Introduction
The gender view
On average, men have less healthy lifestyles than women. They receive less information from physicians, adopt health-promoting behaviours less often (Verbrugge, Reference Verbrugge1985; Roter and Hall, Reference Roter and Hall1997; Sandman et al., Reference Sandman, Simantow and An2000), and are less emotionally expressive than women (Courtenay, Reference Courtenay2000). Men, in turn, suffer increased fatal chronic illness and less acute illness. Men also have higher cholesterol and blood-pressure readings (Galdas et al., Reference Galdas, Cheater and Marshall2005). Furthermore, psychophysiological responses to stress, such as medication use, are common among men, who also recover slowly from stress (Lash et al., Reference Lash, Eisler and Schulman1990; Roter and Hall, Reference Roter and Hall1997; Courtenay, Reference Courtenay2000; Reference Courtenay2003; Courtenay and Keeling, Reference Courtenay and Keeling2000; Sandman et al., Reference Sandman, Simantow and An2000).
Health beliefs and health behaviour
Health beliefs are closely related to adults’ immediate social experiences. Adults prefer to define health in terms of their body and vitality, whereas health behaviour is understood as meaning that individuals are responsible for their own health (Marks et al., Reference Marks, Murray, Evans and Willing1999). Attitudes and beliefs are important determinants of health, affecting both healthy and risky behaviours (Fishbein and Ajzen, Reference Fishbein and Ajzen1975; Ajzen, Reference Ajzen1985). Unrealistic optimism, for instance, has a negative influence on health perceptions, causing people to rate their health risks as less severe than their peers rate (Weinstein, Reference Weinstein1987; Van der Pligt, Reference Van der Pligt1994). Risky health behaviours, such as drinking, are associated with greater perceived susceptibility to related health risks (Vollrath et al., Reference Vollrath, Knoch and Cassano1993). However, these risks are different in men and women.
Although men’s health behaviours are important determinants of excess mortality and premature death (Courtenay, Reference Courtenay2000), it was only during the 1990s that men’s gender-specific health care needs have begun to attract more serious research attention (Lash et al., Reference Lash, Eisler and Schulman1990; Griffiths, Reference Griffiths1996; Roter and Hall, Reference Roter and Hall1997; Moynihan, Reference Moynihan1998; Courtenay, Reference Courtenay2000; Reference Courtenay2003; Courtenay and Keeling, Reference Courtenay and Keeling2000; Sandman et al., Reference Sandman, Simantow and An2000; Galdas et al., Reference Galdas, Cheater and Marshall2005).
Determinants of men’s health and well-being
In his recent review, Courtenay (Reference Courtenay2003) identified 30 key determinants of physical and mental health and well-being in men and boys in the United States. He systematically demonstrated that men and boys are more likely than women and girls to engage in behaviours that increase the risk of disease, injury, and death. Courtenay summarises these findings in terms of three main categories: Health behaviours, health-related beliefs, and expression of emotions and physical distress underlying factors that influence the health behaviours and health beliefs of men and boys, and health care.
Earlier studies in Finland have also shown that men have several high-risk health behaviours that can cause serious physical and mental health problems (Helakorpi et al., Reference Helakorpi, Uutela, Prättälä and Puska2000; Näslindh-Ylispangar et al., Reference Näslindh-Ylispangar, Sihvonen, Vanhanen and Kekki2005). The purpose of this study was to identify determinants of health and well-being in middle-aged men living in Helsinki and to analyse how these key determinants of health behaviour and health beliefs are associated with underlying health covariates.
Methods
Data collection
A sample of 690 men (all of the men who were born in 1961 in the northeastern health care region of the Helsinki area) was drawn from the population register filed by the Statistics Finland Bureau.
The final sample composed of 273 (39%) men who voluntarily completed the Health Behaviour Questionnaire (HBQ; Helakorpi et al., Reference Helakorpi, Uutela, Prättälä and Puska2000) and attended health examinations with public health nurses. Written consent was obtained from each participant.
Measurements
The 96-item HBQ, of which 16 single-item variables were used in this study, was originally developed for the measurement of health in a broad sense. It includes questions regarding socio-demographic factors, subjective health status, chronic diseases, and lifestyle. If two variables, for example, education and occupation, were highly correlated, only one was selected. Similarly, if the distributions of the variables were skewed, the variables were excluded from the analysis.
Measurements, such as body mass index (BMI), waist circumference, and self-reported information from the HBQ (Helakorpi et al., Reference Helakorpi, Uutela, Prättälä and Puska2000) were used to ascertain associations between dependent variables and their underlying factors.
Trained public health nurses determined the BMI and waist circumference at the Helsinki Health District. The clinical measurements were recorded by the same nurses, using the same instruments, and in the same environment, thereby ensuring the reliability of the data.
