Neuropsychological test validation of speech markers of cognitive impairment in the Framingham Cognitive Aging Cohort.

AIM: Although clinicians primarily diagnose dementia based on a combination of metrics such as medical history and formal neuropsychological tests, recent work using linguistic analysis of narrative speech to identify dementia has shown promising results. We aim to build upon research by Thomas JA & Burkardt HA et al. (J Alzheimers Dis. 2020;76:905-22) and Alhanai et al. (arXiv:1710.07551v1. 2020) on the Framingham Heart Study (FHS) Cognitive Aging Cohort by 1) demonstrating the predictive capability of linguistic analysis in differentiating cognitively normal from cognitively impaired participants and 2) comparing the performance of the original linguistic features with the performance of expanded features. METHODS: Data were derived from a subset of the FHS Cognitive Aging Cohort. We analyzed a sub-selection of 98 participants, which provided 127 unique audio files and clinical observations (n = 127, female = 47%, cognitively impaired = 43%). We built on previous work which extracted original linguistic features from transcribed audio files by extracting expanded features. We used both feature sets to train logistic regression classifiers to distinguish cognitively normal from cognitively impaired participants and compared the predictive power of the original and expanded linguistic feature sets, and participants' Mini-Mental State Examination (MMSE) scores. RESULTS: Based on the area under the receiver-operator characteristic curve (AUC) of the models, both the original (AUC = 0.882) and expanded (AUC = 0.883) feature sets outperformed MMSE (AUC = 0.870) in classifying cognitively impaired and cognitively normal participants. Although the original and expanded feature sets had similar AUC, the expanded feature set showed better positive and negative predictive value [expanded: positive predictive value (PPV) = 0.738, negative predictive value (NPV) = 0.889; original: PPV = 0.701, NPV = 0.869]. CONCLUSIONS: Linguistic analysis has been shown to be a potentially powerful tool for clinical use in classifying cognitive impairment. This study expands the work of several others, but further studies into the plausibility of speech analysis in clinical use are vital to ensure the validity of speech analysis for clinical classification of cognitive impairment.

Depression Screening from Voice Samples of Patients Affected by Parkinson's Disease.

Depression is a common mental health problem leading to significant disability world wide. Depression is not only common but also commonly co-occurs with other mental and neurological illnesses. Parkinson's Disease gives rise to symptoms directly impairing a person's ability to function. Early diagnosis and detection of depression can aid treatment, but diagnosis typically requires an interview with a health provider or structured diagnostic questionnaire. Thus, unobtrusive measures to monitor depression symptoms in daily life could have great utility in screening depression for clinical treatment. Vocal biomarkers of depression are a potentially effective method of assessing depression symptoms in daily life, which is the focus of the current research. We have a database of 921 unique patients with Parkinson's disease and their self assessment of whether they felt depressed or not. Voice recordings from these patients were used to extract paralinguistic features, which served as inputs to machine-learning and deep learning techniques to predict depression. The results are presented here and the limitations are discussed given the nature of the recordings which lack language content. Our models achieved accuracies as high as 0.77 in classifying depressed and non-depressed subjects accurately using their voice features and PD severity. We found depression and severity of Parkinson's Disease had a correlation coefficient of 0.3936, providing a valuable feature when predicting depression from voice. Our results indicate a clear correlation between feeling depressed and the severity of the Parkinson's disease. Voice may be an effective digital biomarker to screen for depression among patients suffering from Parkinson's Disease.

Parkinson's Disease Classification of mPower Walking Activity Participants.

Motion sensor data collected using Sage Bionetwork's mPower application on the Apple iPhone to record participant activities is analyzed to classify samples as positive or negative for Parkinson's Diagnosis. Pre-processing of the data showed differences in the time and frequency dimensions for features derived from Apple Core motion data. Several classic machine learning classification algorithms were trained on seventy-seven derived data points for best precision, recall, and F-1 score. Accuracy as high as ninety-two percent were achieved, with the best results attained from decision tree and multi-layered artificial neural network algorithms. This research shows that motion data produced on the Apple iPhone using the mPower application shows promise as an accessible platform to classify participants for presence of Parkinson's Disease signs.

Acceptability of mHealth augmentation of Collaborative Care: A mixed methods pilot study.

