Non-pharmacological treatment of hypertension in primary
health care: a 2-year open randomized controlled trial of
lifestyle intervention against hypertension in eastern Finland Mika J. Kastarinen a , Pekka M. Puska b , Maarit H. Korhonen c , Juha N. Mustonen d , Veikko V. Salomaa e , Jouko E. Sundvall f , Jaakko O. Tuomilehto e,g , Matti I. Uusitupa c and Aulikki M. Nissinen e for the LIHEF Study Group Objective To assess whether lifestyle counselling is effective in non-pharmacological treatment of hypertension in primary health care. Design Open randomized controlled trial. Setting Ten municipal primary health care centres in eastern Finland. Patients Seven hundred and fteen subjects aged 2574 years with systolic blood pressure 140179 mmHg and/or diastolic blood pressure 90109 mmHg or antihypertensive drug treatment. Interventions Systematic health counselling given by local public health nurses for 2 years. Main outcome measures Blood pressure, lipids and lifestyle data were collected annually. Results Among participants with no antihypertensive drug treatment, the net reductions after 1 year both in systolic blood pressure [22.6 mmHg; 95% condence interval (CI), 24.7 to 20.5 mmHg] and in diastolic blood pressure (22.7 mmHg; 95% CI, 24.0 to 21.4 mmHg) were signicant in favour of the intervention group. This difference in blood pressure change was maintained during the second year. In participants with antihypertensive drug treatment, no signicant difference in blood pressure reduction was seen between the groups during the study. Conclusions A relatively modest, but systematic counselling in primary health care can, at least among untreated hypertensive subjects, produce reductions in blood 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 a Department of Public Health and General Practice, University of Kuopio, Kuopio, Finland, b Department of Non-communicable Disease Prevention and Health Promotion, WHO, Geneva, Switzerland, c Department of Clinical Nutrition, University of Kuopio, Kuopio, Finland, d Department of Internal Medicine, North Karelia Central Hospital, Joensuu, Finland, e Department of Epidemiology and Health Promotion, National Public Health Institute, Helsinki, Finland, f Department 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 following institutions: the Social Insurance Institution of Finland, the Juho Vainio Foundation, the National Public Health Institute, the Finnish Ofce for Health Care Technology Assessment, the Graduate School of Public Health in the University of Kuopio, the Finnish Foundation for Cardiovascular Research, the North Savo Regional Fund of the Finnish Cultural Foundation, the Duodecim Society, the Aarne and Aili Turunen Foundation, the Ida Montin Foundation and the EVO Funding from the Kuopio University Hospital. Correspondence and request for reprints to Mika Kastarinen, MD, Department of Public Health and General Practice, University of Kuopio, PB 1627, 70211 Kuopio, Finland. Tel: +358-17-162912; fax: +358-17-162937; e-mail: mika.kastarinen@uku. Received 23 May 2002 Revised 26 July 2002 Accepted 21 August 2002 Introduction Evidence indicates that lifestyle measures such as weight reduction, moderation of alcohol consumption, reduction in salt intake and increase in physical activity are feasible and effective in lowering blood pressure (BP), either alone or in combination with antihypertensive drug therapy [16]. Some studies have suggested that fat-modication of the diet or a diet with low saturated fat intake, but rich in fruits, vegetables and bre, have a signicant BP-lowering effect [4,7]. In addition to its effect on BP, such modication of lifestyle can also reduce the levels of other cardiovascular risk factors. According to the latest international and national hypertension guide- lines, non-pharmacological measures are recommended as the rst-line therapy for the patients with newly diagnosed, uncomplicated hypertension and they should also be applied in the treatment of every Original article 2505 0263-6352 & 2002 Lippincott Williams & Wilkins 10.1097/01.hjh.0000042893.24999.db hypertensive patient treated with antihypertensive drugs [810]. Most of the clinical trials assessing the feasibility and effects of health education targeted to the prevention and control of hypertension have been performed in academic study centres by expert personnel trained for the trial. Because these studies were mainly designed to test the efcacy of such interventions, the interven- tion programmes have been intensive. Due to the limited resources of the public health sector, the implementation of such intensive lifestyle modication programmes in primary health care, i.e. in the setting where most of the hypertensive patients are treated, would be difcult. Therefore, we decided to conduct a randomized, controlled trial to assess the efcacy of a relatively low-intensity patient-counselling programme planned for hypertensive subjects in primary health care. Methods Participants and randomization The Lifestyle Intervention against Hypertension in Eastern Finland (LIHEF) study was conducted in 10 municipal 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 the towns with 50 000 and 90 000 inhabitants (51% of all participants). The study protocol was approved by the Ethics Committee of the Kuopio University Hospital. The study participants were enrolled between February 1996 and June 1997. Eligible subjects were men and women aged 2574 years with systolic blood pressure (SBP) 140179 mmHg and/or diastolic blood pressure (DBP) 90109 mmHg or on antihypertensive drug therapy. Exclusion criteria included secondary hyper- tension, mental or physical illness serious enough to potentially