Abstract

It has been argued that perceptions of familial tendencies to disease are common and important in decisions about health-related behaviours. Indeed, it has been suggested that the increased `geneticization' of society may lead to an increased fatalism about health, which could undermine initiatives aimed at reducing coronary-prone behaviour. To date, much of the research on lay perceptions of inheritance has been based on people at high risk of particular genetic disorders or on qualitative research with small general population samples. Here we investigate perceptions of a family history of heart disease, using quantitative techniques, to test hypotheses about the relationship between a perceived family history (pFH), coronary `candidacy' and adherence to health promotion advice which were raised by earlier anthropological work. We find that reported perceptions of a family history of heart disease are common, particularly amongst women in middle-age. In isolation a pFH is not related to current smoking; however, the odds of smoking are lower for those with a pFH of heart disease when account is also taken of other attitudinal factors (the `salience' of heart disease and the strength of adherence to conventional coronary health promotion).

Introduction

Coronary heart disease (CHD) is a major cause of morbidity and premature mortality, and is strongly patterned by social class and gender. Health-related behaviours have been identified as important risk factors for coronary disease at both a population and individual level [see (Charlton et al., 1997) for a recent review and (Tunstall-Pedoe et al., 1997) for recent evidence from Scotland]. Throughout the 1980s and early 1990s there was a strong, though much criticised [see (Bunton et al., 1995) for recent reviews], emphasis on individual `lifestyle' in coronary prevention [see, e.g. (Department of Health, 1991, 1992; Scottish Office, 1992)]. However, despite the efforts of health promotion and public health agencies, the targets set at the beginning of the decade for improvements in such health behaviours (Department of Health, 1992; Scottish Office, 1992) are unlikely to be met, particularly for smoking (Dong and Erens, 1997), which remains the most important preventable cause of disease and premature death (Scottish Office, 1999). In Scotland, smoking accounts for more than 13 000 deaths each year and is said to cost the NHS £140 million in treating smoking-related disease (Scottish Office, 1999). Therefore, the reduction of `coronary-prone' behaviour (and more detailed understanding of the barriers to change) remains a priority for public health.

Decisions about health-related behaviour are complex, and are made in the context of an individual's background and life circumstances. One factor that plays a part in this process is knowledge about the personal lives and health experiences of other family members (Davison et al., 1989), and speculation about whether individuals follow the same trajectory through life (Richards, 1993). In this paper we examine perceived family histories of heart disease in relation to smoking (a key behavioural risk factor for coronary disease) and relevant attitudes to health.

Background to current research

The importance of perceived family histories as a factor in decisions about health behaviours was highlighted by anthropological research into the lay epidemiology of heart disease in South Wales (Davison et al., 1992). The research focused on `heart trouble' (a broader lay category which included CHD) and on how notions of a healthy personal `lifestyle' were considered in relation to a broader array of influences on health (Davison et al., 1992). Knowledge of the behavioural risk factors for CHD highlighted by health promotion campaigns was found to be detailed, accurate and `almost universal' (Davison et al., 1992), mirroring Mildred Blaxter's observation that `the public has learned well the lessons of health education' [ (Blaxter, 1990), p. 153]. However, this knowledge was not directly associated with behaviours; not, Davison et al. argued, because of irrationality but because the knowledge of behavioural risk factors was set in a wider personal and social context. Similar observations have been noted elsewhere [see, e.g. (Ritchie et al., 1994; Lupton and Chapman, 1995)] and many investigators have demonstrated that decisions about health-related behaviours take place within a risk balancing framework which takes account of other pathogenic or salutogenic influences, including non-health-related benefits [see, e.g. (Backett, 1992; Graham, 1987; Mullen, 1992; Mullen, 1993)].

Davison et al.'s study also drew out key observations in relation to family history, `candidacy', and the juxtaposition of knowledge about behavioural risk factors and more fatalistic attitudes towards health. The once-held dichotomy between adherence to control of `lifestyle' and `fatalism' has little empirical support; as Pill and Stott observed, `most… respondents were probably neither out and out fatalists nor did they believe that a healthy lifestyle would guarantee complete immunity' [ (Pill and Stott, 1987), p. 127]. Davison et al. argued that `fatalism' is not an alternative to a lifestyle-orientated viewpoint, but that more `fatalistic' explanations (and other notions such as coronary `candidacy' or `proneness') arise out of the routine observations of illness and death made within personal social networks and in the public arena. A common focus is on `anomalies' that are not `explained' by the patterns of behavioural coronary risk factors of the coronary `victim'. The emphasis of `conventional' coronary health promotion [or what has been labelled the `Health of the Nation' public health perspective (Davison and Davey Smith, 1995)] on protection from coronary disease through the adoption of a particular personal lifestyle has thus lead to a `paradoxical effect of highlighting anomalies' [ (Davison et al., 1992), p. 677], both `anomalous deaths' and `unwarranted survivals'. These are illustrated by widespread references to `classic' coronary `candidates' (archetypically someone who is very overweight, smokes, drinks heavily and eats `all the wrong things') who survive to a ripe old age and to the `last person' who would be expected to succumb to coronary disease (archetypically the slim, jogging, health conscious individual who unexpectedly `drops dead' of a heart attack).

One important element of `fate' that was commonly identified was the family that we are `born into'. At least half of the South Wales respondents referred to their `inheritance' from their family unprompted when asked general questions about the differential distribution of health in the community. Crucially in this context, some informants discussed their `inheritance' in their accounts of their own decisions about health-related behaviours. The benefits of individual behavioural change could thus `counteract' the effects of heredity, just as `good' and `bad' elements of one's lifestyle could be balanced against one another. Davison et al. concluded that, for those who made some kind of assessment of their own inherited risk of heart disease, there were four orientations to `lifestyle' choices that they may choose to adopt (although they suggested that only in relatively rare cases did these `logical' possibilities appear as concretely formed plans of action) [ (Davison et al., 1989), p. 339]:

(1)I have inherited a high risk of getting heart trouble, and so I will be especially careful about smoking, weight, food and exercise.

