Impact of psychosocial factors on health outcomes in the elderly. A prospective study

One hundred and thirty-five initially healthy men and women older than 65 years of age were studied prospectively to identify factors associated with health outcomes. At study onset, measures of personality, social interaction, and health locus of control were obtained while participants were still healthy. A reliable health outcome measure was developed, based on the annual objective coding of morbidity. Correlations between variables showed significant associations between age, a less independent personality trait, and poor health outcomes. Anxiety and low levels of social interaction were associated with poor health after 8 years. Gender and health locus of control were not significantly related to health outcomes. A path analysis showed significant direct paths between age and trait anxiety and 8-year health outcomes, and indirect paths between anxiety and extroversion and 8-year health, by way of social interaction. There was no evidence that social interaction mediated the effects of either anxiety or extroversion on health. The structure of psychosocial characteristics of the healthy elderly sample was revealed in the pattern of correlations between personality, social interaction, and locus of control.

Differences in humoral insulin-antibody response among inbred Lou/M rats and epitope presentation differences in ELISA and radioimmune titration

Insulin antibodies were not detected in an insulin-capture enzyme-linked immunosorbent assay (ELISA), but they were easily detected in an insulin-copolymer-capture ELISA. Thus, there is a high degree of steric hindrance because of the proximity of the epitopes on the insulin monomer. This is circumvented by substituting an insulin copolymer for insulin in the capture ELISA. A regression analysis comparing the titers of 28 Lou/M rat insulin antiserums measured by liquid-phase radioimmune titration (RIT) with titers obtained in the direct insulin ELISA was not significant (P greater than .05). Thus, epitopes on insulin available and/or masked for antibody binding in the RIT differ from those available and/or masked in the direct insulin ELISA. As more of the epitopes become available when an insulin copolymer is substituted for monomeric insulin in the ELISAs, a significant positive correlation (P less than .05) with the RIT was observed with these 28 insulin antiserums. Twenty-five percent (7 of 28) of these antiserums contained more antibodies that bound to epitopes available in the ELISAs that were masked in the RIT. Conversely, two antiserums contained more antibodies that bound to epitopes that were available in the RIT but were masked in the ELISAs. Thus, the amount of insulin antibodies measured in a given antiserum can vary substantially, depending on which epitopes are made available or are masked in the particular antibody-titration method used. These results demonstrate that the humoral immune response to insulin among inbred Lou/M rats can vary in insulin-antibody levels as well as the epitopes on insulin to which the antibodies bind.