inuence the compliance with the study procedures, alcoholism, type 1 diabetes, current or planned pregnancy and history of myocardial infarction or 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. During each screening visit, BP was measured twice from the right arm of the subject according to the WHO MONICA protocol with a standard mercury sphygmo- manometer [11]. The mean of the BP measurements performed in the second and third visit (four readings) was used as the screening BP. The randomization visit was organized within 30 days after the third screening visit in a participating health centre. A written informed consent was obtained from every eligible person agreeing to participate, after which they were randomized to receive intervention or usual care by the same study physician using a dice (odd numbers intervention group; even numbers control group). Of the 813 subjects originally eligible for the study, 715 were eventually assigned to the intervention group or to the usual care group. Measurements Blood pressure, weight, waist and hip circumferences were measured annually from every participant. Height was measured at the baseline only. Body mass index (BMI) was calculated as kg/m 2 . During the annual visits, BP was measured twice using the same method as during the screening and the mean was used in the analyses. A single trained study nurse, who was blinded to the treatment assignment, performed most of the BP and anthropometrical measurements during the screen- ing and the follow-up. Only in one health centre with 176 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 questionnaires with standard questions. A 4-day food record was collected for the annual visits. The detailed information of the dietary analyses used in the study will be published 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 the participating health centres from the study subjects after 12-h fasting. Serum total and high-density lipopro- tein (HDL) cholesterol, triglycerides and insulin were determined. Low-density lipoprotein (LDL) cholesterol concentrations were calculated using Friedewalds for- mula [13]. Twenty-four-hour urine specimens were collected for the determination of the 24-h potassium and sodium excretion. Intervention The 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 consumption fewer than two drinks per day; (4) to exercise at moderate intensity at least three times per week for 30 min; and (5) to stop smoking, if a smoker. The study physician and a nutritionist trained the local public health nurses who participated in the study. The training sessions dealt with simple counselling and behaviour modication methods targeting weight re- duction, reduction in salt, alcohol and saturated fat consumption, as well as an increase in leisure-time physical activity. The nurses were given a folder with 2506 Journal of Hypertension 2002, Vol 20 No 12 detailed information of the dietary recommendations and 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 health nurses during the rst year of the follow-up (1, 3, 6 and 9 months after randomization), and of three visits during the second year (15, 18 and 21 months after the randomization). At these visits, the participants were systematically instructed to change their health behav- iour primarily on the basis of their individual situation. At each visit, BP and weight were measured, and the values, as well as the changes in lifestyle factors to be reached before the next study visit, were written down using a special follow-up card designed for the study. A written feedback of the 4-day food record was sent to the public health nurse to support the intervention. In addition, a 2-h group session was organized for the intervention group in every health care centre at 6 and 18 months after the randomization. These two group meetings concentrated mainly on advice targeting re- duction of salt intake and overweight. During the 2- year follow-up, the participants in the usual care group were instructed to visit their own physicians and public health nurses according to usual practices. Statistical analyses Statistical analyses were conducted with SPSS for Windows version 10.0 (SPSS Inc., Chicago, Illinois, USA). For continuous variables, the t-test was used to test the differences and changes in mean values be- tween the groups. Condence intervals (CI) for the differences in proportions were calculated using a special 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 the case of missing responses, the last observed response was used when calculating the 1- and 2-year changes in continuous variables (the carry-forward method). The same method was used with dichotomous variables. In a separate analysis of BP changes in subjects without previous antihypertensive drug treatment, the last BP measurement without antihypertensive drug treatment was used if drug treatment was initiated during the trial. Accordingly, in subjects already on antihyperten- sive medication, the last BP measurement with anti- hypertensive drug treatment was used if the treatment was discontinued during the trial. In the calculations of BP changes, the changes in doses of antihypertensive medication were not taken into account. Multiple linear regression analysis was used to examine the associations of changes in body weight, sodium and potassium excretion, alcohol intake and leisure-time physical activity (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 Measurements at baseline, 12 months and 24 months: PN, FFR PN PN, GM PN FFR PN PN, GM PN Intervention 0 1 3 6 9 12 15 18 21 24 Usual care* BP Weight Height (only at baseline) Waist Hip Total cholesterol HDL cholesterol LDL cholesterol Triglycerides 24-hour urinary sodium excretion 24-hour urinary potassium excretion Insulin Questionnaires 4-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 own