OBJECTIVE: To assess the feasibility and acceptability of a mobile health platform supporting Collaborative Care. METHOD: Collaborative Care patients (n=17) used a smartphone app to transmit PHQ-9 and GAD-7 scores and sensor data to a dashboard used by one care manager. Patients completed usability and satisfaction surveys and qualitative interviews at 4weeks and the care manager completed a qualitative interview. Mobile metadata on app usage was obtained. RESULTS: All patients used the app for 4weeks, but only 35% (n=6) sustained use at 8weeks. Prior to discontinuing use, 88% (n=15) completed all PHQ-9 and GAD-7 measures, with lower response rates for daily measures. Four themes emerged from interviews: understanding the purpose; care manager's role in supporting use; benefits of daily monitoring; and privacy / security concerns. Two themes were user-specific: patients' desire for personalization; and care manager burden. CONCLUSIONS: The feasibility and acceptability of the mobile platform is supported by the high early response rate, however attrition was steep. Our qualitative findings revealed nuanced participant experiences and uncovered some concerns about mobile health. To encourage retention, attention may need to be directed toward promoting patient understanding and provider engagement, and offering personalized patient experiences.

Patient-oriented health technologies: Patients' perspectives and use.

BACKGROUND: For patient-oriented mobile health tools to contribute meaningfully to improving healthcare delivery, widespread acceptance and use of such tools by patients are critical. However, little is known about patients' attitudes toward using health technology and their willingness to share health data with providers. AIMS: To investigate primary care patients' comfort sharing health information through mobile devices, and patients' awareness and use of patient portals. METHODS: Patients (n=918) who visited one of 6 primary care clinics in the Northwest US completed a survey about health technology use, medical conditions, and demographics. RESULTS: More patients were comfortable sharing mobile health information with providers than having third parties store their information (62% vs 30%, Somers D=.33, p<0.001). Patients older than 55 years were less likely to be comfortable sharing with providers (AORs 0.37-0.42, p<0.01). Only 39% of patients knew if their clinic offered a patient portal; however, of these, 67% used it. Health literacy limitations were associated with lower portal awareness (AOR=0.55, p=0.005) but not use. Portal use was higher among patients with a chronic condition (AOR= 3.18, p=0.004). CONCLUSION: Comfort, awareness, and use of health technologies were variable. Practices introducing patient-facing health technologies should promote awareness, address concerns about data security, and provide education and training, especially to older adults and those with health literacy limitations. Patient-facing health technologies provide an opportunity for delivering scalable health education and self-management support, particularly for patients with chronic conditions who are already using patient portals.

Determinants of bone mineral density in obese premenopausal women.

Despite being a risk factor for cardiovascular disease and diabetes mellitus, obesity has been thought to protect against osteoporosis. However, recent studies have demonstrated a differential impact of specific fat compartments on bone mineral density (BMD) with visceral adipose tissue (VAT) having potential detrimental effects on BMD. Visceral obesity is also associated with dysregulation of the GH/IGF-1 axis, an important regulator of bone homeostasis. The purpose of our study was to evaluate the differential effects of abdominal fat depots and muscle, vitamin D, and hormonal determinants, including insulin-like growth factor-1 (IGF-1), testosterone, and estradiol, on trabecular BMD of the lumbar spine. We studied 68 healthy obese premenopausal women (mean BMI, 36.7±4.2 kg/m(2)). Quantitative computed tomography (QCT) was used to assess body composition and lumbar trabecular BMD. There was an inverse association between BMD and VAT, independent of age and BMI (p=0.003). IGF-1 correlated positively with BMD and negatively with VAT and, in stepwise multivariate regression modeling, was the strongest predictor of BMD and procollagen type 1 amino-terminal propeptide (P1NP). Thigh muscle cross sectional area (CSA) and thigh muscle density were also associated with BMD (p<0.05), but 25-hydroxyvitamin D [25(OH)D], testosterone, free testosterone, and estradiol levels were not. 25(OH)D was associated inversely with BMI, total, and subcutaneous abdominal adipose tissue (p<0.05). These findings support the hypothesis that VAT exerts detrimental effects, whereas muscle mass exerts positive effects on BMD in premenopausal obese women. Moreover, our findings suggest that IGF-1 may be a mediator of the deleterious effects of VAT on bone health through effects on bone formation.

Vertebral bone marrow fat is positively associated with visceral fat and inversely associated with IGF-1 in obese women.

