GHA 1654-9716 1654-9880 Global Health Action Vol. 4, September 2011, pp. 1–7 sici xxx PUBMED Abbreviation Yulia Blomstedt

*Yulia Blomstedt Epidemiology and Global Health Dept. Public Health and Clinical Medicine Umeå University SE-901 85 Umeå Sweden 46 (0)90 785 526 77

Maria Emmelin Lars Weinehall Department of Public Health and Clinical Medicine, Epidemiology and Global Health Umeå University

Umeå Sweden

Umeå Centre for Global Health Research Umeå University

Umeå Sweden

Centre for Population Studies, Ageing and Living Conditions Programme Umeå University

Umeå Sweden

Global Health Action The Västerbotten Intervention Programme What about healthy participants? The Västerbotten Intervention Programme (VIP) addresses cardiovascular disease and diabetes in the middle-aged population of Västerbotten County, Sweden. Self-reported health (SRH) is one of the risk factors for both conditions. The aim of this study was to analyse the development patterns of SRH among the VIP participants. Cross-sectional data from 1990 to 2007 were used to analyse the prevalence of poor SRH among 101,396 VIP participants aged 40–60 years. Panel data were used to study the change in SRH among 25,695 persons aged 30–60 years, who participated in the VIP twice within a 10-year interval. Prevalence of poor SRH fluctuated between 1990 and 2007 in Västerbotten County. There was a temporary decline around 2000, with SRH continuously improving thereafter. The majority of panel participants remained in good SRH; over half of those with poor or fair SRH at baseline reported better SRH at follow-up. SRH declined in 19% of the panel participants, mostly among those who had good SRH at the baseline. The decline was common among both women and men, in all educational, age and marital status groups. The SRH improvement among those with poor and fair SRH at baseline suggests that VIP has been successful in addressing its target population. However, the deterioration of SRH among 21% of the individuals with good SRH at baseline is of concern. From a public health perspective, it is important for health interventions to address not only the risk group but also those with a healthy profile to prevent the negative development among the seemingly healthy participants. self-reported health intervention Västerbotten Sweden All Västerbotten residents are invited to participate in the VIP upon becoming 40, 50 or 60 years old, thus having a chance to participate in the VIP once every 10 years. Individuals aged 30 were also invited to participate until 1995. The participation is voluntary and the participation rate is typically about 60%. VIP is conducted by medical personnel at local healthcare centres. First, some biometrical measurements are taken and then the participants fill in a questionnaire concerning their socioeconomic status, SRH, life satisfaction, social support, lifestyle (eating habits, alcohol intake and smoking, physical activity), personal history and family history of CVD and diabetes. Finally, individual counselling by means of a motivational interview based on the results of the medical examination and the questionnaire takes place regarding individual's risk factors and required lifestyle changes. A more detailed description of the VIP survey design can be found elsewhere 11. The current study was based on the VIP questionnaire data obtained during 1990–2007 and analysed both cross-sectionally and longitudinally (a 10-year panel). The cross-sectional sample included 101,396 individuals who were 40, 50 or 60 years old and participated in the VIP at least once during 1990–2007. The 30-year olds were excluded since data were not available for the entire study period. The panel analysis included all 25,695 individuals aged 30, 40 or 50 years at the initial survey (1990–1997) who participated in a 10-year follow-up study when they turned 40, 50 or 60 years, respectively, during 2000–2007. SRH was determined by response to the question ‘How would you describe your health status during the past year: very good, rather good, fair, rather poor or poor?’ Individuals who responded ‘very good’ or ‘rather good’ were considered to have ‘good’ SRH. Those who responded ‘fair’ were considered to have ‘fair’ SRH and those who responded ‘rather poor’ or ‘poor’ were considered to have ‘poor’ SRH. SRH at follow-up was compared with SRH at baseline. Those who reported the same SRH at follow-up as at baseline were categorised as ‘remained’; those who reported poorer SRH at follow-up were categorised as ‘deteriorated’; and those who reported better SRH at follow-up were categorised as ‘improved.’Sex was categorised as ‘man’ or ‘woman.’ Age was categorised as ‘30’, ‘40’, ‘50’ or ‘60’ years. Educational level was categorised into ‘basic’ (elementary and comprehensive compulsory school), ‘mid-level’ (residential college for adult education or high school) and ‘high’ (college or higher) educational levels. Marital status was dichotomised into ‘married/cohabiting’ with or without children or ‘single’ with or without children. The latter group also included divorced, separated or widowed individuals. Marital status at follow-up was compared with marital status at baseline. Respondents were categorised accordingly into ‘remained married/cohabiting,’ ‘remained single,’ ‘became single’ or ‘became married/cohabiting.’ Prevalence of good, fair and poor SRH was determined over 18 years of observations in the cross-sectional sample (n = 101,396). Ten-year changes in SRH were evaluated using the panel data (n=25,695). Individuals with incomplete information on SRH and/or the demographic variables of interest listed earlier comprised approximately 2% of the samples and were excluded. All tables and figures were adjusted for sex, age, educational level and marital status, unless otherwise stated. The differences between the groups were assessed with the Pearson's chi-square test (p< 0.05). Data analyses were performed using STATA Version 10.1 15. This study was approved by the Regional Research Ethics Board in Umeå, Sweden (Dnr 08-131 M). Individuals provided written consent prior to each survey and health screening. The results of the cross-sectional and panel data analyses are presented separately in the following sections. The descriptive characteristics of both samples are presented in Table 1. Sample characteristics Cross-sectional sample (N=101,396) Panel sample baseline (N=25,695) Panel sample follow-up (N=25,695) Sex  Men 48.2% (48,917) 46.7% (11,997) 46.7% (11,997)  Women 51.8% (52,479) 53.3% (13,698) 53.3% (13,698) Age (years)  30 – 19.4% (4,992) –  40 33.9% (34,392) 40.4% (10,375) 19.4% (4,992)  50 35.7% (36,205) 40.2% (10,328) 40.4% (10,375)  60 30.4% (30,799) – 40.2% (10,328) Marital status  Married/cohabiting 81.3% (82,440) 84.2% (21,629) 82,4% (21,184)  Single 18.7% (18,956) 15.8% (4,066) 17.6% (4,511) Education  High 25.0% (25,359) 21.2% (5,436) 19.9% (5,113)  Mid-level 49.1% (49,766) 54.8% (14,071) 52.9% (13,582)  Basic 25.9% (26,271) 24.1% (6,188) 27.2% (7,000) Self-reported health  Good 71.3% (72,256) 78.4% (20,145) 71.2% (18,303)  Fair 21.6% (21,929) 16.8% (4,325) 21.6% (5,549)  Poor 7.1% (7,211) 4.8% (1,225) 7.2% (1,843) On average, 71% of individuals reported good SRH, 22% reported fair SRH and 7% reported poor SRH in 1990–2007 (Table 1). Poor SRH was more likely to be reported by women than men, older than younger adults, single than married individuals and by those in the lowest educational category (Table 2). Self-reported health (%) Sample characteristics Good Fair Poor *Significant difference between the groups (p<0.05) Sex*  Men 73.5 20.8 5.7  Women 69.2 22.4 8.4 Age* (year)  40 75.0 18.8 6.2  50 72.7 20.2 7.1  60 65.4 26.5 8.1 Marital status*  Married/cohabiting 72.7 20.7 6.6  Single 64.9 25.9 9.2 Education*  High 75.7 17.9 6.5  Mid-level 71.9 21.2 6.9  Basic 65.5 26.4 8.1 There was deterioration in SRH around 2000 for both men and women of all educational levels. The deterioration remained significant after adjustment for age and marital status (Fig. 1). The fall in good SRH was most pronounced among women with basic education. The recovery was gradual, and in 2007, the prevalence of good SRH was still lower than in 1990 among all groups.

