Occupation, Marital Status, and Low-Grade Inflammation
Mutual Confounding or Independent Cardiovascular Risk Factors?
Objective— We explored the relationships between inflammatory proteins, occupation, and marital status, and their independent associations with incidence of cardiovascular disease (CVD).
Methods and Results— Five inflammation-sensitive proteins (ISPs) (fibrinogen, ceruloplasmin, haptoglobin, α1-antitrypsin, orosomucoid) were measured in 6075 apparently healthy men. Incidence of coronary events and stroke was followed over 18 years in relation to occupation and marital status. All ISPs showed higher concentrations in divorced men and in manual workers. Except for fibrinogen, this remained significant after adjustments for confounding factors. Adjusted for traditional cardiovascular risk factors, incidence of coronary events was significantly increased in unskilled manual workers and in divorced men. The relative risks were slightly reduced after further adjustments for ISPs (from 1.79 to 1.70 in unskilled manual workers; from 1.58 to 1.51 in divorced men). All ISPs were significantly associated with incidence of coronary events, after adjustments for traditional risk factors. This relationship was essentially unchanged after further adjustments for occupation and marital status.
Conclusion— Inflammation could contribute to, but not fully explain, the increased cardiovascular risk in manual workers and divorced men. Although the ISPs vary greatly by occupational and marital status, this does not confound the relationship between ISPs and incidence of CVD.
It is well-documented that men with low occupation level have increased incidence of cardiovascular disease (CVD).1–5 Furthermore, marital status has shown relationships with cardiovascular disease, with higher incidences among divorced men.6–8 The reasons for these relationships are still unclear, despite many years of research. Even though individuals with lower socioeconomic position and unmarried men generally have higher levels of the traditional cardiovascular risk factors, eg, smoking, diabetes, cholesterol and hypertension, this can only explain a minor share of the increased incidence of CVD.1 Recent studies have shown that subjects with low socioeconomic position also have higher levels of C-reactive proteins (CRP), fibrinogen and other markers of inflammation,9–12 although some studies found weak or no relationships.13,14 It is unclear to what extent relationships with plasma markers of inflammation accounts for the increased incidence of cardiovascular disease in individuals with low socioeconomic level or among unmarried individuals.
Many studies have shown that increased plasma levels of various inflammatory markers, including fibrinogen, orosomucoid, CRP, white blood cells and serum amyloid A, are associated with incidence of myocardial infarction and stroke.15–19 The nature of this association is controversial. Many researchers claim that the inflammatory markers, or the underlying inflammatory processes they represent, are causally involved in the development of atherosclerosis and plaque rupture.20,21 However, this view is questioned by other researchers. The inflammatory markers are associated with many cardiovascular risk factors, including smoking, hypertension, and low socioeconomic position, and it has been argued that the relationship between inflammatory markers and CVD is caused by confounding, rather than being an independent risk factor.5,22,23 Even though some studies have reported relationships between socioeconomic position and inflammation, few have explored the role of occupation or marital status for the relationship between inflammation and CVD.5
The present study from a population-based cohort of middle-aged men sought to explore the relationships between marital status, occupation, and 5 inflammatory markers, and to what extent these factors independently are associated with incidence of CVD.
Between 1974 and 1984, 22 444 men participated in a screening program for detection of individuals with high risk for cardiovascular diseases.24 Complete birth cohorts from the city of Malmö were invited. Participation rate was 71%. Determination of 5 plasma proteins was part of the program for 6193 men. These men were randomly selected from birth cohorts examined between 1974 and 1982. After exclusion of men with a history of myocardial infarction, stroke, or cancer (according to questionnaire), 6075 men remained. Mean age was 46.8±3.7 years (range, 28 to 61). The health service authority of Malmö approved and funded the screening program. All participants gave informed consent.
Information about occupation was retrieved by data linkage with the national census investigations performed in 1970, 1980, and 1985, which are total registers of the Swedish population in those years. The subjects were classified according to the census that was closest to the screening examination. Occupation was categorized into unskilled manual workers, skilled manual workers, low-level nonmanual workers, medium-level nonmanual workers, high-level nonmanual workers, and others (early retired people, unemployed, farmers, enterprisers, etc.).25 Information on occupation was available for 6056 men (99.7%).
