Influence of early life characteristics on psychiatric admissions and impact of psychiatric disease on inflammatory biomarkers and survival: a Danish cohort study

Most psychiatric research has focused on the identification of etiologic and prognostic factors for specific psychiatric diagnostic categories, in particular depression and schizophrenia. A limited number of studies have examined risk factors such as composite measures of cognitive ability (IQ) or mortality across different psychiatric diseases (1–3). Furthermore, although mental disorders seem to have their roots early in life (4,5), few studies have explored the influence of early life characteristics on these disorders over the life course and potential underlying mechanisms such as systemic inflammation (6).

In a cohort of Danish men, we examined the impact of social, mental and physical characteristics assessed in childhood, young adulthood and midlife on the incidence of all psychiatric admissions and of schizophrenia, depression, alcohol and drug abuse. We further explored the influence of the above psychiatric diseases on inflammatory biomarkers and survival.

The information used was extracted from birth registers (birth weight and father’s socioeconomic position); a school survey in 1965 (IQ); conscript examinations (IQ, education and body mass index); a health survey in 2004 (Major Depression Inventory, body mass index and smoking) (7); and a follow-up examination in 2010 (adult socioeconomic position, IQ, Major Depression Inventory, body mass index, smoking and inflammatory biomarkers) (8). Biomarkers included high-sensitivity C-reactive protein (hsCRP), interleukin (IL)-6 and IL-18, IL-10, tumor necrosis factor alpha (TNF-alpha) and interferon gamma (IFN-gamma).

Participants were recruited from the Metropolit cohort, consisting of 11,532 men born in 1953 in the Copenhagen Metropolitan area. Data from birth certificates were available for all members of the cohort. For 11,108 men, additional information was collected from conscript board examinations at about age 20 years. Further, 7,987 cohort members participated in a school-based survey in 1965 around age 12 years, and 6,292 were followed up by mailed questionnaire in 2004 around age 51 years (7). In 2010, around age 57 years, 7,799 cohort members living in the Eastern part of Denmark were invited to The Copenhagen Ageing Midlife Biobank, and 2,486 participated in a health examination including blood sampling and psychological tests (8).

Information on any admission to a psychiatric ward from 1972 to 2009 was obtained by linkage with the Danish Psychiatric Central Registry. All-cause mortality was followed from 1968 to 2009 by register linkage with the Danish civil registration system.

Data were analyzed using chi-square test, t-test, linear and Cox regression analysis in STATA version 12. Psychiatric admission was entered as time-dependent variable in the survival analysis.

Of the cohort members, 1,640 (14.2%) had ever been admitted to a psychiatric ward between age 19 and 56 years. The most frequent diagnosis (34.5%) was alcohol or drug abuse, while 18.2% were diagnosed with schizophrenia and 17.1% had an affective disorder.

Men with any psychiatric admission had lower socioeconomic position and lower IQ from childhood to middle age. They had lower mean birth weight and lower body mass index at age 20 and in middle age, and were more often smokers. In adjusted regression analysis, low IQ at age 20 increased the likelihood of developing schizophrenia (hazard ratio, HR per SD decrease = 1.39, 95% CI: 0.79-1.61), but decreased that of developing depression (HR per SD decrease = 0.81, 95% CI: 0.68-0.96). Birth weight and low education at age 20 were associated with alcohol or drug abuse (HR per 100 g increase = 0.98, 95% CI: 0.97-1.00; and HR low versus high = 1.51, 95% CI: 1.16-1.97, respectively).

Among the 2,486 men who participated in the health examination in 2010, the 242 men with a psychiatric admission had significantly higher levels of hsCRP, IL-6 and IL-18, while IL-10, TNF-alpha and IFN-gamma were not associated with psychiatric morbidity. The regression coefficients for having any psychiatric admission were β = 0.37 (95% CI: 0.16-0.57) for hsCRP; β = 0.16 (95% CI: 0.02-0.30) for IL-6; and β = 0.09 (95% CI: 0.01-0.18) for IL-18.

Analyses of the relation between psychiatric diagnosis and biomarkers showed that alcohol or drug abusers (β = 0.79, 95% CI: 0.42-1.17) and those with affective disorders (β = 0.44, 95% CI: 0.00-0.89) had higher levels of hsCRP. Men with an abuse diagnosis also had higher IL-6 levels (β = 0.52, 95% CI: 0.28-0.76).

During the follow-up period, 1,392 (12.7%) of the 11,532 cohort members died. Of them, 511 (37%) had a psychiatric diagnosis. Men with a psychiatric admission had higher mortality rates at age 55 years (HR = 5.43, 95% CI: 4.76-6.20), after adjustment for early life characteristics. The analyses also showed increased mortality in all four psychiatric diagnostic categories. The highest HRs were observed for alcohol or drug abuse (8.23, 95% CI: 6.98-9.68) and schizophrenia (6.43, 95% CI: 5.20-8.12).

These findings suggest that low birth weight, socioeconomic position and IQ early in life increase the risk of psychiatric disease, in particular of alcohol or drug abuse, in adult men. Alcohol or drug abuse is strongly associated with inflammatory biomarkers and poor survival.