Dependent and independent variables
From the HBQ, five variables that served as equivalents to the categories identified by Courtenay (Reference Courtenay2003) functioned as dependent variables: (1) use of medications (behavioural responses to stress), (2) alcohol consumption (risk-taking behaviour), (3) self-rated health (health beliefs), (4) burnout (expression of emotions) and, (5) motivation (readiness to change unhealthy behaviours). Twelve independent variables (marital status, occupation, BMI, waist circumference, joint ache, dental health, illness or injury, insomnia, depression, alcohol counselling by family members, physical exercise, and alcohol consumption) were selected from the HBQ on the basis of the literature and correlations between outcome measures and eligible predictors were examined.
BMI and waist circumference were continuous variables. To be considered a risk factor, the following thresholds were required: BMI (overweight or obese ⩾25.0 kg/m2; normal range, <25.0 kg/m2) and waist circumference (abdominal obesity ⩾94 cm; normal, <94 cm). The criteria for objective health measurements were based on the literature (Lakka et al., Reference Lakka, Laaksonen, Lakka, Niskanen, Kumpusalo, Tuomilehto and Salonen2002; De Backer et al., Reference De Backer, Ambrosioni, Borch-Jonhsen, Brotons, Cifkova, Dallongeville, Ebrahim, Faergeman, Graham, Mancia, Manger Cats, Orth-Gomer, Perk, Pyörälä, Rodicio, Sans, Sansoy, Sechtem, Silber, Thomsen and Wood2003).
In the HBQ, the questions of use of medications, alcohol consumption, burnout, motivation, joint ache, illness or injury, insomnia, depression, and alcohol counselling by family members were originally posed as dichotomies (ie, yes or no). Also other variables, except alcohol consumption and physical exercise, were dichotomized as follows: marital status (0 = married and 1 = single/divorced/widowed), occupation (0 = office work and 1 = industrial/construction work), and self-rated health (1 = good health (excellent/good) and 0 = average/poor). Average health was combined with poor health on the basis of the extant (Manderbacka et al., Reference Manderbacka, Lahelma and Martikainen1998; Manor et al., Reference Manor, Sharon and Chris2000; Kivimäki et al., Reference Kivimäki, Head, Ferrie, Shipley, Vahtera and Marmot2003).
With respect to the question of alcohol use, ‘How often do you drink alcohol, 6 units at a time?’, the response options were as follows: never, seldom, once a month, once a week, and daily. The variable was changed into three categories: 0 = never/seldom, 1 = once a month, and 2 = once a week/daily. The question regarding physical exercise, ‘How often do you have physical exercises lasting at least half an hour and getting winded and sweaty?’, had three variables: 2 = daily/4–6 times a week, 1 = 1–3 times a week, and 0 = 2–3 times a month or less.
Table 1 lists the dependent and independent variables of each model. The theoretical equivalent by Courtenay (Reference Courtenay2003) is presented in parentheses.
The general characteristics of the study sample are shown in Table 2, which outlines the distribution of outcome factors of the logistic models and the predictors of each model.
Statistical analysis
A stepwise logistic regression model was used to estimate the adjusted odds of each outcome by socio-demographic factors, health-behaviour factors, health-belief factors, and clinical factors. Six models were developed to test the effects of outcomes. Marital status and occupation were analysed by using force-entered procedures while the rest of the predictors were analysed by using each step conditional method.
The analysis was carried out using SPSS for Windows, version 11.0 (SPSS Inc., Chicago, IL, USA).
Findings
Models 1–5 present how the use of medications, alcohol consumption, self-rated health status, burnout, and motivation are associated with underlying health covariates (Table 3).
Use of medications and alcohol consumption
In Model 1, marital status (P = 0.050), occupation (P = 0.004), illness (P ⩽ 0.001), and joint ache (P = 0.043) were significantly associated with medication use. Married men were slightly more likely to use medications than single or divorced men. Industrial workers were more likely to use medications than office workers (OR = 2.75; 95% CI, 1.38–5.46). Men who had an illness or injury were eight times more likely to use medications (OR = 8.02; 95% CI, 3.71–17.30).
Alcohol counselling by family members was the most significant predictor of alcohol consumption in Model 2. Men who were counselled regarding alcohol use were 4.5-fold more likely not to use alcohol, 6 units at a time, than their peers (OR = 4.47; 95% CI, 2.21–9.03). The other statistically significant predictor for alcohol consumption was BMI. Obese men (BMI ⩾ 25.0 kg/m2) were two times more likely to use alcohol than men who had a normal BMI.
Self-rated health and burnout
In Model 3, marital status (P = 0.040), waist circumference (P ⩽ 0.001), perceived dental health (P ⩽ 0.001), and illness or injury (P ⩽ 0.001) were significant predictors for self-rated health. Men who were married (OR = 2.02; 95% CI, 1.03–3.96), men who had a normal waist circumference, (OR = 2.58; 95% CI, 1.48–4.47), men who rated their dental health as good (OR = 0.36; 95% CI, 0.21–0.62) or who did not suffer from illness or injury were more likely to regard their overall health as good (OR = 0.28; 95% CI, 0.13–0.59).