Information about marital status was similarly retrieved from the national censuses. Marital status was categorized into married, never married, divorced men, and widowers. Information about marital status was available for all men.
Subjects were categorized into smokers and nonsmokers. The smokers were categorized into consumers of up to 9 cigarettes per day, 10 to 19 cigarettes, and daily consumption of 20 cigarettes or more.26
Body mass index (BMI) was calculated as weight/height2 (kg/m2).
Blood pressure (mm Hg) was measured twice in the right arm after a 10-minute rest. The average of 2 measurements was used. A sphygmomanometer and a rubber cuff of appropriate size were used. Use of antihypertensive medication was assessed in a questionnaire.
Physical inactivity in spare time was assessed using the question “Are you mostly engaged in sedentary activities in spare time, for example, watching TV, reading, going to the movie?” Subjects who confirmed a doctor’s diagnosis of angina pectoris or who used nitrates were considered to have angina pectoris.
Blood samples were taken after an overnight fast. Serum cholesterol was analyzed with standard methods at the laboratory of the university hospital. Men with fasting whole blood glucose ≥6.1 mmol/L, men with 2-hour glucose values ≥10.0 mmol/L (glucose load, 30g/m2 body surface area), and men who reported treatment for diabetes were considered as having diabetes.
High alcohol consumption was assessed by means of the modified shortened version of the Michigan Alcoholism Screening Test.27 Men with >2 (of 9) affirmative answers were considered to have high alcohol consumption.
Electroimmunoassay was used to assess the plasma levels of acute phase proteins.28 The analysis was performed consecutively at the time of study entry. The detection limits were 20 mg/L for ceruloplasmin, 50 mg/L for α1-antitrypsin, and 350 mg/L for orosomucoid, haptoglobin, and fibrinogen. The coefficient of variation was <5%.28 We have previously shown that all 5 ISPs are associated with cardiovascular disease and that the cardiovascular risk increases with the number of ISPs in the top quartile.17 The top quartiles were as follows: fibrinogen >4.0 g/L, orosomucoid >0.93 g/L, α1-antitrypsin >1.42 g/L, haptoglobin >1.76 g/L, and ceruloplasmin >0.36 g/L.
All men were followed from the baseline examination until death, emigration from Sweden, or December 31, 1997 (mean follow-up, 18.7±3.7 years). A coronary event was defined as nonfatal myocardial infarction (code 410 according to the International Classification of Diseases, ICD, 9th revision) or death caused by ischemic heart disease (ie, ICD codes 410 to 414). Stroke was defined as cases coded 431 (intracerebral hemorrhage), 434 (ischemic stroke), or 436 (unspecified) according the ICD. All strokes were validated by review of hospital records, except for 34 (15%) cases who were hospitalized outside the city of Malmö or whose records were unavailable.
The Swedish Hospital Discharge Register and the Stroke register of Malmö were used for case retrieval.29,30 A validation study from the Swedish Hospital Discharge Register showed that the diagnosis “myocardial infarction” was false only in 5% of the cases.30
Analysis of variance was used to study the relationships between the inflammatory proteins and occupation or marital status. A general linear model was used to adjust the relationships for confounding factors. High-level nonmanual workers and married men were reference categories for the occupational and marital status groups, respectively.
Cox proportional hazards model was used to analyze the event rates in relation to inflammatory proteins, categories of occupation, and marital status, respectively. First, the relationships were adjusted for age. Second, traditional risk factors were entered (smoking, tobacco consumption, diabetes, BMI, systolic blood pressure, medication for hypertension, physical inactivity, angina pectoris, cholesterol, high alcohol consumption, triglycerides [log normalized]). Finally, the relationships with occupation and marital status were adjusted for inflammatory proteins and the inflammatory proteins were adjusted for occupation and marital status. The inflammatory proteins, BMI, systolic blood pressure, cholesterol, triglycerides, and age were used as continuous variables. Tobacco consumption was used as an ordinal variable. Four categories of marital status and 6 categories occupation were used. The remaining variables were dichotomous.
