UC Merced

Understanding determinants of health behavior engagement is key to promoting health outcomes. One such determinant is health literacy, the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions (Baker et al., 2006). In this study we aimed to meta-analyze the effect of health literacy on health behavior engagement and outcomes across studies. We conducted a systematic search of published and unpublished literature (k = 203) to examine the overall relationship between health literacy, health behaviors, and health outcomes as mediated by social cognition constructs (attitudes, self-efficacy, knowledge, and risk perceptions). Meta-analysis indicated significant correlations between model constructs, particularly health literacy with knowledge and well-being, between self-efficacy and knowledge, attitudes, risk perceptions, and health behavior, and between risk perceptions and attitudes. A structural equation model based on the synthesized correlations indicated that self-efficacy and attitudes mediated the relationship between health literacy and health behavior. There were also total indirect effects of social cognition constructs on the health literacy-health behavior relationship. Analysis of health literacy measure type, sample origin, and behavior type revealed few differences in model results. Findings support that health literacy is associated with health behavior engagement, and this relationship may be explained by various health beliefs. Understanding the effects of health literacy or health beliefs on health behaviors may potentially be informative for future research aiming to promote health outcomes broadly.

In this dissertation, “Internet Meme Culture: Affective Response and Political Indoctrination,” I examine the affective power of internet memes. As there was no pre-existing theoretical framework by which to analyze internet memes, I use a combination of literary theory, art history, and image theory, as well as performance studies to create a framework which I then used to analyze both internet memes and the ways in which people were responding to these memes, both on and offline. Internet memes often interpolate people into greater cultural memes, sometimes without their knowledge or consent. Internet memes are often encoded with certain information, ideas, and ideologies, and yet are frequently decoded by those who view into them completely different messages. This is due to the interpretive communities and reading strategies of each individual host as well as to the affective power of internet memes, which value emotional responses over intellectual ones and enable the spread of false information just as readily as accurate information. The success of an internet meme has nothing to do with the relevance or factuality of the information contained therein but is based entirely on the affective response it generates within a host—that is, someone who interprets, appropriates, and redistributes it. Given the unpredictable nature of affect, these responses can be hard to quantify, understand, or even identify, and are often only recognized through the actions they produce within a person, further complicating the process. Social media provide a stage for acting out social drama (breach, crisis, and redress, as outlined by Victor Turner) similar to both traditional theater and the real world, and internet memes are a means of enacting the first two stages of social drama. However, when enacted online, these social dramas are unable to reach the third and final stage—the redress. Unlike the theater, internet meme culture is not a vehicle of transformation. Whether through social activism meant to change the system or through radicalization meant to overthrow the system, no lasting change can be made through internet meme culture. 

How changes in bone homeostasis affect immune development is not fully understood. The von-Hippel Lindau protein (VHL) regulates hypoxia-inducible factor (HIF) degradation, which is involved in cellular adaptation to low oxygen environments. Conditional deletion of Vhl in osteoblasts and hematopoietic progenitors have demonstrated a role for VHL in these cell types. Studies have demonstrated that the B cell development is mediated by crosstalk between the skeletal and hematopoietic systems. To understand how changes in bone homeostasis may affect immune cell development, we utilized Dmp1-Cre;Vhlfl/fl conditional knockout mice (VhlcKO), in which Vhl is deleted primarily in osteocytes, mature osteoblasts, and a small subset of MSCs. The VhlcKO mice display dysregulated bone growth, high bone density, smaller bone marrow (BM) cavity volume, and an overall decrease in BM cellularity compared to wild-type (WT) controls. In line with this, the frequencies, and numbers of B cells in the BM were significantly decreased. These data suggest that changes in bone homeostasis may adversely affect B cell development in a cell-extrinsic manner. We hypothesized that normal B cell development was not supported in VhlcKO mice due to alterations of the microenvironment niche, such as reduction of key niche cells and decreased production of B cell-supporting cytokines. Furthermore, we have obtained evidence that Vhl deletion also affects myeloid and erythroid development. We found elevated Epo levels in cKO peripheral blood serum and BM fluid by 6 weeks of age, and evidence for dysregulated erythropoiesis. Moreover, the cKO displayed an increased frequency of common myeloid progenitors, CD11b+ Gr1- monocytes, and CD11b+ Gr1+ granulocytes by 10 weeks of age. We also hypothesized that alterations in skeletal glucose metabolism directly affect myeloerythroid development in the BM through increased Epo-receptor (EpoR) signaling. EpoR is expressed on osteoprogenitors, hematopoietic stem cells, and B cells, but whether it is expressed on myeloid progenitors is unclear. Studies are in progress to test the hypothesis that overproduction of Epo by Vhl-deficient Dmp1+ cells result in elevated EpoR signaling in myeloid progenitors and lineages; this in turn, results in an increase in myeloid cell glycolysis and Glut1 expression. These studies are relevant for the understanding how BM microenvironmental changes can dysregulate B cell development and trigger adaptations in immunometabolism of myeloid cells. My thesis works expands on the potential mechanisms by which VHL in Dmp1-expressing cells regulate the development of distinct hematopoietic cell lineages in the BM, further elucidating and expanding our definition of “immune niches” to include osteocytes, bone, and vasculature.