(2)I have inherited a high risk of getting heart trouble, and so I may as well not bother to follow advice about smoking, weight food and exercise.

(3)I have inherited a low risk of getting heart trouble, and so I will build on that by being careful about smoking, weight, food and exercise.

(4)I have inherited a low risk of getting heart trouble, and so I don't have to take any notice about advice about smoking, weight, food and exercise.

In this paper we present analyses of data collected during structured interviews with large general population samples which aim to test quantitatively some of the hypotheses generated by this anthropological research. We examine the relationship between a major coronary behavioural risk factor, smoking, and perceived family history (pFH) of coronary disease, self-ascribed `candidacy' and orientations towards conventional coronary health promotion. Specifically we address the following questions:

(1)How common are perceptions of a family history of heart disease in the population?

(2)Do perceptions of a family history of heart disease relate to other attitudes and beliefs about heart disease aetiology and prevention?

(3)Are perceptions of a family history of heart disease related to smoking?

Method

Subjects

Data are presented from the West of Scotland Twenty-07 study, a longitudinal study of the social patterning of health amongst three age cohorts, aged 15, 35 and 55 when first studied in 1987/88. Respondents were sampled from residents in the Central Clydeside Conurbation, a socially varied but mainly urban area centred on Glasgow. Initial sample sizes in 1987/88 were around 1000 per cohort. Respondents have completed lengthy home-based structured interviews on two subsequent contacts in 1990/91 and 1995/96. Further details on the sample and methods are available elsewhere (Ecob, 1987; Macintyre et al., 1989; Ford et al., 1994; West et al., 1994; Der, 1998).

Data are presented here from the 1995/96 interviews, when 676 people in the youngest cohort (then around 23 years), 754 in the middle cohort (then around 43 years) and 723 people in the oldest cohort (then aged around 63 years) were re-interviewed in their own homes by nurses trained in interview techniques. A wide range of measures of self-reported health and health behaviour, of physical development and functioning, and of personal and social circumstances has been collected at each face-to-face contact. In 1995/96, this included a number of questions that were specifically designed to explore the relationship, if any, between perceptions of family history of illness and health attitudes and health-related behaviours. Respondents were asked about health and deaths in the family (reported numbers of close and more distant family with specified illnesses or conditions, etc.); perceptions of illnesses and weaknesses that run in their own family; their own perceived vulnerability to a number of conditions, including heart disease; and a range of health attitudinal questions that were intended to be related to these perceptions. The questions on family deaths and family health appeared earliest in the questionnaire, and were separated from questions on current health behaviours by another large section of the questionnaire. The main block of questions on perceived family histories and attitudes towards the salience and implications of these appeared towards the end of the questionnaire.

Measures

A summary of the basic descriptive characteristics of the sample is shown in Table I.

Socio-demographic variables

Age (23, 43, 63 years), gender (men, women), housing tenure [owner-occupiers versus others; this was chosen as main indicator of socio-economic status (SES) since many in the youngest cohort were not in the labour force at 23 and had to be classified according to their class of origin].

Perceptions of family history (pFH)

Respondents were asked the following question: `Some people think that particular illnesses or weaknesses run in their families, others don't. Do you think that there are any conditions, weaknesses or illnesses which run in your family?'. Nurses asked those who responded positively which illnesses or weaknesses ran in their families, but were instructed not to probe so that respondents were not presented with any lists of illnesses or conditions. The nurses recorded as many conditions as the respondents mentioned unprompted by allocating their responses to pre-defined codes, including `heart disease or heart trouble'. This was intentionally broad (including both specific references to ischaemic heart disease and angina, for example, and less specific references to `heart trouble' and `heart problems') because previous research and pilot work had indicated that these terms were commonly used when discussing heart disease.

Health attitudinal variables

Respondents were presented with a series of statements.

Factors in the aetiology of heart disease. Respondents were asked to indicate (on a five-point scale) how important they thought various factors were in the aetiology of heart disease; these included `housing and living conditions', `aspects of lifestyle like smoking, drinking and diet', `illnesses and weaknesses that run in the family' and `stress in daily life'.

Adherence to conventional coronary health promotion. Respondents were assigned to three groups (very high, mid and low) according to the strength of their agreement with three statements (five-point scale, from agree strongly to disagree strongly): `If heart trouble runs in someone's family (a) …it is particularly important for them not to smoke; (b) …it is particularly important for them to take exercise; (c) …it is particularly important for them to eat a healthy diet'. Only a few were ambivalent about or disagreed with the statements. Those who strongly agreed with all three statements were assigned to the `very high' adherence group; those who failed to agree (i.e. were ambivalent, disagreed or disagreed strongly) with one or more of the statements were assigned to the `low' adherence group; the remainder who agreed with all of the statements but said `just agree' (rather than `strongly agree') to at least one of the statements were allocated to the `mid' adherence group.

Adherence to `fatalism'. Respondents were asked the extent to which they agreed (five-point scale, from agree strongly to disagree strongly) with the statement: `If heart trouble runs in someone's family it doesn't matter what they do; whether they get heart disease or not is out of their hands'. Those who disagreed strongly were classed as having low, those who agreed (strongly or just) were classified as having high and the remainder as having middling `fatalism'.

Self-ascribed coronary `candidacy'. Respondents were asked how likely they thought they were to get heart disease. Those who said they were very or quite unlikely to develop heart disease or that they did not know are contrasted for the purposes of this analysis with those who said that they were very or quite likely to get it, and those who already had heart disease. For brevity we describe this as `salience' in this paper. Table I shows that, as expected, the salience of heart disease increases with age.