Recent studies have demonstrated an important physiologic link between bone and fat. Bone and fat cells arise from the same mesenchymal precursor cell within bone marrow, capable of differentiation into adipocytes or osteoblasts. Increased BMI appears to protect against osteoporosis. However, recent studies have suggested detrimental effects of visceral fat on bone health. Increased visceral fat may also be associated with decreased growth hormone (GH) and insulin-like growth factor 1 (IGF-1) levels which are important for maintenance of bone homeostasis. The purpose of our study was to assess the relationship between vertebral bone marrow fat and trabecular bone mineral density (BMD), abdominal fat depots, GH and IGF-1 in premenopausal women with obesity. We studied 47 premenopausal women of various BMI (range: 18-41 kg/m², mean 30 ± 7 kg/m²) who underwent vertebral bone marrow fat measurement with proton magnetic resonance spectroscopy (1H-MRS), body composition, and trabecular BMD measurement with computed tomography (CT), and GH and IGF-1 levels. Women with high visceral fat had higher bone marrow fat than women with low visceral fat. There was a positive correlation between bone marrow fat and visceral fat, independent of BMD. There was an inverse association between vertebral bone marrow fat and trabecular BMD. Vertebral bone marrow fat was also inversely associated with IGF-1, independent of visceral fat. Our study showed that vertebral bone marrow fat is positively associated with visceral fat and inversely associated with IGF-1 and BMD. This suggests that the detrimental effect of visceral fat on bone health may be mediated in part by IGF-1 as an important regulator of the fat and bone lineage.

Comparison of 3.0 T proton magnetic resonance spectroscopy short and long echo-time measures of intramyocellular lipids in obese and normal-weight women.

PURPOSE: To compare correlations of intramyocellular lipids (IMCL) measured by short and long echo-time proton magnetic resonance spectroscopy (1H-MRS) with indices of body composition and insulin resistance in obese and normal-weight women. MATERIALS AND METHODS: We quantified IMCL of tibialis anterior (TA) and soleus (SOL) muscles in 52 premenopausal women (37 obese and 15 normal weight) using single-voxel 1H-MRS PRESS at 3.0 T with short (30 msec) and long (144 msec) echo times. Statistical analyses were performed to determine correlations of IMCL with body composition as determined by computed tomography (CT) and insulin resistance indices and to compare correlation coefficients from short and long echo-time data. Signal-to-noise ratio (SNR), linewidth, and coefficients of variation (CV) of short and long echo-time spectra were calculated. RESULTS: Short and long echo-time IMCL from TA and SOL significantly correlated with body mass index (BMI) and abdominal fat depots (r = 0.32 to 0.70, P = <0.05), liver density (r = -0.39 to -0.50, P < 0.05), and glucose area under the curve as a measure of insulin resistance (r = 0.47 to 0.49, P < 0.05). There was no significant difference between correlation coefficients of short and long echo-time spectra (P > 0.5). Short echo-time IMCL in both muscles showed significantly higher SNR (P < 0.0001) and lower CVs when compared to long echo-time acquisitions. Linewidth measures were not significantly different between groups. CONCLUSION: IMCL quantification using short and long echo-time 1H-MRS at 3.0 T is useful to detect differences in muscle lipid content in obese and normal-weight subjects. In addition, IMCL correlates with body composition and markers of insulin resistance in this population with no significant difference in correlations between short and long echo-times. Short echo-time IMCL quantification of TA and SOL muscles at 3.0 T was superior to long echo-time due to better SNR and better reproducibility.

Breath-hold 1H-magnetic resonance spectroscopy for intrahepatic lipid quantification at 3 Tesla.

OBJECTIVE: To compare breath-hold 1H-magnetic resonance spectroscopy (1H-MRS) with respiratory-gated 1H-MRS and computed tomography (CT) for quantification of hepatic lipid content. METHODS: Twenty-three premenopausal women underwent breath-hold point-resolved single-voxel 1H-MRS of the liver followed by respiratory-gated 1H-MRS at 3 Tesla and CT slice through the liver. Interscan variability for 1H-MRS was assessed in 6 volunteers. Pearson correlation coefficients, Bland-Altman 95% limit of agreement, and concordance correlation coefficients were calculated. RESULTS: There was a strong correlation between breath-hold and respiratory-gated 1H-MRS (r = 0.94, P < 0.0001; concordance correlation coefficient, 0.75). Using Bland-Altman analysis, all but 2 data points were within the limits of agreement. Both 1H-MRS techniques had low interscan variability. There was an inverse correlation of both 1H-MRS techniques with CT attenuation values of the liver. CONCLUSIONS: Breath-hold 1H-MRS is a reliable method to measure hepatic lipid content at 3 Tesla. Breath-hold 1H-MRS of the liver provides data that closely correlates with that obtained from longer-duration respiratory-gated technique.

Comparison of DXA and CT in the assessment of body composition in premenopausal women with obesity and anorexia nervosa.