public health nurses as usual. BP, blood pressure; HDL, high-density lipoprotein; LDL, low-density lipoprotein. Non-pharmacological treatment of hypertension Kastarinen et al. 2507 was not reached due to relatively numerous dropouts before the randomization among the already recruited subjects. It was estimated that this sample size would enable detection of a 3.2 mmHg difference in change of SBP and a 1.6 mmHg difference in change of DBP between the intervention and usual care groups, with 80% power at the 5% signicance level. Results Baseline characteristics and adherence to treatment The mean age of all participants was 54.3 years. Of the participants, 52% were on antihypertensive drug treat- ment at the beginning of the trial (Table 1). There were no statistically signicant differences between the groups in any of the baseline variables analysed. Attendance rates at the 1-year and 2-year study visits were satisfactory (Fig. 2). The attendance rate at both group meetings organized for the intervention group after 6 and 18 months of intervention was 50%. The subjects who dropped out during the different phases of the study were, at baseline measurements, signi- cantly younger (50 versus 55 years) and heavier (83.1 versus 80.1 kg) compared with the attenders of the 2- year visit. Also, alcohol use (73 versus 42 g/week) and proportion of smokers (13 versus 7%) were signicantly higher among the drop-outs at the baseline. Changes in blood pressure and lifestyle factors The changes in BP and other continuous variables in the two groups are shown in Table 2. Without taking into 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 at 1-year follow-up and at 2-year follow-up and in DBP from 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 mean 6SD or percentage Intervention (n 360) Usual care (n 355) Age (years) 54.4 10.1 54.2 9.9 Female (%) 52 54 Antihypertensive drug treatment (%) 52 53 Current smoker (%) 9 7 History of coronary heart disease (%) 5 3 Moderate physical activity at least three times per week (%) 51 51 Alcohol consumption (g/week) 47 83 48 74 Body weight (kg) 81.1 15.7 80.0 14.8 Body mass index (kg/m 2 ) 28.9 4.6 28.5 4.5 Waist circumference (cm) 97.2 13.1 95.8 12.8 Hip circumference (cm) 104.7 10.4 104.1 10.0 Systolic blood pressure (mmHg) 149 16 148 16 No antihypertensive drug treatment 152 14 150 14 Antihypertensive drug treatment 147 18 146 18 Diastolic blood pressure (mmHg) 91 9 91 8 No antihypertensive drug treatment 93 8 93 9 Antihypertensive drug treatment 89 9 89 8 Total cholesterol (mmol/l) 5.66 0.91 5.59 0.93 LDL-cholesterol (mmol/l) 3.64 0.81 3.56 0.79 HDL-cholesterol (mmol/l) 1.32 0.33 1.36 0.38 Triglycerides (mmol/l) 1.56 1.01 1.49 1.00 24-h urinary sodium excretion (mmol) 146 56 142 56 24-h urinary potassium excretion (mmol) 83 27 83 28 Serum 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 Intervention n 360 Usual care n 355 Attended 1-year study visit 317/360 (88%) Attended 1-year study visit 275/355 (77%) Attended 2-year study visit 304/360 (84%) Attended 2-year study visit 283/355 (80%) Fig. 2 The numbers of participants involved throughout the trial. 2508 Journal of Hypertension 2002, Vol 20 No 12 level of statistical signicance. In the subgroup with no antihypertensive drug treatment, the reductions in both SBP and DBP were signicantly greater in the inter- vention group compared with the usual care during both the rst and the second year of follow-up (Table 3). In subjects with antihypertensive drug treatment at baseline, 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 change signicantly in either of the randomized groups during the follow-up. In both groups, 70% of the drug-treated patients 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 more in the intervention group during the rst year (net change 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 groups during 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/m 2 ) assigned to intervention had achie- ved normal weight at the end of the trial, which was signicantly more than that in the usual care group (Table 4). Also, the waist and hip circumferences fell signicantly more in the intervention group than in usual care throughout the study. The changes in 24-h urinary sodium and potassium excretion were small, with no signicant differences between the groups. Self-reported alcohol consumption fell signicantly more during the rst study year in the intervention group, but this difference disappeared during the second year. Compared to the usual care group, a signicantly larger proportion of the participants in the intervention group had increased their physical activity to the target level at both 1-year and 2-year visits. The net reduction in weight between the randomized groups was signicantly greater in subjects with no antihypertensive drug treatment during the rst year compared to the group with antihypertensive drug treatment (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 01 year 02 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 treatment status 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 152 150 147 146 Change 01 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 02 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 01 year 2.6 (4.7, 0.5) 0.8 (3.4, 1.9) Net change 02 years 2.4 (4.7, 0.0) 1.3 (4.5, 1.8) Diastolic blood pressure (mmHg) Baseline mean 93 93 89 89 Change 01 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 02 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 01 year 2.7 (4.0, 1.4) 0.8 (2.3, 0.6) Net change 02 years 2.0 (3.4, 0.6) 0.1 (1.8, 1.6) CI, condence interval. Non-pharmacological treatment of hypertension Kastarinen et al. 2509 term 0.021), but such a difference was not detected in the analysis from baseline to 2 years. Similarly, the net reductions in alcohol consumption from baseline to 1-year and 2-year visits were signicantly greater in this group compared to the group with antihypertensive drug treatment (from baseline to 1-year visit 17 versus 1 g/week, P for the interaction term 0.025; from base- line to 2-year visit 9 versus 5 g/week, P for the interaction 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 the baseline age. In a multiple linear regression analysis, a positive relation with 2-year change in blood pressure was found for weight change in both subjects with no antihypertensive 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 among the subjects with no antihypertensive drug treatment, adjusting for both BP and weight at baseline, the estimated effect of the 1 kg weight lost at 2 years was a reduction of 0.55 mmHg in SBP and of 0.50 mmHg in DBP Changes in other cardiovascular risk factors After 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 during the study, the net reductions in total cholesterol (0.11 mmol/l; 95% CI 0.20 to 0.02) and in LDL cholesterol (0.13 mmol/l; 95% CI 0.22 to 0.03) were larger 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 observed during the study. The serum insulin concentration de- creased more among the persons assigned to interven- tion, but the net change was not statistically signicant. Discussion The lifestyle counselling provided in the LIHEF study could 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 was signicantly greater in the intervention group at the 1- year visit, and this difference was maintained during the second year. The intervention programme could not induce any signicant reductions in salt intake. Despite the fact that in 95% of the participants the weekly alcohol consumption was already at the recom- mended level at baseline, a small but signicant reduc- tion in alcohol intake occurred in the intervention group during the rst year. The self-reported leisure- time physical activity increased signicantly more in the intervention group throughout the study. Large-scale randomized trials reporting the effects of lifestyle intervention on BP and other cardiovascular risk factors in hypertensive persons in the primary care setting are rare. The very few previous studies have included relatively small numbers of patients, with the maximum follow-up time of 12 months [1517]. Iso et al. [18] reported that a community-based trial lowered BP among hypertensive subjects. In their study, the intervention was based mainly on group sessions instead of the individualized health counselling used in our study. The net changes in BP and lipid levels detected in our study are in accordance with trials of multiple risk factor intervention or dietary intervention with people at high risk but not necessarily hypertensive [1921]. The reduction achieved in body weight after 2 years of intervention was almost the same as in some trials with more intensive intervention [1]. In contrast, the BP reduction observed in the subjects without antihyper- tensive drug treatment was smaller compared to some clinical trials of non-pharmacological treatment of hypertension [4,22]. These trials used very intensive intervention compared to our study. Thus it seems that at least some of the results obtained in high-intensity intervention trials can be translated successfully to primary 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 goals at baseline Year 1 Year 2 Intervention a Usual care a Difference in change (95% CI) Intervention Usual care Difference in change (95% CI) BMI , 25 kg/m 2 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 at least three times per week and at least to 30 minutes. 2510 Journal of Hypertension 2002, Vol 20 No 12 this study, the proportion of fat, and especially of saturated fats, in total energy intake decreased signi- cantly more in the intervention group compared to usual care [12]. Also, the total energy intake tended to decrease more in the intervention group, although not reaching the level of statistical signicance. In addition to the increase in physical activity, these changes in diet may have contributed to the observed differences in changes of body weight, lipid levels and BP between the randomized groups. The dietary data were in accordance with the results of the 24-h urinary sodium excretion, showing no signicant changes in sodium intake. These results repeat the ndings of the many other studies that have demonstrated the difculties in achieving the recommended level of salt intake in free- living subjects [23,24]. It has been suggested that the main reason for this difculty in salt restriction seen in all Western countries is the still relatively high concen- tration of salt in processed foods [25]. The differences in BP reduction observed between the groups could not be explained by accustomization with BP measurement, because there was not any difference in self-reported frequency of BP measurements be- tween the groups during the study. One explanation for the greater fall in blood pressure among the participants who continued antihypertensive drug treatment com- pared with the participants without antihypertensive drugs could be a more regular use of antihypertensive drugs 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 are usually more susceptible to accept the recommended intervention than the population at large. On the other hand, many volunteers in lifestyle intervention