Inflammation in Relation to Occupation and Marital Status
All proteins, except fibrinogen, showed significant differences between the occupational groups after adjustments for other risk factors. Men with high- and medium-level nonmanual occupation generally had lower levels than unskilled manual workers (Table 1).
Marital status was similarly associated with the inflammatory proteins (Table 2). The highest levels were observed in divorced men, and married men had the lowest concentrations. After adjustments for other risk factors, ceruloplasmin, orosomucoid, and α1-antitrypsin were significantly related to marital status.
The relationships were essentially the same in separate analyses of smokers and nonsmokers, and in men younger than and older than the median age of 47.4 years (data not shown). However, after adjustments for confounding factors, orosomucoid was significantly associated with marital status in smokers and not in nonsmokers. Fibrinogen and haptoglobin were significantly associated with marital status in young men and not in older groups.
Occupation and Marital Status and Incidence of Cardiovascular Disease
The age-adjusted incidence of coronary events was substantially higher among men with low occupation level. Adjustment for traditional risk factors reduced this relationship, but it remained significant (Table 3). Further adjustment for the inflammatory proteins reduced the relative risk in unskilled manual workers from 1.79 (CI, 1.3 to 2.5) to 1.70 (CI, 1.2 to 2.4).
Unskilled manual workers and low-level nonmanual workers had similarly higher incidence of stroke. In low-level nonmanual workers, this relationship remained significant after adjustment for risk factors (Table 3).
Divorced men had higher incidence of coronary events. This association was reduced, but remained significant, after adjustments for traditional risk factors. After further adjustments for inflammatory proteins, the relative risk of coronary events among divorced men was reduced from 1.58 (CI, 1.3 to 2.0) to 1.51 (CI, 1.2 to 1.9) (Table 4).
Both for occupation level and marital status, the relationships with cardiovascular disease were consistent in smokers and nonsmokers, and in men older than and younger than the median age (47.4 years), and the effects of adjustment were essentially the same irrespective of smoking status or age group.
Inflammatory Proteins and Incidence of Cardiovascular Disease
All ISPs were strongly associated with incidence of coronary events. These relationships were reduced after adjustments for traditional risk factors but remained clearly significant (Table 5). Further adjustments for occupation and marital status had small effects of the relative risks. These relationships were very similar when smokers and nonsmokers, and older and younger men were analyzed separately (not shown).
With exception of α1-antitrypsin, the inflammatory proteins showed essentially the same relationships with incidence of stroke, and the effects of adjustments for risk factors, occupation, and marital status were similar.
The nature of the relationships between inflammatory plasma proteins and incidence of CVD is still controversial. This study shows that the inflammatory proteins vary greatly between men with different occupation, and between married and divorced men, even after adjustments for several risk factors. Even though marital status and occupation were clearly associated with inflammation and incidence of cardiovascular disease, adjustment for marital status and measures of socioeconomic position had very small effects on the relationship between inflammation and CVD. It is unlikely that relationships with socioeconomic position or low social support explain why inflammatory markers are associated with incidence of cardiovascular disease.
The results also show that traditional cardiovascular risk factors and the present inflammatory markers only partially explain the increased risk of CVD in manual workers and divorced men. Approximately 50% of the increased risk associated with unskilled manual occupation could be explained by other cardiovascular risk factors. However, all risk factors were assessed at one single occasion, and they do not fully reflect the exposure during the life-course. Hence, the role of traditional risk factors for the relationships between occupation, marital status, and incidence of CVD could be more important than the present results indicate.