In Model 4, burnout was highly predicted by insomnia (P = 0.006) and depression (P ⩽ 0.001). Men who suffered from insomnia (OR = 2.49; 95% CI, 1.30–4.74) or who were depressed (OR = 4.87; 95% CI, 2.18–10.87) had a strong association with a history of burnout.
In Model 5, motivation was significantly predicted by marital status (P = 0.034) and occupation (P = 0.014). Married men (OR = 0.48; 95% CI, 0.24–0.95) and office workers (OR = 0.49; 95% CI, 0.28–0.86) were more likely to be motivated than their peers. Additionally, physical exercise and alcohol consumption served statistically significant predictors for motivation. Men who exercised 4–7 times a week were over two times more likely to get motivated than inactive men (OR = 2.36; 95% CI, 1.26–4.42). Similarly, no or infrequent use of alcohol was associated with readiness to change unhealthy behaviours (OR = 0.28; 95% CI, 0.13–0.59).
Discussion
Main findings of this study
Many relevant predictors that were significantly associated with the selected key determinants of health behaviour and health beliefs of middle-aged men were found in this study.
The use of medications analysis represents the response to stress. Existing illness or injury and occupation showed the strongest associations with medication use, in which industrial workers were more likely to use medication than office personnel.
Although, over two-thirds of men used medications for various symptoms, joint ache was only one indication. The two most common categories of occupational disease are musculoskeletal disorders and occupational stress (Coggon, Reference Coggon2005), which may increase the use of medications. Additionally, risk-taking behaviour, such as dangerous sports and other high-risk activities are more common among men than among women (Courtenay, Reference Courtenay2003). Santana (Reference Santana2005) has highlighted that changes in occupation-related morbidity require new research approaches to provide protection for workers and improvements in health standards for workers in contemporary society.
Compared with single or divorced responders, married men were more likely to follow medical regimens, which may be a consequence of input by their spouses (Ramirez Garcia et al., Reference Ramirez Garcia, Chang, Young, Lopez and Jenkins2006). On the basis of our results, it is not possible to know, however, what has been the utility of this usage for men’s health.
Alcohol counselling by family members was strongly associated with less use of alcohol in our study. This finding supports a recent study, in which 22% of the Finnish men aged 35–44 years reported to be counselled by family members, 3% by their physician, and 4% by public health nurses (Helakorpi et al., Reference Helakorpi, Uutela, Prättälä and Puska2005). Thus, the role of family members as ‘health keepers,’ is important not only for men but for entire families. According to Courtenay (Reference Courtenay2000), ‘health matters are women’s matters;’ such collective beliefs and assumptions are referred to by social scientists and called social norms or subjective norms (Ajzen, Reference Ajzen1985).
High levels of alcohol consumption have been associated with various physical diseases, accident-related injuries (Vollrath et al., Reference Vollrath, Knoch and Cassano1993; Courtenay, Reference Courtenay2003), mental health problems, difficulties at work, violence, and family problems (Courtenay, Reference Courtenay2003). In Finland, heavy-drinking habits are on the rise (Lahelma et al., Reference Lahelma, Karisto, Mannila and Notkola2006); 20% of the excess mortality is attributable to alcohol-related deaths, nearly 20% to other deaths from accidents and violence, and 13% to suicide (Martelin et al., Reference Martelin, Koskinen and Valkonen2006). Therefore, campaigns to educate spouses to have a more active role in alcohol counselling should lead to improved family health.
Lack of illness or injury, good dental health, and normal waist circumference had the strongest associations with perceived good health. These findings are consistent with previous results, according to which a variety of pathways link clinical measures to self-rated health (Manderbacka et al., Reference Manderbacka, Lahelma and Martikainen1998; Goldman et al., Reference Goldman, Glei and Chang2004; Myint et al., Reference Myint, Welch, Luben, Wainwright, Surtees, Bingham, Wareham, Smith, Harvey and Khaw2006) Special attention must be paid to a male’s dental health. Recent findings indicate an association between dental disease and coronary heart diseases (Jansson et al., Reference Jansson, Laustedt, Frithiof and Theobald2001).
From two adjusting factors, marital status was associated with self-rated health. Married men were more likely to perceive their health as good than single or divorced responders. This is in agreement with earlier results that compared with married men, unmarried men have more serious health risks and they engage in poorer health behaviour (Courtenay, Reference Courtenay2000; Courtenay and Keeling, Reference Courtenay and Keeling2000; Sandman et al., Reference Sandman, Simantow and An2000; Näslindh-Ylispangar et al., Reference Näslindh-Ylispangar, Sihvonen, Vanhanen and Kekki2005; Lahelma et al., Reference Lahelma, Karisto, Mannila and Notkola2006).