Smoking

Current smokers (one or more cigarettes per day) are contrasted with never or ex-smokers. The prevalence of smokers was around a third in each cohort (Table I).

Analysis

Reports of a perception of family history of illness were examined by age, and by gender and SES within age. The health attitudes of those with and without a pFH of heart disease were also compared within age. A series of logistic regression models examined the relationship between pFH, each of these attitudinal factors and smoking, controlling for SES, sex and age, and then examined the relationship between a pFH and smoking in more detail. Unless otherwise stated, the reference category is the first mentioned category in the description of measures above. We also tested for interactions between a pFH and the socio-demographic and attitudinal variables.

Results

Prevalence of pFH

A substantial proportion, around 40%, reported having some kind of family history of illness (Table II). The prevalence did not differ by cohort nor by SES. Women were significantly more likely to report a family history in the older two cohorts.

Around 20% in each cohort said that they had a family history of heart disease or heart trouble; this was the most commonly mentioned illness. The prevalence did not differ by cohort or by SES, but again women were more likely to report a family history of heart trouble than men. Perception of a family history of heart disease was very strongly related to reported number of relatives who had ever had heart disease (Table III). The vast majority of those who had reported that they had no close family (parents or siblings) with heart disease said they had no family history of heart disease (89, 96 and 97% at ages 23, 43 and 63, respectively). Of those with neither close nor more distant family (grandparents, aunts, uncles and first cousins) with heart disease, 99% said they had no family history. Conversely, 93% of those aged 43 who reported that they had three or more close family with heart disease said they had a family history; this fell to 52% for those with two relatives, 28% with one relative and 3% with none. A similar though less pronounced relationship was seen at age 23 (when many family members are still very young for coronary disease) and at age 63.

Relationship of pFH of CHD to health attitudes

In each cohort a pFH of heart disease was strongly related to self-ascribed candidacy (salience); about twice as many without a pFH described themselves as unlikely to get heart disease (Table IV). Similarly, respondents with a pFH were more likely to have ascribed a greater level of importance to `family illnesses and weaknesses' in the aetiology of heart disease than were those with no pFH (see Table V). Most respondents ascribed a high importance to `lifestyle' (`aspects of lifestyle like smoking, drinking and diet') in the aetiology of heart disease. Fewer dissociated themselves from the view that behavioural risk factors had some effect at the very least than for the other factors, although the propensity to endorse a `lifestylist' line did decrease with age (cohort differences were highly significant, P < 0.001). However, there were still differences between those with and without a pFH of heart disease. The tendency of those with a pFH of heart disease to be even more likely to ascribe a `very' important effect to lifestyle factors (75 and 77% at 23 and 43, respectively) was significant in the younger two cohorts (Table V). It is of interest that the majority of the respondents at all ages still regarded `stress' as being an important cause of heart disease. Indeed it was accorded similar importance to `family illnesses or weaknesses'. It is also interesting that the ratings of the importance of stress did not differ between those with and without a pFH of heart disease (Table V). Housing and living conditions were rated as having less importance in the aetiology of heart disease than `lifestyle', stress and family tendencies, and ratings did not differ in any of the cohorts according to pFH of heart disease.

Table VI shows attitudes to `conventional' coronary health promotion given a family history of heart disease (for `someone' in general) by cohort and pFH. The vast majority in all groups agree that for `someone with a family history of heart disease' it is `particularly important' for them `not to smoke', `to take exercise' and `to eat a healthy diet'. For each of these, respondents with and without a pFH of heart disease did not differ in the older two cohorts. In the youngest cohort, a significant difference was seen for smoking and exercise, with those with a pFH being more likely to strongly endorse the conventional health promotion line. However, those with and without a pFH of heart disease did not differ in any of the cohorts in the extent of their agreement with a more `fatalistic' statement about coronary disease. Many more were ambivalent about this statement than the other three, but still a sizeable proportion disagreed that, for `someone' with heart disease in their family it `doesn't matter what they do. Whether they get heart disease or not is out of their hands'.

pFH of heart disease, health-related attitudes and smoking

As shown in Table I, 39, 36 and 33% of the youngest, middle and oldest cohort, respectively, were current smokers; these figures are consistent with contemporary data for Scotland as a whole (Turner, 1997). As expected, those in poorer socio-economic circumstances were more likely to smoke (e.g. the odds of smoking were more than twice as great for those not living in owner-occupied households in comparison with those in owner-occupied households). Overall, women were slightly less likely to smoke, although the odds ratio is not significant, the difference in prevalence was statistically significant in the youngest (35 versus 44%, P < 0.05) and the oldest cohorts (31 versus 35%, P < 0.01), but smoking prevalence was similar by gender for the middle cohort (38 versus 34%). All further analyses of smoking are controlled for age, sex and tenure.

Table VII shows results of four logistic regression models which examine the relationship of smoking to (a) pFH of heart disease, (b) salience, (c) adherence to health promotion and (d) degree of fatalism. People who described themselves as more vulnerable to heart disease, who had the lowest adherence to health promotion advice for the prevention of heart disease and who had the most fatalistic attitudes to heart disease were significantly more likely to be smokers. However, in isolation, a pFH was unrelated to smoking (see Tables VIIa and VIII, model 1).

Table VIII presents the results of logistic regression models which examine the odds for smoking for those with a pFH whilst taking account of the other attitudinal variables, first separately then together. A pFH remains unrelated to current smoking when controlling also for either degree of adherence to coronary health promotion (Table VIII, model 2) or to fatalism (Table VIII, model 3). However, when salience (i.e. how likely the respondents are to perceive the risk of getting heart disease themselves) is included in the model, those with a pFH of heart disease are less likely to smoke [odds ratio 0.69, 95% confidence intervals (CI) 0.54–0.89, Table VIII, model 4]. The final model in Table VIII shows that decreased likelihood of smoking amongst those with a pFH remains significant when further controlling for adherence to health promotion and fatalism, in addition to salience.