Accurate methods for assessing body composition in subjects with obesity and anorexia nervosa (AN) are important for determination of metabolic and cardiovascular risk factors and to monitor therapeutic interventions. The purpose of our study was to assess the accuracy of dual-energy X-ray absorptiometry (DXA) for measuring abdominal and thigh fat, and thigh muscle mass in premenopausal women with obesity, AN, and normal weight compared to computed tomography (CT). In addition, we wanted to assess the impact of hydration on DXA-derived measures of body composition by using bioelectrical impedance analysis (BIA). We studied a total of 91 premenopausal women (34 obese, 39 with AN, and 18 lean controls). Our results demonstrate strong correlations between DXA- and CT-derived body composition measurements in AN, obese, and lean controls (r = 0.77-0.95, P < 0.0001). After controlling for total body water (TBW), the correlation coefficients were comparable. DXA trunk fat correlated with CT visceral fat (r = 0.51-0.70, P < 0.0001). DXA underestimated trunk and thigh fat and overestimated thigh muscle mass and this error increased with increasing weight. Our study showed that DXA is a useful method for assessing body composition in premenopausal women within the phenotypic spectrum ranging from obesity to AN. However, it is important to recognize that DXA may not accurately assess body composition in markedly obese women. The level of hydration does not significantly affect most DXA body composition measurements, with the exceptions of thigh fat.

Peak growth hormone-releasing hormone-arginine-stimulated growth hormone is inversely associated with intramyocellular and intrahepatic lipid content in premenopausal women with obesity.

CONTEXT: Visceral adiposity is a strong determinant of GH secretion, and low endogenous GH secretion is associated with increased insulin resistance, a key component of the metabolic syndrome. Increased fat accumulation in skeletal muscle and liver may play an etiological role in the development of insulin resistance and other complications of the metabolic syndrome. Little is known about the role of decreased endogenous GH secretion in the pathogenesis of insulin resistance in obesity. OBJECTIVE: To investigate the relationship between intramyocellular lipids (IMCL), intrahepatic lipids, and peak-stimulated GH in premenopausal women with obesity. DESIGN AND SETTING: We conducted a cross-sectional study at a clinical translational research center. PATIENTS: Patients included 21 premenopausal women with obesity (mean body mass index, 34.0 +/- 4.5 kg/m(2)) and 17 normal-weight controls (mean body mass index, 21.9 +/- 2.0 kg/m(2)) of comparable mean age. MAIN OUTCOMES MEASURES: IMCL and intrahepatic lipids were measured with proton magnetic resonance spectroscopy ((1)H-MRS). Body composition was measured with magnetic resonance imaging. Peak GH was measured after stimulation with GHRH-arginine. RESULTS: Obese subjects had higher IMCL, intrahepatic lipids, abdominal and thigh fat, and thigh muscle mass compared with normal-weight controls. There were strong inverse associations between peak GH and both IMCL and intrahepatic lipids independent of age and visceral adiposity. There were positive associations between IMCL and intrahepatic lipids with measures of insulin resistance and serum triglycerides. CONCLUSION: In premenopausal women with obesity, peak GH is inversely associated with IMCL and intrahepatic lipids independent of age and visceral adiposity. This suggests that low GH may contribute to insulin resistance in obesity through effects on muscle and intrahepatic lipids.

Increased bone marrow fat in anorexia nervosa.

CONTEXT: Although women with anorexia nervosa (AN) have severe depletion of body fat, a paradoxical increase in bone marrow fat has been described. Recent data suggest that marrow fat measured by 1H-magnetic resonance spectroscopy (MRS) in combination with bone mineral density (BMD) may be more valuable than either parameter alone in detecting bone weakness. OBJECTIVE: The objective of the study was to investigate the effect of AN on accumulation of marrow fat in spine and femur using 1H-MRS and the relationship between marrow fat, BMD, and body composition in subjects with AN and normal-weight controls. DESIGN: This was a cross-sectional study. SETTING: The study was conducted at a referral center. PATIENTS: Patients included 10 women with AN (29.8 +/- 7.6 yr) and 10 normal-weight age-matched women (29.2 +/- 5.2 yr). INTERVENTIONS: There were no interventions. MAIN OUTCOMES MEASURE: Marrow fat content of the fourth lumbar vertebra and femur measured by 1H-MRS. BMD of spine and hip measured by dual-energy x-ray absorptiometry. RESULTS: Subjects with AN had higher marrow fat at the fourth lumbar vertebra and femur compared with controls (P = 0.004-0.01). There was an inverse correlation between marrow fat of L4 and femur and BMD of the spine and hip (r = -0.56 to -0.71, P = 0.01-0.0002) and body mass index and sc adipose tissue of the thigh (r = -0.49 to -0.71, P = 0.03-0.0007). There was an inverse correlation between femur marrow fat and sc and total abdominal adipose tissue (r = -0.53 to -0.67, P = 0.003-0.03). CONCLUSION: Women with AN have greater lumbar and femoral marrow fat than controls, and marrow fat correlates inversely with BMD. This paradoxical increase in marrow fat at a time when sc and visceral fat are markedly reduced raises important questions about functional consequences of this process.