The reason why cardiovascular risk factors and inflammatory proteins cluster in groups with low socioeconomic position is an important research question. Many explanations have been suggested and it is likely that more than one factor contribute to this relationship.4 In most studies, men with low socioeconomic position have worse health behavior, eg, more smoking, unhealthy diet, and higher psychosocial stress.31 Experimental studies have shown that the response of inflammatory cytokines10 after a stress test is higher in groups with low socioeconomic position. Hence, the vulnerability to various stress factors seems to be higher in groups with low socioeconomic level.4
Our results disagree with the findings from the British Women’s Heart and Health Study.5 In their report, neither fibrinogen nor CRP was associated with incidence of CVD after adjustments for other risk factors. The authors suggested that the socioeconomic position and other confounding factors, rather than causal relationships, explain the association between inflammatory proteins and CVD. However, in contrast with our results, CRP and fibrinogen showed very weak relationships with incidence of CVD in their study, and none of them was significant after adjustments for behavioral risk factors like smoking and BMI. A recent study from Finland reported rather weak relationships between CRP and intima-media thickness, and the association was attenuated after adjustments for obesity.23 The authors suggested that obesity explains the associations between CRP and socioeconomic level and between CRP and intima-media thickness. Many studies from the present cohort have shown strong and very consistent relationships between markers of inflammation and incidence of CVD.15,17 Differences between study populations, end-points, as well as different inflammatory markers could explain the discrepancies between our results and the British and Finnish studies. The correlations between different inflammatory markers are often moderate (r=0.3 to 0.6 in this study), and the inflammatory markers could therefore pick up different aspects of the underlying inflammatory processes. It is likely that the prognostic information can be improved if a combination of inflammatory markers is used instead of one specific protein.
It has sometimes been suggested that reverse causation explain why inflammatory proteins are associated with CVD, ie, that advanced atherosclerosis increases the plasma levels of inflammatory markers. It is well-known that the atherosclerotic process starts early, and that many middle-aged men have advanced atherosclerotic lesions. Studies of subjects with angina pectoris suggest that uncomplicated atherosclerosis per se has rather small effects on the inflammatory proteins.32 However, even though the men initially were healthy and the follow-up time was long, we cannot completely rule out that pre-existing asymptomatic atherosclerosis could have increased the degree of inflammation.
Many different measures of socioeconomic position have been associated with fibrinogen and CRP. Even though other measures of socioeconomic position during the life-course probably could add more information,5 there is no doubt that occupation and marital status are important measures of the social and socioeconomic situation. These factors showed significant relationships with inflammation as well as incidence of CVD, in accordance with many previous studies.1–12 If the social and socioeconomic situation had been an important confounding factor for the relationship between inflammation and CVD, there should have been a clear attenuation after adjustment for marital status and occupation, even though they do not pick up all aspects of these dimensions. Furthermore, in an additional analysis not presented here, the results were almost identical after further adjustments for several measures of the socioeconomic position earlier in life (car ownership in 1960, education level in 1960, crowded housing in 1960).
There are some limitations of the study that should be mentioned. The end-points were retrieved from national and local hospital register. The hospital registers cover the southern parts of Sweden during the entire follow-up period. For some parts of Sweden, this register does not cover nonfatal cases before 1987, and some cases that moved to other areas in Sweden before 1987 may have been lost. A few subjects who moved away from Sweden were censored at the time of emigration. This number was small and cannot influence the results.
The laboratory analyses were limited to those who were available in clinical practice at the time of screening. For example, we have no information about CRP or low-density lipoprotein cholesterol.
Yet another limitation is that we do not know about changes that occurred during the follow-up period, eg, smoking cessation, divorces, or new occupations.
In conclusion, relationships with inflammatory markers could contribute to, but not fully explain, the increased incidence of CVD in divorced men and men with low occupation. The inflammatory proteins vary greatly by occupation level and marital status. However, this does not confound the relationship between inflammatory markers and incidence of CVD.
The study was supported by the Swedish Research Council, the Swedish Heart-Lung foundation, the Swedish Stroke foundation, and Syskonen Svenssons foundation.
- Received October 7, 2005.
- Accepted November 29, 2005.
Kaplan GA, Keil JE. Socioeconomic factors and cardiovascular disease: a review of the literature. Circulation. 1993; 88: 1973–1998.