A series of further logistic regression models tested for interactions between these attitudinal variables (salience, adherence to health promotion, fatalism and pFH), and between these variables and the socio-demographic variables (age, sex and tenure) (data not shown). The only significant interaction was between pFH and adherence to health promotion (P < 0.05). Those with a pFH of heart disease and the strongest adherence to conventional health promotion had the lowest odds for smoking (0.29, 95% CI 0.12–0.71). The inclusion of this term rendered the main effect of pFH non-significant.

Discussion

Most research on perceptions of family history has involved participants from `high-risk' groups or families, from clinical populations or has used small general population samples. We demonstrate here that, in a large-scale general population sample, perceptions of family history (familial weaknesses and tendencies) in general, and of heart disease in particular, are a common and an important element in lay epidemiology and risk formulation. This supports other smaller studies which have shown that `family history' figures highly in popular (and professional) accounts of the aetiology of major diseases, including `heart trouble' (Davison et al., 1989, 1991; Hastrup et al., 1992; Green et al., 1993; Brorsson et al., 1995).

Although there is increasing interest in lay perceptions of familial risk (and inherited susceptibilities to common diseases with a multi-factorial aetiology), few studies have related such perceptions to relevant attitudes or behaviours (Ponder et al., 1996). Marteau et al. (Marteau et al., 1995) have also argued that little is known about how self-assessments of coronary risk relate to epidemiologically defined risk. Their research suggests that family history of heart disease is one of the coronary risk factors that is accorded more weight by the lay public than by epidemiologically derived indices of risk, perhaps because greater weight is given to `visible' risk factors (such as weight or family history) rather than to `hidden' ones (such as cholesterol level or blood pressure) (Marteau et al., 1995). We have shown here that people who say they have a family history of heart disease are more likely to see themselves as `at risk' of heart disease and to think behavioural risk factors have an important influence on the development of heart disease; however, they were no more or less likely to have a `fatalistic' attitude towards heart disease. People who think they have a family history of heart disease and see themselves as at risk of the disease are much less likely to smoke.

We have concentrated here on perceptions of family history of heart disease, as any `risk-balancing' in behavioural decisions clearly has to be mediated through an individual's own perceptions of risk (which may or may not concur with external evaluations) (Love et al., 1985). We dichotomized people into `having' or `not having' a family history of heart disease on the basis of their responses to two simple questions. Survey methods such of these leave little room for the expression of ambiguity or qualification. Further research using qualitative methods has demonstrated that deciding whether or not one has a family history is not always straightforward (Emslie et al., submitted). Whilst coronary events in the family are a necessary condition, they are not sufficient. As we have seen here, there is a strong relationship to the number of relatives (particularly close family members), but other factors are carefully considered before familial coronary events are attributed to a familial tendency. For example, older family members are `discounted' by people when weighing up whether they have a family history of heart disease (Emslie et al., submitted). This may account for our observation that fewer among the oldest (63 year) cohort who reported having three or more close family members who had had heart disease regarded themselves as having a `family history' than among people in early mid-life (43 year cohort).

In the current paper we could only examine the relationship between a pFH and number of affected relatives on the basis of respondents' own reports of family ill-health. However, research in another population in the west of Scotland has demonstrated a high degree of accuracy of reporting of parental (age at and cause of) death in comparison with the information recorded on death certificates (Watt et al., submitted). Other studies which have compared medical and lay perceptions of family history suggest that knowledge about family history seems to be higher for heart disease (Kee et al., 1993) and cancer than for other chronic conditions, to be acquired at an early age (Hastrup et al., 1992), and to be highest for first-degree relatives (Love et al., 1985).

The finding that current smoking is related to perceptions of family history of heart disease in the context of other attitudes is relevant for health promotion. The reduction of coronary heart disease continues to be a major public health priority (Scottish Office, 1999); the government has recently stated that: `tackling smoking is central to cutting deaths from cancer and heart disease. Tackling smoking is central to improving health in Britain' (Department of Health, 1998). At the time that these data were collected, the strategic health promotion policy aims which were focused on the general public in Scotland were to `influence people's health-related knowledge, motivations and skills' [ (Health Education Board for Scotland, 1997), p. 1)] Research, including that reported here, shows that the public are generally well-informed about the behavioural risk factors for CHD, e.g. the Health Education Board for Scotland's own monitoring has suggested that `Respondents showed high awareness of the importance of not smoking, taking regular exercise and having a healthy diet' [ (Health Education Board for Scotland, 1997), p. 4]. Our data were collected prior to the recent shift in public health policy to an explicit recognition that health-related behaviours are important contributors to inequalities in health and `are linked strongly to social class and underlying life circumstances' (Scottish Office Department of Health, 1998; Scottish Office, 1999).

Although recent studies suggest that notions of familial inheritance may be important influences on attitudes, behaviour and (in)action, at present there is a paucity of empirical information about how people deal with thoughts, beliefs and information about their own inherited predisposition (Green et al., 1993; Richards, 1993; Macintyre, 1995). Macintyre has argued that, `A particularly important question is whether what has been called `the increasing geneticization of society' will lead to an increasingly general scepticism about the capacity of behaviour modification to improve health, and a consequent rejection of health promotion messages' [ (Macintyre, 1995), p. 230]. Richards too has suggested that `geneticization' may lead to fatalism about health, disease and life chances in general becoming more common (Richards, 1993). The quantitative data presented here support the view that a pFH of heart disease is important in decisions about smoking but its role in individual formulations of risk may only be understandable in the context of an array of other perceived risks that are `weighed' or `balanced' against this perception. It is increasingly accepted that both health promotion and interventions negotiated in clinical practice (Brorsson et al., 1995; Stott et al., 1995) are more likely to be well received and effective when based on a clearer understanding of cultural norms informing health beliefs, attitudes and practice (Backett and Davison, 1995; Kreuter and Strecher, 1995; Butler et al., 1998). We would thus argue that understanding people's construction of familial risk of heart disease is important for health promotion as it is common and linked to behavioural risk factors for coronary disease.