Self-reported health in men and women by educational level, Västerbotten County, 1990–2007 (the cross-sectional sample N=101,396). The figure is adjusted for age and marital status.

At follow-up, 70% of respondents reported the same SRH as at baseline, 19% reported poorer SRH and 11% reported better SRH. The most pronounced deterioration was among persons with good SRH at baseline (Table 3). In general, 21% (n = 4,236) of individuals with baseline good SRH had poorer SRH at follow-up (fair or poor). On the other hand, 71% of those with poor SRH at baseline reported better SRH at follow-up. However, this group constituted only 3% (869) of the panel sample. Self-reported health at 10-year follow-up,% (n) Self-reported health at baseline Remained Deteriorated Improved *Significance p< 0.05 Good 79.0% (15,909) 21.0% (4,236) – Fair 39.5% (1,706) 13.6% (590) 46.9% (2,029) Poor 29.1% (356) – 70.9% (869) The deterioration of SRH was significantly (p< 0.05) more prevalent in women than in men, in older than in younger adults, among those with basic education than among those with high- or mid-level education at baseline (Table 4). Baseline characteristics Prevalence of deteriorated self-reported health,% (n) *Significant difference between the groups (p< 0.05) Sex *  Men 16.7% (2,002)  Women 20.6% (2,824) Age * (years)  30 18.2% (908)  40 17.8% (1,845)  50 20.1% (2,073) Marital status  Married/cohabiting 18.7% (4,048)  Single 19.1% (778) Education *  High 17.2% (1,064)  Mid-level 18.5% (2,596)  Basic 21.5% (1,166) Over 3,500 persons had changed their marital status during the follow-up period. We, therefore, analysed the association between SRH deterioration and change in marital status. Change in marital status was associated with change of SRH. Interestingly, any change of marital status (becoming married/cohabiting or single/divorced) was associated with a greater deterioration of SRH than remaining in the same marital status at follow-up. As shown in Fig. 2, becoming single was associated with greater deterioration in younger than in older adults, and in women more than men. However, even a change to being married/cohabiting was associated with greater deterioration of SRH in men who turned 60 at follow-up.

Deteriorated self-reported health at 10-year follow-up (the panel sample N=25,695). This figure is stratified by follow-up sex, age and change in marital status and is adjusted by educational level at follow-up.