studies have already previously changed their lifestyle, which could 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 sub- jects in the population-based FINRISK study in 1997 [2628]. The study participants also came from a geographical area with a long history of cardiovascular disease prevention activities, and thus many of them already had a relatively good knowledge about lifestyle factors affecting the cardiovascular risk [29]. In addition, the contamination of the control group, due to the fact that their follow-up visits were done by the same nurses as with the intervention group, might have reduced the difference 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 usual in 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 the trials with more intensive interventions. However, the principal aim of this study was to nd out the extent to which lifestyle intervention will work in the usual primary health care setting. From this point of view, our results were quite satisfactory, considering the limited requirement for the use of health care re- sources. The potential of the intervention shown in this study can certainly be much improved by further development and systematic dissemination, especially concerning newly detected hypertensive persons. Non- pharmacological treatment of hypertension has been advocated for a long time, but so far only limited evidence and experience has been available as to its effective implementation within primary care. The task is not easy, due to the limited time and resources that the public health service can allocate for such preven- tive services. However, we have shown that this approach works. Acknowledgements The statistician Pirjo Halonen, MSc, advised M. Kastar- inen about data analysis. Mr Veli Koistinen was respon- sible for preparing the database. Study nurses Anneli Mitrunen and Mari Aalto screened the study subjects and performed BP and anthropometrical measurements. Registered dietitians Sari Aalto and Sointu Lassila were responsible for the training of nurses in dietary issues and for the group sessions organized for subjects assigned to intervention. We thank for the staff of the North Karelia Project for the help given in coordination of 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 members of the LIHEF study group and helped in designing the study. The authors are grateful to all practice staff working in the participating health care centres, as well as to the hypertensive persons participating the study. References 1 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 the Trials of Hypertension Prevention, phase II. Ann Intern Med 2001; 134: 111. 2 Puddey IB, Beilin LJ, Vandongen R. Regular alcohol use raises blood pressure 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 Dietary Approaches 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 Hum Hypertens 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 of Mild Hypertension Study Research Group. JAMA 1993; 270:713724. Non-pharmacological treatment of hypertension Kastarinen et al. 2511 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 blood pressure. 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 OrganizationInternational Society 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 working party 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 MONICA project. http:www.ktl./publications/monica/bp/bpqa.htm. [Accessed 13 March 2002]. 12 Korhonen M, Kastarinen M, Uusitupa M, Puska P, Nissinen A. The effect of intensied diet counselling on the diet of hypertensive subjects in primary health care. A 2-year open randomised controlled trial of lifestyle intervention against hypertension in Eastern Finland (LIHEF). Prev Med 2002, (in press). 13 Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative 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. Dietary treatment of patients with mild to moderate hypertension in a general practice: a pilot intervention study (1). The rst three months. J Hum Hypertens 1990; 4:368371. 16 Cohen MD, DAmico FJ, Merenstein JH. Weight reduction in obese hypertensive patients. Fam Med 1991; 23:2528. 17 Woollard J, Beilin L, Lord T, Puddey I, MacAdam D, Rouse I. A controlled trial of nurse counselling on lifestyle change for hypertensives treated in general 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-year randomised controlled trial. Hypertension 1996; 27:968974. 19 Ebrahim S, Smith GD. Systematic review of randomised controlled trials of multiple 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. Behavioural counselling in general practice for the promotion of healthy behaviour among 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 of hypertension in older persons: a randomized controlled trial of non- pharmacologic interventions in the elderly (TONE). TONE Collaborative Research Group. JAMA 1998; 279 (11):839846. 23 Pietinen P, Tanskanen A, Nissinen A, Tuomilehto J, Puska P. Changes in dietary habits and knowledge concerning salt during a community-based prevention programme for hypertension. Ann Clin Res 1984; 16 (suppl 43):150155. 24 Evers SE, Bass M, Donner A, McWhinney IR. Lack of impact of salt restriction advice on hypertensive patients. Prev Med 1987; 16:213220. 25 Kumanyika S. Behavioral aspects of intervention strategies to reduce dietary 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 hypertensive and normotensive Finnish population from 1982 to 1997. J Hypertens 2000; 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 and normotensive 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 1982 1997. 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. 2512 Journal of Hypertension 2002, Vol 20 No 12