Table I.

Basic descriptive characteristics of sample

 Youth cohort(age c. 23) Early mid-life cohort(age c. 43) Late mid-life cohort(age c. 63) 
Total n 676 754 723 
Sex (% female) 53.0 56.1 55.7 
Tenure (% in owner-occupied) 63.4 76.4 65.0 
Percent current smokers 39.0 36.2 32.5 
Salience of heart disease (% who feel likely to get or already got heart disease) 24.3 34.0 42.3 
 Youth cohort(age c. 23) Early mid-life cohort(age c. 43) Late mid-life cohort(age c. 63) 
Total n 676 754 723 
Sex (% female) 53.0 56.1 55.7 
Tenure (% in owner-occupied) 63.4 76.4 65.0 
Percent current smokers 39.0 36.2 32.5 
Salience of heart disease (% who feel likely to get or already got heart disease) 24.3 34.0 42.3 
Table II.

Percentage reporting that they have `illnesses or weaknesses' in their family by age and sex

 Cohort 
 23 year olds Significancea 43 year olds Significancea 63 year olds Significancea 
 Men Women  Men women  Men Women  
aSignificance of sex difference within age. NS, not significant; *P < 0.05; **P < 0.01; ***P < 0.001. 
Any condition 37 43 NS 36 53 *** 31 50 *** 
Heart disease 17 20 NS 15 25 *** 16 23 * 
Cancer NS NS 10 ** 
Joint problems NS ** ** 
Respiratory problems * 10 * ** 
 Cohort 
 23 year olds Significancea 43 year olds Significancea 63 year olds Significancea 
 Men Women  Men women  Men Women  
aSignificance of sex difference within age. NS, not significant; *P < 0.05; **P < 0.01; ***P < 0.001. 
Any condition 37 43 NS 36 53 *** 31 50 *** 
Heart disease 17 20 NS 15 25 *** 16 23 * 
Cancer NS NS 10 ** 
Joint problems NS ** ** 
Respiratory problems * 10 * ** 
Table III.

Relationship between number of family members with heart disease and a pFH

 No. of family members χ2 (significance) 
 0 (%) 1 (%) 2 (%) 3+ (%)  
****P < 0.0001. 
a`Close' family defined as `parents or siblings'. 
b`All' χ2 family defined as close family (as above) and more distant family (grandparents, aunts, uncles and first cousins). 
All χ2 have 3 d.f., except that for `close family' for 23 years olds (2 d.f.). 
Close family onlya      
23 year olds      
no pFH 89 56 40 – 102.64 
pFH 11 47 60 – (****
43 year olds      
no pFH 96 72 48 162.74 
pFH 28 52 93 (****
63 year olds      
no pFH 97 83 52 24 218.40 
pFH 17 49 76 (****)All familyb 
23 year olds      
no pFH 99 83 62 33 183.06 
pFH 17 38 67 (****
43 year olds      
no pFH 98 88 72 41 199.93 
pFH 12 28 59 (****
63 year olds      
no pFH 99 92 66 41 220.62 
pFH 34 60 (****
 No. of family members χ2 (significance) 
 0 (%) 1 (%) 2 (%) 3+ (%)  
****P < 0.0001. 
a`Close' family defined as `parents or siblings'. 
b`All' χ2 family defined as close family (as above) and more distant family (grandparents, aunts, uncles and first cousins). 
All χ2 have 3 d.f., except that for `close family' for 23 years olds (2 d.f.). 
Close family onlya      
23 year olds      
no pFH 89 56 40 – 102.64 
pFH 11 47 60 – (****
43 year olds      
no pFH 96 72 48 162.74 
pFH 28 52 93 (****
63 year olds      
no pFH 97 83 52 24 218.40 
pFH 17 49 76 (****)All familyb 
23 year olds      
no pFH 99 83 62 33 183.06 
pFH 17 38 67 (****
43 year olds      
no pFH 98 88 72 41 199.93 
pFH 12 28 59 (****
63 year olds      
no pFH 99 92 66 41 220.62 
pFH 34 60 (****
Table IV.

Salience of heart disease by age and pFH

 `Unlikely' to getheart disease (%) `Likely' to getheart disease (%) Already gotheart disease (%) χ2 (significance) 
***P < 0.001. 
23 year olds     
no pFH 82 18 66.16 
pFH 48 51 (***)43 year olds 
no pFH 73 26 73.04 
pFH 38 55 (***
63 year olds     
no pFH 65 20 15 58.22 
pFH 30 41 30 (***
 `Unlikely' to getheart disease (%) `Likely' to getheart disease (%) Already gotheart disease (%) χ2 (significance) 
***P < 0.001. 
23 year olds     
no pFH 82 18 66.16 
pFH 48 51 (***)43 year olds 
no pFH 73 26 73.04 
pFH 38 55 (***
63 year olds     
no pFH 65 20 15 58.22 
pFH 30 41 30 (***
Table V.

