williams B, Johnston G, MacGregor G, Poston L, Potter J, et al .Guidelines for management of hypertension: report of the third workingparty of the British Hypertension Society. J Hum Hypertens 1999;13 :569592.

11 Kuulasmaa K, Hense HW, Tolonen H for the WHO MONICA project.Quality assessment of data on blood pressure in the WHO MONICAproject. http:www.ktl./publications/monica/bp/bpqa.htm. [Accessed 13March 2002].

12 Korhonen M, Kastarinen M, Uusitupa M, Puska P,NissinenA. Theeffect ofintensied diet counselling on thediet of hypertensivesubjectsin primaryhealth care.A 2-year open randomised controlled trialof lifestyleinterventionagainst hypertension in Eastern Finland (LIHEF). PrevMed 2002, (in press).

13 Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentrationof low-density lipoprotein cholesterol in plasma, without use of thepreparative ultracentrifuge. Clin Chem 1972; 18 :499502.

14 Gardner M, Gardner S, Winter P. Condence interval analysis (CIA).Microcomputer program manual . London: BMJ Publishing Group and MJ

Gardner; 1991.15 Koopman H, Spreeuwenberg C, Westerman RF, Donker AJ. Dietarytreatment of patients with mild to moderate hypertension in a generalpractice: a pilot intervention study (1). The rst three months. J HumHypertens 1990; 4 :368371.

16 Cohen MD, DAmico FJ, Merenstein JH. Weight reduction in obesehypertensive patients. Fam Med 1991; 23 :2528.

17 Woollard J, Beilin L, Lord T, Puddey I, MacAdam D, Rouse I. A controlledtrial of nurse counselling on lifestyle change for hypertensives treated ingeneral practice: preliminary results. Clin Exp Pharmacol Physiol 1995;22 :466468.

18 Iso H, Shimamoto T, Yokota K, Sankai T, Jacobs DR Jr, Komachi Y.Community-based education classes for hypertension control. A 1.5-yearrandomised controlled trial. Hypertension 1996; 27 :968974.

19 Ebrahim S, Smith GD. Systematic review of randomised controlled trials ofmultiple risk factor interventions for preventing coronary heart disease.BMJ 1997; 314 :16661674.

20 Brunner E, White I, Thorogood M, Bristow A, Curle D, Marmot M. Can

dietary interventions change diet and cardiovascular risk factors? A meta-analysis of randomised controlled trials. Am J Public Health 1997;87 :14151422.

21 Steptoe A, Doherty S, Rink E, Kerry S, Kendrick T, Hilton S. Behaviouralcounselling in general practice for the promotion of healthy behaviouramong adults at increased risk of coronary heart disease: randomised trial.BMJ 1999; 319 :943947.

22 Whelton PK, Appel LJ, Espeland MA, Applegate WB, Ettinger WH Jr,Kostis JB, et al . Sodium reduction and weight loss in the treatment ofhypertension in older persons: a randomized controlled trial of non-pharmacologic interventions in the elderly (TONE). TONE CollaborativeResearch Group. JAMA 1998; 279 (11) :839846.

23 Pietinen P, Tanskanen A, Nissinen A, Tuomilehto J, Puska P. Changes indietary habits and knowledge concerning salt during a community-basedprevention programme for hypertension. Ann Clin Res 1984; 16 (suppl43) :150155.

24 Evers SE, Bass M, Donner A, McWhinney IR. Lack of impact of saltrestriction advice on hypertensive patients. Prev Med 1987; 16 :213220.

25 Kumanyika S. Behavioral aspects of intervention strategies to reducedietary sodium. Hypertension 1991; 17 (suppl 1) :190195.26 Kastarinen MJ, Nissinen AM, Vartiainen EA, Jousilahti PJ, Korhonen HJ,

Puska PM, et al . Blood pressure levels and obesity trends in hypertensiveand normotensive Finnish population from 1982 to 1997. J Hypertens2000; 18 :255262.

27 Kastarinen M, Tuomilehto J, Vartiainen E, Jousilahti P, Sundvall J, Puska P,et al . Trends in lipid levels and hypercholesterolaemia in hypertensive andnormotensive Finnish adults from 1982 to 1997. J Intern Med 2000;247 :5362.

28 Kastarinen M, Tuomilehto J, Vartiainen E, Jousilahti P, Nissinen A, Puska P.Smoking trends in hypertensive and normotensive Finns during 19821997. J Hum Hypertens 2002; 16 :299303.

29 Puska P, Tuomilehto J, Nissinen A, Vartiainen E. The North Karelia Project: 20-year results and experiences . Helsinki: National Public Health Institute;1995.


kastarinen ht mott

Documents