SRH of middle-aged men and women fluctuated between 1990 and 2007 in Västerbotten County. There was a temporary deterioration in the overall SRH around 2000, but this trend reversed, and SRH continuously improved thereafter. Panel data showed that the majority of the VIP participants remained in good SRH at follow-up and over half of those with poor or fair SRH at baseline reported better SRH at follow-up. SRH declined in 19% of participants, mostly among those who had good SRH at baseline. The decline was common among both women and men, in all educational, age and marital status groups. The overall SRH level observed in the cross-sectional data (71% good, 22% fair and 7% poor) corresponded well with national figures. According to Swedish National Statistical Bureau, 75% of adults aged 18–84 years had good SRH and 6% had poor SRH in Sweden during 1996–2005 16. Also, in line with previous research 17 18, women, individuals with basic education, single and older persons had poorer SRH than their comparison groups. The SRH deterioration around 2000 was not unique for Västerbotten County. A similar temporary decline in SRH was observed in Stockholm County 19 20 and at the national level 21. According to the National Board of Health and Welfare, self-reported prevalence of psychological problems, pain, sleeping problems and long-standing illness also increased during the same period across the whole country 21. An increase in the number of people on long-term sickness pension, disability pension and sickness allowance in Sweden and in Västerbotten County, in particular, were also seen at this time 22–24. However, the causality behind this decline and the consequent improvement of SRH are yet to be understood. These questions are addressed in detail elsewhere by Blomstedt et al 25. The panel data showed that the majority of the VIP participants remained in the same SRH at follow-up as at baseline. Moreover, 71% of those with poor SRH and 47% of those with fair SRH at baseline experienced improvement in their SRH. In total, 11% (n=2,898) of the sample (n=25,695) had better SRH at follow-up. This is in line with a previous study of SRH changes in a sub-sample of the VIP population followed between 1986 and 1996 13. The panel data also revealed a deterioration of SRH among 17% of men and 21% of women over a 10-year period. The deterioration was common for all respondents, but was more pronounced among the 50 years old and those with the basic educational level at baseline. The change of marital status was also associated with the change in SRH. Becoming single was associated with a decline in SRH among younger persons (30 years old at baseline), and surprisingly, becoming married/cohabiting was also associated with a decline in SRH, mainly in older men (50 years at baseline). The decline of SRH with age is well-known and has been widely discussed in the literature 26 27. In line with our observations, other follow-up studies in European populations have also shown a decline of SRH with time 28 29. A previous 10-year follow-up study of a VIP sub-sample (n=174) found the same deterioration rate. Nineteen percent of those with good SRH at baseline in 1986 reported fair or poor SRH in 1996 13. SRH at baseline is usually the strongest predictor of SRH at follow-up 30. Unexpectedly, in our study, the proportion of participants with poorer SRH at follow-up was higher among those who had good rather than fair SRH at baseline (21% vs. 14%). This is a concern and suggests the possibility that those who were healthier at the baseline did not benefit from the intervention programme as much as those who were less healthy. Poor SRH is a risk factor itself, as well as one that accompanies CVD and diabetic conditions. Being designed to target persons at risk for developing or already having CVD or diabetes, VIP is successful in helping persons to reduce their risk factor burden and improve their SRH 13 31–34. This is also confirmed in the current study by the improvement in SRH of the majority of those with poor SRH and almost half of those with fair SRH at baseline. Since 2009, the healthcare personnel within VIP have been specifically educated to give particular attention to persons who have poor or fair SRH along with other CVD risk factors. This recommendation is based on findings from Weinehall et al 35 and Emmelin et al 36, which show that poor SRH dramatically increases the risk for acute myocardial infarction and stroke among persons with other CVD risk factors. However, such focus on CVD risk factors and poor SRH may cause an unforeseen and undesired side effect. Those with good SRH at baseline might feel left out and exhibit deterioration in their SRH at follow-up. These results should not undermine the success of VIP in addressing the target population as illustrated