Beliefs about importance of various factors in aetiology of heart disease by age and pFH

 No effect(%) Very smalleffect (%) Some effect(%) Quite importanteffect (%) Very importanteffect (%) Significance 
Some categories combined to get valid χ2. Number in brackets indicates number of degrees of freedom. NS, not significant; *P < 0.05; **P < 0.01; ***P < 0.001. 
(a) Housing and living conditions      χ(4)2 
23 year olds       
no pFH 13 24 39 18  1.24 
pFH 11 23 43 19 (NS) 
43 year olds       
no pFH 12 20 38 23 4.50 
pFH 13 21 31 22 13 (NS) 
63 year olds       
no pFH 18 15 32 24 11  1.49 
pFH 15 14 36 26 10 (NS) 
(b) Lifestyle      χ(2)2 
23 year olds       
no pFH 28 67  6.00 
pFH 24 75 (*
43 year olds       
no pFH 29 65  8.71 
pFH 18 77 (*
63 year olds       
no pFH 10 31 55  1.78 
pFH 12 25 61 (NS) 
(c) Family illnesses or weaknesses      χ(4)2 
23 year olds       
no pFH 35 44 13 40.55 
pFH 20 44 35 (***
43 year olds       
no pFH 31 43 20 15.43 
pFH 20 47 30 (**
63 year olds       
no pFH 33 40 18 18.75 
pFH 22 51 25 (***
(d) Stress      χ(4)2 
23 year olds       
no pFH 27 44 23  4.68 
pFH 24 41 31 (NS) 
43 year olds       
no pFH 22 43 30  1.73 
pFH 20 44 33 (NS) 
63 year olds       
no pFH 20 40 33  5.17 
pFH 18 46 33 (NS) 
 No effect(%) Very smalleffect (%) Some effect(%) Quite importanteffect (%) Very importanteffect (%) Significance 
Some categories combined to get valid χ2. Number in brackets indicates number of degrees of freedom. NS, not significant; *P < 0.05; **P < 0.01; ***P < 0.001. 
(a) Housing and living conditions      χ(4)2 
23 year olds       
no pFH 13 24 39 18  1.24 
pFH 11 23 43 19 (NS) 
43 year olds       
no pFH 12 20 38 23 4.50 
pFH 13 21 31 22 13 (NS) 
63 year olds       
no pFH 18 15 32 24 11  1.49 
pFH 15 14 36 26 10 (NS) 
(b) Lifestyle      χ(2)2 
23 year olds       
no pFH 28 67  6.00 
pFH 24 75 (*
43 year olds       
no pFH 29 65  8.71 
pFH 18 77 (*
63 year olds       
no pFH 10 31 55  1.78 
pFH 12 25 61 (NS) 
(c) Family illnesses or weaknesses      χ(4)2 
23 year olds       
no pFH 35 44 13 40.55 
pFH 20 44 35 (***
43 year olds       
no pFH 31 43 20 15.43 
pFH 20 47 30 (**
63 year olds       
no pFH 33 40 18 18.75 
pFH 22 51 25 (***
(d) Stress      χ(4)2 
23 year olds       
no pFH 27 44 23  4.68 
pFH 24 41 31 (NS) 
43 year olds       
no pFH 22 43 30  1.73 
pFH 20 44 33 (NS) 
63 year olds       
no pFH 20 40 33  5.17 
pFH 18 46 33 (NS) 
Table VI.

Attitudes to health promotion advice if someone has a family history of heart disease, by age and own pFH

 Agreestrongly (%) Justagree (%) Neither agree/disagree (%) Just disagree(%) Disagreestrongly (%) Significance 
Some categories combined to get valid χ2. Number in brackets indicates number of degrees of freedom. NS, not significant;*P < 0.05; **P < 0.01; ***P < 0.001. 
If heart trouble runs in someone's family…       
…it is particularly important for them not to smoke      χ(2)2 
23 year olds       
no pFH 57 30 10 16.67 
pFH 76 19 (***) 
43 year olds       
no pFH 79 16  3.32 
pFH 82 16 (NS) 
63 year olds       
no pFH 77 17  1.27 
pFH 79 18 (NS) 
…it is particularly important for them to take exercise      χ(2)2 
23 year olds       
no pFH 46 42 10 14.1 
pFH 64 31 (***
43 year olds       
no pFH 56 36  4.42 
pFH 57 40 (NS) 
63 year olds       
no pFH 48 43  0.58 
pFH 47 46 (NS) 
…it is particularly important for them to eat a healthy diet      χ(2)2 
23 year olds       
no pFH 60 32  8.66 
pFH 73 25 (NS) 
43 year olds       
no pFH 65 30  0.26 
pFH 66 30 (NS) 
63 year olds       
no pFH 64 30  0.58 
pFH 64 32 (NS) 
…it is doesn't matter what they do; whether they get heart disease or not is out of their hands      χ(4)2 
23 year olds       
no pFH 22 33 34  6.89 
pFH 16 28 47 (NS) 
43 year olds       
no pFH 10 12 30 45  1.46 
pFH 15 30 44 (NS) 
63 year olds       
no pFH 14 17 30 34  3.37 
pFH 11 19 36 30 (NS) 
 Agreestrongly (%) Justagree (%) Neither agree/disagree (%) Just disagree(%) Disagreestrongly (%) Significance 
Some categories combined to get valid χ2. Number in brackets indicates number of degrees of freedom. NS, not significant;*P < 0.05; **P < 0.01; ***P < 0.001. 
If heart trouble runs in someone's family…       
…it is particularly important for them not to smoke      χ(2)2 
23 year olds       
no pFH 57 30 10 16.67 
pFH 76 19 (***) 
43 year olds       
no pFH 79 16  3.32 
pFH 82 16 (NS) 
63 year olds       
no pFH 77 17  1.27 
pFH 79 18 (NS) 
…it is particularly important for them to take exercise      χ(2)2 
23 year olds       
no pFH 46 42 10 14.1 
pFH 64 31 (***
43 year olds       
no pFH 56 36  4.42 
pFH 57 40 (NS) 
63 year olds       
no pFH 48 43  0.58 
pFH 47 46 (NS) 
…it is particularly important for them to eat a healthy diet      χ(2)2 
23 year olds       
no pFH 60 32  8.66 
pFH 73 25 (NS) 
43 year olds       
no pFH 65 30  0.26 
pFH 66 30 (NS) 
63 year olds       
no pFH 64 30  0.58 
pFH 64 32 (NS) 
…it is doesn't matter what they do; whether they get heart disease or not is out of their hands      χ(4)2 
23 year olds       
no pFH 22 33 34  6.89 
pFH 16 28 47 (NS) 
43 year olds       
no pFH 10 12 30 45  1.46 
pFH 15 30 44 (NS) 
63 year olds       
no pFH 14 17 30 34  3.37 
pFH 11 19 36 30 (NS) 
Table VII.