Rosvall M, Östergren PO, Hedblad B, Isacsson SO, Janzon L, Berglund G. Occupational status, educational level, and the prevalence of carotid atherosclerosis in a general population sample of middle-aged Swedish men and women: results from the Malmö Diet and Cancer Study. Am J Epidemiol. 2000; 152: 334–346.
Steptoe A, Marmot M. The role of psychobiological pathways in socio-economic inequalities in cardiovascular disease risk. Eur Heart J. 2002; 23: 13–25.
Ebrahim S, Wannamethee G, McCallum A, Walker M, Shaper AG. Marital status, change in marital status, and mortality in middle-aged British men. Am J Epidemiol. 1995; 142: 834–842.
Rosengren A, Wedel H, Wilhelmsen L. Marital status and mortality in middle-aged Swedish men. Am J Epidemiol. 1989; 129: 54–64.
Jousilahti P, Salomaa V, Rasi V, Palosuo T. Association of markers of systemic inflammation, C reactive protein, SAA, And fibrinogen with socioeconomic status. J Epidemiol Comm Health. 2003; 57: 730–733.
Wilson TW, Kaplan GA, Kauhanen J, Cohen RD, Wu M, Salonen R, Salonen JT. Association between plasma fibrinogen concentration and five socioeconomic indices in the Kuopio Ischemic Heart Disease Risk Factor Study. Am J Epidemiol. 1993; 137: 292–300.
Rosengren A, Wilhelmsen L, Welin L, Tsipogianni A, Teger-Nilsson AC, Wedel H. Social influences and cardiovascular risk factors as determinants of plasma fibrinogen concentration in a general population sample of middle aged men. BMJ. 1990; 300 (6725): 634–638.
Danesh J, Muir J, Wong YK, Ward M, Gallimore JR, Pepys MB. Risk factors for coronary heart disease and acute-phase proteins. A population-based study. Eur Heart J. 1999; 20: 954–959.
Engström G, Hedblad B, Stavenow L, Lind P, Tydén P, Janzon L, Lindgärde F. Inflammation-Sensitive Plasma Proteins and Incidence of Myocardial Infarction in Men With Low Cardiovascular Risk. Arterioscler Thromb Vasc Biol. 2003; 23: 2247–2251.
Engström G, Lind P, Hedblad B, Stavenow L, Janzon L, Lindgärde F. Effects of cholesterol and inflammation-sensitive plasma proteins on incidence of myocardial infarction and stroke in men. Circulation. 2002; 105: 2632–2637.
MacLeod J, Smith GD. Fibrinogen, social position, and risk of heart disease. J Epidemiol Community Health. 2004; 58: 157.
Kivimaki M, Lawlor DA, Juonala M, Davey Smith G, Elovainio M, Keltikangas-Jarvinen L, Vahtera J, Viikari JS, Raitakari OT. Lifecourse Socioeconomic Position, C-Reactive Protein, and Carotid Intima-Media Thickness in Young Adults. The Cardiovascular Risk in Young Finns Study. Arterioscl Thromb Vasc Biol. 2005; 25: 2197–2202.
Lind P, Engström G, Stavenow L, Janzon L, Lindgärde F, Hedblad B. Risk of myocardial infarction and stroke in smokers is related to plasma levels of inflammation-sensitive proteins. Arterioscler Thromb Vasc Biol. 2004; 24: 577–582.
Engström G, Jerntorp I, Hedblad B, Berglund G, Janzon L. Geographical distribution of stroke in an urban population: Relationships with prevalence of risk factors and socioeconomic circumstances. Stroke. 2001; 32: 1098–1103.
Värdering av diagnoskvaliten för akut hjärtinfarkt i patientregistret 1987 och 1995 [in Swedish]. The National board of health and welfare, Stockholm, Sweden.
Lynch JW, Smith GD, Kaplan GA, House JS. Income inequality and mortality: importance to health of individual income, psychosocial environment, or material conditions. BMJ. 2000; 320: 1200–1204.