Burnout was strongly predicted by insomnia and depression. This finding strengthened previous results that men rest less often and take care of themselves less well than women. This situation may decrease a male’s well-being and cause burnout (Koskinen et al., Reference Koskinen, Martelin and Rissanen1999; Marks et al., Reference Marks, Murray, Evans and Willing1999; Courtenay, Reference Courtenay2003; Lahelma et al., Reference Lahelma, Karisto, Mannila and Notkola2006).
In 2005, 21% of Finnish, working men reported sleeplessness during the last month, and 17% thought that they were more stressed than other people. During the past year, 5% of adults experienced a severe episode of depression in Finland (Lönnqvist, Reference Lönnqvist2006).
Active physical exercise and no or infrequent alcohol consumption were the most significant predictors for motivation to change unhealthy behaviours. In our study, over one-half of the men reported doing physical exercise at least once a week. It, however, was not enough to get them as motivated as men who did exercise at least four times a week (OR = 1.7 versus OR = 2.4). This supports the finding that physical activity must be regular and repeated often to promote good health (Kukkonen-Harjula et al., Reference Kukkonen-Harjula, Borg, Nenonen and Fogelholm2005; Vuori, Reference Vuori2006).
A deeper understanding and further studies are needed to analyse the content and motivation of physical activities. Many questions remain unanswered in this study, such as: Are men taking risks while having sports during leisure time? or, Do they stress while having physical activity after work? or, Can they afford to have expensive sport hobbies?
The results concerning the association between excessive alcohol consumption and lack of motivation was in concordance with the findings of Courtenay (Reference Courtenay2003) that men respond to stress in less-healthy way than do women, and they are more likely than women to use avoidant coping strategies, such as alcohol consumption and social isolation.
As with the use of medication, both adjusting demographic factors explained motivation. Married men and office workers were more likely to be ready to promote their health than their peers. This result again strengthens previous findings that social support (ie, family members) and higher work status have an influence on one’s willingness to take care of their own health (Franklin et al., Reference Franklin, Larson, Khan, Wong, Leip, Kannel and Levy2001).
The health psychological ideology that individuals are both responsible for and also experts regarding their own health is notable in discussing lifestyles and risk-taking behaviour (Marks et al., Reference Marks, Murray, Evans and Willing1999; Courtenay, Reference Courtenay2003). Attitudes and beliefs are important determinants of health, affecting both healthy and risky behaviours. Male gender is strongly associated with risk-taking behaviour; the more traditional health beliefs of masculinity they have, the more health risks they exhibit. These beliefs have been linked with smoking, excessive alcohol consumption, drug use, insomnia, depression, cardiovascular diseases, stress, and unhealthy dietary habits. (Fishbein and Ajzen, Reference Fishbein and Ajzen1975; Ajzen, Reference Ajzen1985; Courtenay, Reference Courtenay2000; Courtenay and Keeling, Reference Courtenay and Keeling2000).
Gender-specific health is a deep and multidimensional issue to analyse, in which the determinants influence each other depending on social, mental, physiological, and environmental views. Actually, this issue is endless, and finally there are always interesting questions left waiting new answers and solutions.
Limitations of this study
At least some limitations to our study have to be considered. The first is the topic of interest. Research in matters as complicated as health behaviour and health-related beliefs are difficult and fragmentary, especially when there is a lack of a specific model or general theory. The second is the sampling. Our study population consisted only of men who were born in 1961 and lived in one area of a city. Third, the response rate was low because the participating men in this study were all volunteering, thus, results can be skewed. The fourth limitation is the measurement. Although the HBQ included several relevant questions that were considered equivalent to Courtney’s determinants, many of them provided no new information owing to the low response rates.
Despite the limitations outlined above, these findings will help to improve our understanding of the influences of key determinants and their underlying factors on health and well-being among middle-aged men.
What this study adds
Many relevant predictors that were significantly associated with the selected key determinants of health behaviour and health beliefs of middle-aged men were found in this study.
• Married men, office workers, men who were physically active, and men who used alcohol less frequently, were more likely to be motivated to change their health behaviours.
• Special attention must be paid to a male’s dental health.
• Counselling by family members seems to be an important factor in the maintenance of less alcohol consumption.
The promotion of regular physical activity and healthy lifestyles among middle-aged men is currently one of the most important priorities of preventive work in primary health care.
Ethical approval
The ethics committee for human research at Helsinki approved the study protocol. Informed consent was obtained from all participants and they were aware of the possibility to withdraw their participation at any time.
Acknowledgements
This study was supported by the Helsinki Heart District and the City of Helsinki. We wish to thank the public health nurses at these organisations for their contribution to this study.