Relationship of heart-related health attitudes to smoking behaviour [odds ratios from logistic regressionsa(n = 2079)]

  Odds ratio 95% CI 
aAll models are adjusted for age, class, housing tenure and sex. 
(a) pFH of heart disease (NS) no 1.00  
 yes 0.89 0.70–1.13 
(b) Salience (P < 0.0001) not likely to get heart disease 1.00  
 likely to get heart disease 1.97 1.60–2.43 
 already got heart disease 1.20 0.81–1.77 
(c) Adherence to health promotion (P < 0.0001) high 1.00  
 mid 1.14 0.93–1.40 
 least 1.86 1.42–2.43 
(d) `Fatalism' regarding heart disease (P = 0.0002) low 1.00  
 mid 1.22 0.99–1.50 
 high 1.83 1.37–2.44 
  Odds ratio 95% CI 
aAll models are adjusted for age, class, housing tenure and sex. 
(a) pFH of heart disease (NS) no 1.00  
 yes 0.89 0.70–1.13 
(b) Salience (P < 0.0001) not likely to get heart disease 1.00  
 likely to get heart disease 1.97 1.60–2.43 
 already got heart disease 1.20 0.81–1.77 
(c) Adherence to health promotion (P < 0.0001) high 1.00  
 mid 1.14 0.93–1.40 
 least 1.86 1.42–2.43 
(d) `Fatalism' regarding heart disease (P = 0.0002) low 1.00  
 mid 1.22 0.99–1.50 
 high 1.83 1.37–2.44 
Table VIII.

Odds ratio for smoking for pFH after adjusting for various factorsa (n = 2079)

Model Odds ratio (pFH versus none) 95% CI 
aAll models are adjusted for age, class, housing tenure and sex. 
Model 1 = pFH 0.89 0.70–1.13 
Model 2 = pFH + adherence to health promotion 0.93 0.73–1.18 
Model 3 = pFH + fatalism 0.90 0.71–1.14 
Model 4 = pFH + salience 0.69 0.54–0.89 
Model 5 = pFH + adherence to health promotion + fatalism + salience 0.73 0.56–0.94 
Model Odds ratio (pFH versus none) 95% CI 
aAll models are adjusted for age, class, housing tenure and sex. 
Model 1 = pFH 0.89 0.70–1.13 
Model 2 = pFH + adherence to health promotion 0.93 0.73–1.18 
Model 3 = pFH + fatalism 0.90 0.71–1.14 
Model 4 = pFH + salience 0.69 0.54–0.89 
Model 5 = pFH + adherence to health promotion + fatalism + salience 0.73 0.56–0.94 

The authors gratefully acknowledge all those involved in the Twenty-07 Study, particularly the participants, the interviewers, Geoff Der who provided statistical advice, and Lindsay Macaulay, Patricia Fisher, Margaret Reilly and Barbara Jamieson. K. H. and G. F. are employed by the Medical Research Council of Great Britain who fund the Twenty-07 Study. C. E. is supported by ESRC grant no. L128251028. Sally Macintyre and the journal's reviewers provided helpful comments on an earlier draft.