in earlier studies 13 31–34. Moreover, earlier VIP studies have shown that SRH deterioration among VIP participants is slower than in the reference area 37. Nevertheless, our findings raise the question of how to best design a health intervention to reach the whole population of a defined geographical area. According to our analysis, 21% of those with good SRH at baseline deteriorate over 10 years. Since poor SRH is a proxy of general health and may reflect preclinical manifestations of disease 38 39, we need to consider whether it is possible to prevent this negative development among seemingly healthy participants as part of the VIP. Our analysis also illustrated that other factors (e.g. marital status) influence SRH. This shows that deterioration of SRH is not only associated with lifestyle risk factors or chronic conditions such as CVD or diabetes, even if lifestyle is a strong predictor of a future SRH 28. Therefore, it is important to emphasise that from a public health perspective, VIP should use its influence to target not only those who are clearly at risk for developing CVD or diabetes and have poor SRH but also those with good SRH at baseline. The methods of addressing such participants (e.g. confirmation, encouragement) are not discussed here, but should be considered elsewhere. This argument is supported by a qualitative sub-study of VIP participants done in 1995–1996. That study found a desire among the participants to be seen and confirmed by the VIP personnel even if they did not have any of the risk factors that VIP was designed to target. Otherwise, they felt disappointed and left out – feelings that increase their risk of moving from a low to high CVD risk factor load 13. The analysis of the SRH development among the VIP participants during 1990–2007 in Västerbotten County, Sweden, suggested a positive effect of the VIP. The majority of the VIP participants remained in good SRH, and the majority of those with fair or poor SRH at baseline revealed improvement in their SRH at 10-year follow-up. Nevertheless, a fifth of the VIP participants, mostly those with good SRH at baseline, had deterioration in SRH at a 10-year follow-up. These findings raise a question of how to best design a health intervention to reach the whole population of a defined geographical area. From a public health perspective, it is important to support health maintenance and prevention of future disease in those with a healthy profile as well as the high-risk individuals. The authors have not received any funding or benefits from industry or elsewhere to conduct this study. This study was funded by FAS, the Swedish Council for Working Life and Social Research (Grant no. 2006-1512) and the Swedish Research Council (Grant no. 2008-6592). This work was undertaken at the Umeå Centre for Global Health Research, funded by FAS, the Swedish Council for Working Life and Social Research (Grant no. 2006-1512), and within the Ageing and Living Conditions Programme, funded by a Linnaeus grant from the Swedish Research Council (Grant no. 2008-6592). All authors contributed to discussion and presentation of results as well as revision of the manuscript. YB conducted the data analyses. The funding sources were not involved in any parts of the study conduct or data analyses. Manderbacka K Lundberg O Martikainen P. Do risk factors and health behaviours contribute to self-ratings of health? Soc Sci Med 1999 48 1713 20 10.1016/S0277-9536(99)00068-4 Lundberg O Manderbacka K. Assessing reliability of a measure of self-rated health. Scand J Soc Med 1996 24 218 24 Idler EL Benyamini Y. Self-rated health and mortality: a review of twenty-seven community studies. J Health Soc Behav 1997 38 21 37 Fayers PM Sprangers MAG. Understanding self-rated health. Lancet 2002 359 187 8 DeSalvo KB Predicting mortality and healthcare utilisation with a single question Health Serv Res 2005 40 1234 46 10.1111/j.1475-6773.2005.00404.x Orth-Gomer K Unden AL Edwards ME. Social isolation and mortality in ischemic heart disease. A 10-year follow-up study of 150 middle-aged men. Acta Med Scand 1988 224 205 15 Pijls LT Feskens EJ Kromhout D. Self-rated health, mortality, and chronic diseases in elderly men. The Zutphen Study, 1985–1990 Am J Epidemiol 1993 138 840 8 Appels A Bosma V Grabauskas A. Self-rated health and mortality in a Lithuanian and a Dutch population. Soc Sci Med 1996 42 681 9 Moller L Kristensen TS Hollnagel H. Self rated health as a predictor of coronary heart disease in Copenhagen Denmark. J Epidemiol Community Health ’ 1996 50 423 8 10.1136/jech.50.4.423 Heistaro S Self rated health and mortality: a long term prospective study in eastern Finland J Epidemiol Community Health 2001 55 227 32 10.1136/jech.55.4.227 Norberg M , Boman K , Weinehall L. The Västerbotten Intervention Programme: Background, design and implications. Global Health Action 2010; 3 . DOI: 10.3402/gha.v3i0.4643. Kottke TE Changes in perceived heart disease risk and health during a community-based heart disease prevention programme: the North Karelia project Am J Public Health 1984 74 1404 5 10.2105/AJPH.74.12.1404 Emmelin M Weinehall L Stenlund H Wall S Dahlgren L. To be seen, confirmed and involved--a ten year follow-up of perceived health and cardiovascular risk factors in a Swedish community intervention programme BMC Public Health 2007 7 190 10.1186/1471-2458-7-190 Norberg M , . Changing patterns of tobacco use in a middle-aged population – the role of snus, gender, age, and education. Glob Health Action 2011; 4. StataCorp. Stata Statistical Software: Release 10.1. College Station TX : Stata Corporation; 2009 . Population Statistics, part 3. Örebro : Statistics Sweden; 2007 . Undén A-L Elofsson S. Do different factors explain self-rated health in men and women? Gender Medicine 2006 3 295 308 10.1016/S1550-8579(06)80218-4 af Sillen U Self-rated health in relation to age and gender: influence on mortality risk in the Malmo Preventive Project Scand J Public Health 2005 33 183 9 10.1080/14034940410019235 Public Health Report. Stockholm : National Board of Health and Welfare; 2007 . Burström K Johannesson M Rehnberg C Deteriorating health status in Stockholm 1998–2002: results from repeated population surveys using EQ-5D Qual Life Res 2007 16 1547 53 10.1007/s11136-007-9243-z Public Health Report ( ). Stockholm : National Board of Health and Welfare; 2005 . Statistical information. Days covered by sickness allowance, December 1998 – June 2002. Stockholm : The Social Insurance Office; 2002 . Statistical information. Diseases that caused retiring. Stockholm : The State Insuarance Office; 2004 . Lidwall U Marklund S Thoursie PS. Sickness absence in Sweden Worklife and health in Sweden Gustafsson RÅ Lundberg I National Institute for Working Life Stockholm 2004 183 204 Blomstedt Y , Blomstedt P , Janlert U , Weinehall L. The millennium shift: might a symbolic date affect the level of self-reported health in a modern secularised society? (submitted). Andersen FK Christensen K Frederiksen H. Self-rated health and age: a cross-sectional and longitudinal study of 11,000 Danes aged 45–102 Scand J Public Health 2007 35 164 71 10.1080/14034940600975674 Chen H Cohen P Kasen S. Cohort differences in self-rated health: Evidence from a three-decade, comunity-based, longitudinal study of women Am J Epidemiol. 2007 166 439 46 10.1093/aje/kwm100 Borg V Kristensen TS. Social class and self-rated health: can the gradient be explained by differences in life style or work environment? Soc Sci Med 2000 51 1019 30 10.1016/S0277-9536(00)00011-3 Disney R Emmerson C Wakefield M. Ill health and retirement in Britain: a panel data-based analysis J Health Econ 2006 25 621 49 10.1016/j.jhealeco.2005.05.004 Miller TR Wolinsky FD. Self-rated health trajectories and mortality among older adults J Gerontol B Psychol Sci Soc Sci 2007 62 S22 7 Weinehall L Shifting the distribution of risk: results of a community intervention in a Swedish programme for the prevention of cardiovascular disease J Epidemiol Community Health 1999 53 243 50 10.1136/jech.53.4.243 Weinehall L Can a sustainable community intervention reduce the health gap? 10-year evaluation of a Swedish community intervention programme for the prevention of cardiovascular disease Scand J Public Health Suppl 2001 56 59 68 Jansson JH Boman K Messner T. Trends in blood pressure, lipids, lipoproteins and glucose metabolism in the Northern Sweden MONICA project 1986–99 Scand J Public Health Suppl 2003 61 43 50 10.1080/14034950310001397 Lindahl B Trends in lifestyle1986 –1999 in a 25- to 64-year-old population of the Northern Sweden MONICA project Scand J Public Health Suppl 2003 61 31 7 10.1080/14034950310001414 Weinehall L Perceived health modifies the effect of biomedical risk factors in the prediction of acute myocardial infarction. An incident case-control study from northern Sweden J of Internal Med 1998 243 99 107 10.1046/j.1365-2796.1998.00201.x Emmelin M Self-rated ill-health strengthens the effect of biomedical risk factors in predicting stroke, especially for men – an incident case referent study J of Hyperten 2003 21 887 96 10.1097/00004872-200305000-00012 Emmelin M. Self-rated health in public health evaluation (PhD thesis) Umea University Umea 2004 Froom P Melamed S Triber I Ratson NZ Hermoni D Predicting self-reported health: the CORDIS study Prev Med 2004 39 419 23 10.1016/j.ypmed.2004.02.006 Szwarcwald CL Souza-Junior PR Esteves MA Damacena GN Viacava F. Socio-demographic determinants of self-rated health in Brazil Cadernos de saude publica / Ministerio da Saude, Fundacao Oswaldo Cruz, Escola Nacional de Saude Publica 2005 21 Suppl 54 64