Is Demography Destiny? Application of Machine Learning Techniques to Accurately Predict Population Health Outcomes from a Minimal Demographic Dataset

For years, we have relied on population surveys to keep track of regional public health statistics, including the prevalence of non-communicable diseases. Because of the cost and limitations of such surveys, we often do not have the up-to-date data on health outcomes of a region. In this paper, we examined the feasibility of inferring regional health outcomes from socio-demographic data that are widely available and timely updated through national censuses and community surveys. Using data for 50 American states (excluding Washington DC) from 2007 to 2012, we constructed a machine-learning model to predict the prevalence of six non-communicable disease (NCD) outcomes (four NCDs and two major clinical risk factors), based on population socio-demographic characteristics from the American Community Survey. We found that regional prevalence estimates for non-communicable diseases can be reasonably predicted. The predictions were highly correlated with the observed data, in both the states included in the derivation model (median correlation 0.88) and those excluded from the development for use as a completely separated validation sample (median correlation 0.85), demonstrating that the model had sufficient external validity to make good predictions, based on demographics alone, for areas not included in the model development. This highlights both the utility of this sophisticated approach to model development, and the vital importance of simple socio-demographic characteristics as both indicators and determinants of chronic disease.     

Association of Iodine and Iron with Thyroid Function

Iodine and iron are essential elements for healthy thyroid function. However, little is known about the association of iron and iodine with thyroid function in the general US population. We investigated iron and iodine status in relation to concentrations of thyroid hormones. We included 7672 participants aged 20 and older from three surveys (2007–2008, 2009–2010, and 2011–2012) of the National Health and Nutrition Examination Survey. Serum thyroid measures (including free and total T3 and T4, and TSH), serum iron concentration, and urinary iodine concentrations were measured. Multivariate linear regression models were conducted with serum thyroid measures as dependent variables and combinations of serum iron concentration and urinary iodine concentration as predictors with covariate adjustment. Logistic regression models were performed with TSH levels (low, normal, and high) and combinations of serum iron concentration and urinary iodine concentration. Overall, 10.9% of the study population had low iron; 32.2 and 18.8% had low or high iodine levels, respectively. Compared with normal levels of iron and iodine, normal iron and high iodine were associated with reduced free T3 and increased risk of abnormal high TSH. Combined low iron and low iodine was associated with reduced free T3 and increased TSH. In addition, high iodine was associated with increased risk of abnormal high TSH in females but not in males. Thyroid function may be disrupted by low levels of iron or abnormal iodine, and relationships are complex and sex-specific. Large prospective studies are needed to understand the mechanisms by which iron interacts with iodine on thyroid function.     

Identification of lipopolysaccharide binding site on high molecular weight kininogen

Plasma kallikrein kinin system (KKS) activation along with its cellular receptors expression are increased after injury and in patients with septic shock, hypotensive bacteremia and rhesus monkey infected with Salmonella typhimurium. KKS signaling cascade is activated by activated factor XII (FXIIa, Hageman factor)- and prolylcarboxypeptidase (PRCP)-dependent pathways on endothelial cells. Among the many entities that comprise the KKS, high molecular weight kininogen (HK), a bradykinin precursor, is critical in the assembly and activation of this system. HK is primarily expressed in the liver and secreted into the bloodstream. The activation of the KKS influences the permeability of the endothelium by liberating bradykinin (BK) from HK. BK is a potent inflammatory peptide which stimulates constitutive bradykinin B2 and inducible B1 receptors to release nitric oxide and prostacyclin. Regardless of the triggers, PK can only be activated on HK bound to the artificial negatively charged or to cell membrane surfaces. Since LPS has a negatively charged moiety and the ability to induce inflammatory responses in human, we determined the interaction between LPS and HK. HKH19 (HK cell binding site) and heparin inhibited LPS binding to HK with IC₅₀s of 15 nM and 20 μg/ml, respectively. C1-inhibitor and N-acetylglucosamine glycan inhibited LPS binding to HK with IC₅₀s of about 10 μg/ml and 10 mM, respectively. This novel study underscores the implication of HK in infection. We propose that HKH19, heparin, and C1-inhibitor present therapeutic potential for the treatment of sepsis and hypotensive bacteremia.