References

Backett, K. (
1992
) Taboos and excesses: lay health mortalities in middle class families.
Sociology of Health and Illness
 ,
14
,
2
.
Backett, K. C. and Davison, C. (
1995
) Lifecourse and lifestyle: the social and cultural location of health behaviours.
Social Science and Medicine
 ,
40
,
629
–638.
Blaxter, M. (1990) Health and Lifestyles. Routledge, London.
Brorsson, A., Troein, M., Lindbladh, E., Selander, S., Widlund, M. and Rastam, L. (
1995
) My family dies from heart attacks. How hypercholesterolaemic men refer to their family history.
Family Practice
 ,
12
,
433
–437.
Bunton, R., Nettleton, S. and Burrows, R. (eds) (1995) The Sociology of Health Promotion. Critical Analyses of Consumption, Lifestyle and Risk. Routledge, London.
Butler, C. C., Pill, R. and Stott, N. C. (
1998
) Qualitative study of patients' perceptions of doctors' advice to quit smoking: implications for opportunistic health promotion.
British Medical Journal
 ,
316
,
1878
–1881.
Charlton, J., Murphy, M., Khaw, K., Ebrahim, S. and Davey Smith, G. (1997) Cardiovascular diseases. In Charlton, J. and Murphy, M. (eds), The Health of Adult Britain 1841–1994. The Stationery Office, London, pp. 60–81.
Davison, C. and Davey Smith, G. (1995) The baby and the bath water: examining socio-cultural and free-market critiques of health promotion. In Bunton, R., Nettleton, S. and Burrows, R. (eds), The Sociology of Health Promotion. Critical Analyses of Consumption, Lifestyle and Risk, Routledge, London, pp. 91–99.
Davison, C., Frankel, S. and Smith, G. D. (
1989
) Inheriting heart trouble: the relevance of common-sense ideas to preventive measures.
Health Education Research
 ,
4
,
329
–340.
Davison, C., Smith, G. D. and Frankel, S. (
1991
) Lay epidemiology and the prevention paradox: the implications of coronary candidacy for health education.
Sociology of Health and Illness
 ,
13
,
1
–19.
Davison, C., Frankel, S. and Davey Smith, G. (
1992
) The limits of lifestyle: re-assessing `fatalism' in the popular culture of illness prevention.
Social Science and Medicine
 ,
34
,
675
–685.
Department of Health (1991) The Health of the Nation. A Consultative Document for Health in England. HMSO, London.
Department of Health (1992) The Health of the Nation: A Strategy for Health in England. HMSO, London.
Department of Health (1998) Smoking Kills. A White Paper on Tobacco. The Stationery Office, London.
Der, G. (1998) A Comparison of the West of Scotland Twenty-07 Study Sample and the 1991 Census SARs. MRC Medical Sociology Unit, Glasgow.
Dong, W. and Erens, B. (1997) Scotland's Health. Scottish Health Survey 1995. The Stationery Office, Edinburgh.
Ecob, R. (1987) West of Scotland Twenty-07 Study: the sampling scheme, frame and procedures for the cohort studies. Working Paper 6. MRC Medical Sociology Unit, Glasgow.
Ford, G., Ecob, R., Hunt, K., Macintyre, S. and West, P. (
1994
) Patterns of class inequality in health through the lifespan: class gradients at 15, 35 and 55 years in the West of Scotland.
Social Science and Medicine
 ,
39
,
1037
–1050.
Graham, H. (
1987
) Women's smoking and family health.
Social Science and Medicine
 ,
25
,
47
–56.
Green, J., Murton, F. and Statham, H. (
1993
) Psychosocial issues raised by a familial ovarian cancer register.
Journal of Medical Genetics
 ,
30
,
575
–579.
Hastrup, J. L., Phillips, S. M., Vullo, K., Kang, G. and Slomka, L. (
1992
) Adolescents' knowledge of medical terminology and family health history.
Health Psychology
 ,
11
,
41
–47.
Health Education Board for Scotland (1997) Indicators for Health Education in Scotland. HEBS, Edinburgh.
Kee, F., Tiret, L., Robo, J. Y., Nicaud, V., McCrum, E., Evans, A. and Cambien, F. (
1993
) Reliability of reported family history of myocardial infarction.
British Medical Journal
 ,
307
,
1528
–1530.
Kreuter, M. W. and Strecher, V. J. (
1995
) Changing inaccurate perceptions of health risk: results from a randomized trial.
Health Psychology
 ,
14
,
56
–63.
Love, R. R., Evans, A. M. and Josten, D. M. (
1985
) The accuracy of patient reports of a family history of cancer.
Journal of Chronic Disease
 ,
38
,
289
–293.
Lupton, D. and Chapman, S. (
1995
) `A healthy lifestyle might be the death of you': discourses on diet, cholesterol control and heart disease in the press and among the lay public.
Sociology of Health and Illness
 ,
17
,
477
–494.
Macintyre, S. (
1995
) The public understanding of science or the scientific understanding of the public? A review of the social context of the `new genetics'.
Public Understanding of Science
 ,
4
,
223
–232.
Macintyre, S., Annandale, E., Ecob, R., Ford, G., Hunt, K., Jamieson, B., Maciver, S., West, P. and Wyke, S. (1989) The West of Scotland Twenty-07 Study: health in the community. In Martin, C. J. and McQueen, D. V. (eds), Readings for a New Public Health. Edinburgh University Press, Edinburgh, pp. 56–73.
Marteau, T. M., Kinmouth, A.-L., Pyke, S., Thompson, S. G. and Group, F. H. S. (
1995
) Readiness for lifestyle advice: self-assessments of coronary risk prior to screening in the British family heart study.
British Journal of General Practice
 ,
45
,
5
–8.
Mullen, K. (
1992
) A question of balance: health behaviour and work context among male Glaswegians.
Sociology of Health and Illness
 ,
14
,
73
–97.
Mullen, K. (1993) A Healthy Balance: Glaswegian Men Talking about Health, Tobacco and Alcohol. Avebury, Aldershot.
Pill, R. and Stott, N. (
1987
) Development of a measure of potential health behaviour: a salience of lifestyle index.
Social Science and Medicine
 ,
24
,
125
–134.
Ponder, M., Lee, J., Green, J. and Richards, M. (
1996
) Family history and perceived vulnerability to some common diseases: a study of young people and their parents.
Journal of Medical Genetics
 ,
22
,
485
–492.
Richards, M. (
1993
) The new genetics: some issues for social scientists.
Sociology of Health and Illness
 ,
15
,
567
–586.
Ritchie, J. E., Herscovitch, F. and Norfor, J. B. (
1994
) Beliefs of blue collar workers regarding coronary risk behaviours.
Health Education Research
 ,
9
,
95
–103.
Scottish Office (1992) Scotland's Health: A Challenge to Us All. HMSO, Edinburgh.
Scottish Office (1999) Towards a Healthier Scotland. A White Paper on Health. The Stationery Office, Edinburgh.
Scottish Office Department of Health (1998) Working Together for a Healthier Scotland: A Consultation Document. The Stationery Office, Edinburgh.
Stott, N. C. H., Rollnick, S., Rees, M. R. and Pill, R. M. (
1995
) Innovation in clinical method: diabetes care and negotiating skills.
Family Practice
 ,
12
,
413
–418.
Tunstall-Pedoe, H., Woodward, M., Tavendale, R., Brook, R. A. and McCluskey, M. K. (
1997
) Comparison of the prediction by 27 different factors of coronary heart disease and death in men and women of the Scottish heart health study: coronary study.
British Medical Journal
 ,
315
,
722
–729.
Turner, R. (1997) Smoking. In Dong, W. and Erens, B. (eds), Scotland's Health. Scottish Health Survey 1995. The Stationery Office, Edinburgh, vol. 1, pp. 127–168.
West, P., Ford, G., Hunt, K., Macintyre, S. and Ecob, R. (
1994
) How sick is the West of Scotland? Age specific comparisons with national datasets on a range of health measures.
Scottish Medical Journal
 ,
39
,
101
–109.

Comments

0 Comments