In this population based nation wide cohort study we found an increased risk of MI in mothers of HIV-infected patients compared with mothers of the general population. The increased risk of MI was mainly seen in mothers of HIV patients reporting to be infected heterosexually or by IVD. In the fathers the increased risk of MI was only seen in case the offspring reported IVD as route of HIV infection.
The strength of our study is the nationwide population based design as well as a long and complete follow-up and the access to valid data on family members. We included parents of HIV patients and parents of a well-matched population control cohort. Further we were able to stratify our results on several important factors as parental gender, route of HIV transmission as well as calendar periods. As immigrants may retain higher risk of a disease due to susceptibility periods during childhood for some diseases as well as the potential effect of acculturation, immigrants were excluded from our study [22
]. In the start of the epidemic in the 1950s coronary heart disease was associated with high socio-economic status, but over time became a disease of the lower social classes. This, as well as the fact that treatment and prevention of ischaemic heart disease has changed considerably over time, makes control of calendar periods essential [8
]. We are not aware of other studies with a similar design.
Our study has some limitations. To identify patients with the MI diagnosis we relied on hospital registry based discharge diagnoses, which may not be entirely accurate and includes a risk of incorrectly registered diagnoses due to pre-hospital death. Importantly though we used the same source of data to ascertain MI for all study objects. Furthermore identification of MI on behalf of discharge diagnoses has previously been shown to be valid [25
]. We could not adjust for specific indicators of social class and socio-economic status such as parental education, occupation, family income, housing conditions, maternal marital status or illegitimacy. However, reported route of HIV infection in the offspring may act as a surrogate marker for socio-economic status.
Denmark harbours well-designed registries, which provide full access to data on all inhabitants concerning vital status and hospital contacts. As the Danish National Hospital Registry was not initiated until January 1977, some of the study subjects (i.e. the parents) might have had MI prior to study inclusion. These subjects are at increased risk of MI and ideally should have been excluded. However, as the Danish National Hospital Registry does not allow us to distinguish between first and subsequent episodes of MI in case the first MI occurred before 1977, we have no means of excluding these patients. Nevertheless, this phenomenon could hardly have biased our results substantially as the study subjects were young at inclusion, which makes previous episodes of MI unlikely. Further, age at study inclusion did not differ markedly between the compared groups (mothers versus mothers, fathers versus fathers) and therefore the proportion of post MIs misclassified as primary MI probably did not differ between groups. The relative risk measures therefore would remain rather unchanged even in case of inclusion of some post MI subjects. Parents were not included in the study if they had never been registered as parents in the Danish Civil Registration System, or had died prior to year 1977 as diagnoses of MI in the Danish National Hospital Registry were first available from that year. In 188 cases (7.6%) we were not able to trace the mothers in The Danish Civil registration System. We presume that the bias this may introduce is equivalent in the patient/control populations and does not influence our estimates.
In general the absolute risk of MI is higher in males than in females. However, in our study the increase in relative risk of MI was substantially higher in the mothers compared to the fathers. Several factors may explain this observation. First low socio-economic class, associated with particular lifestyle patterns, has been described as a stronger risk factor of MI for women than for men [26
]. Second the offspring may be more prone to share life style characteristics with their mothers than their fathers possible due to a preponderance of single family homes with mothers raising the children. And third the absolute risk of MI was almost three times higher in the fathers compared to the mothers. Therefore a minor increase in absolute risk of MI in the mothers will result in a greater increase in relative risk in this population compared to the fathers.
Although ischaemic heart disease manifests itself in adulthood, the disease and underlying atherosclerosis, develops throughout life [23
]. An increased risk of MI has previously been shown to be associated with lower social class in adulthood [22
]. However, adverse circumstances during pregnancy and childhood, including poor socio-economic conditions, seem to be important determinants of risk of adult cardiovascular disease. These risk factors are not merely due to the continuity of disadvantage throughout life [8
]. In a study by Kaplan et al it was suggested that HIV-infected individuals from economically disadvantaged populations bear a disproportionate burden of risk factors that might explain the association of MI and HIV [27
]. In addition to this it has previously been established that the prevalence of smokers are consistently higher in HIV patients than in age-matched controls [14
]. We cannot exclude that smoking may be a major contribution to the increased risk of MI in parents of HIV patients acting independently of other risk factors.
Our results showed a strong association between IVD as route of HIV infection in the offspring and risk of MI in both mothers and fathers of HIV patients. A strong association between initiation of IVD in the offspring and adverse socio-environmental factors in childhood has been described and this association probably explains the association observed in our study [32
Our study cannot unambiguously discriminate to what extend genetic and environmental factors exert their impact on risk of MI in the parents of HIV patients. But the findings imply that shared socio-economic disadvantages to some extent explain the excess risk of MI observed in the mothers. The genetic element in MI is not to be questioned but to our knowledge no genetic factors have been described, which is associated with increased risk of both HIV acquisition and MI. However, health-risk behaviours are frequently interrelated [34
] and tendency to "risk taking behaviour" has been explained as a result of a complex combination of social (e.g. socio-economic and environmental factors), genetic and developmental factors [35
]. In that case initiating IVD, sharing needles and having sex without barrier protection as well as heavy smoking, alcohol consumption, inappropriate diet and physical inactivity, could all be a result of the same genetic predisposition.
The increased risk of MI in HIV infected individuals has mainly been ascribed to the HIV infection per se and exposure to HAART [4
]. Additionally it has been questioned whether a HIV-positive status simply serves as a marker for differences in the prevalence of conventional risk factors such as smoking [29
]. We found an increased risk of MI in mothers of HIV patients of almost the same order (IRR: 1.31) as that seen in their HIV infected offspring when not on HAART (IRR: 1.39) [4
]. Our data therefore indicate that a main proportion of the increased risk of MI in these mothers and their HIV-infected offspring may stem from the same family related life style factors.
The implications of our findings extend beyond establishing the risk of MI in the parents of the HIV infected patients. Recent studies suggest premature aging, accelerated cognitive decline, increased risk of non-AIDS associated cancers and osteoporosis in HIV infected patients despite successful HAART [41
]. These studies may potentially be hampered by substantial influence from family related risk factors as well. Thus our data indicate that the association between risk for common diseases and HIV exposure may make it difficult to definitively establish HIV or antiretroviral drugs as causal risk factors for comorbidity in HIV infected patients.
In conclusion, mothers of HIV-infected patients have increased risk of MI. We presume, that the increased risk is related to family related life style risk factors, some of which may also influence the risk of MI in the HIV-infected patients. Thus, a major fraction of the increased risk of MI observed in HIV-infected patients may stem from non-HIV-related factors.