Adolescent disclosure to parents has been associated with a variety of positive adolescent outcomes. Studies utilizing primarily self-report have found both concurrent and prospective associations between disclosure with parental behaviors. However, self-report does not allow researchers to determine what parental behaviors directly inhibit or facilitate adolescent disclosure in real-time. Additionally, research on adolescent disclosure has focused primarily on European American middle-class families. Less is known about disclosure in underrepresented families. The goal of this dissertation is to examine a) the impact of parental emotion-related responses (ERRs) on observed adolescent disclosure, b) associations between parental ERRs, disclosure, and physical health and c) associations between disclosure, early cultural factors, and social behaviors within a diverse population. Study I explored the impact of mothers’ ERRs and adolescent perspective taking on subsequent disclosures during real-time discussions. Interest and validation predicted the shortest lag times compared with other maternal ERRs. Findings highlight the importance of parental communication of both validation and interest in promoting disclosure in the context of parent-adolescent conversations. Study II examined associations between disclosure, parental ERRs, and diabetes management. Higher levels of anger and of positive affect, relative to parental baseline levels, predicted longer lag times to subsequent disclosures, whereas higher levels of expressive suppression predicted shorter lag times. However, these patterns varied depending on HbA1c. Specifically, adolescents with lower HbA1c had shorter lag times to subsequent disclosures, and adolescents with higher HbA1c had longer lag times when parents responded with increased anger compared to those with lower HbA1c. Findings highlight that parental ERRs to disclosures have implications for adolescent physical health. Study III explored associations between observed disclosure to parents, early cultural factors, and adjustment in a sample of diverse families. Adolescents from Latinx families were found to engage in lower levels of disclosure compared to adolescents from African American families. In Latinx families, speaking English in the home was prospectively associated with higher levels of disclosure. Adolescents from Spanish-speaking Latinx families engaged in lower levels of disclosure compared to adolescents from African American and English-speaking Hispanic families. Findings highlight that disclosure may vary among ethnic groups due to unique challenges with communication. Collectively, these studies identify the impact of parent behavior on adolescent disclosure during real-time interactions. These studies also provide information about how observed adolescent disclosure relates to adolescent adjustment in a variety of contexts (i.e., diverse families and those with chronic illnesses). Findings from this dissertation will inform research on adolescent disclosure from underrepresented populations and guide interventions aimed at families who struggle with a lack of disclosure from adolescents.

Tensors, which generalize matrices to more than two dimensions, are fundamental to many disciplines, such as scientific computing and machine learning. Improving the performance and scalability of tensor computation is essential to those domains. The recent advance of heterogeneous memory is promising to deliver large-scale, high-performance tensor computation. However, it is challenging to leverage memory heterogeneity because of performance disparity between memory components. Tensor computation, often characterized with irregular memory access patterns, large working set size, and unknown tensor dimension size, makes the usage of heterogeneous memory more challenging.

In this dissertation, we propose efficient and scalable heterogeneous memory systems for tensor computation to solve the challenges. The core innovation in our proposed systems is to introduce system-architecture-tensor co-designs, taking advantage of intersectional domain knowledge in runtime system policies, architecture characteristics, and tensor features. In particular, our approach takes into account runtime system policies (e.g., policies of data migration, prefetching, concurrency control), architecture characteristics (e.g., characteristics in emerging non-volatile memories, 3D-stacked memories, accelerators with massive parallelism), and tensor features (e.g., high data dimensionality, varying memory access patterns, irregular data distribution in the data structure) for tensor computation.

The evaluation results show that: (1) with evaluating various sparse tensor contraction datasets, our design brings 28−576 times speedup over the state-of-the-art sparse tensor contraction design; (2) with evaluating various sparse tensor contraction sequence datasets, our design brings 327-7362 times speedup over the state-of-the-art work; (3) with evaluating various tensor-based neural network training workloads, our design achieves up to 24 times and 4 times better energy consumption compared to CPU and GPU respectively; (4) with evaluating various tensor-based neural network training workloads, our design achieves up to 50% (33% on average) performance improvement compared to the state-of-the-art TensorFlow.

This dissertation intervenes in the fields of arts and cultural industries with the case study of art biennial exhibitions. As the Olympics of the art world, the international art biennial is a site of global knowledge production with cultural value and soft power. This dissertation examines Sinophone art biennials and these international events' global outreach and digital campaign that also recreate new geopolitical meanings in the virtual sphere. This research also emphasizes female curatorship in the art biennial scene, and aims to uncover the complex cultural and geopolitical values circulating in today's increasingly globalized and digitized world.

In the last decade alone fish hearts have become an increasingly popular model for studying heart function. Although fish hearts contain a single atrium and ventricle and present a fundamentally different cardiovascular system when compared to mammalian models, there are many developmental, structural, and functional commonalities between the two vertebrate classes. The goldfish heart, specifically, has remarkably similar electrical properties to that of humans’. For example, the heart rate, action potential morphology, and Ca2+ transient kinetics and dynamics of adult goldfish closely parallel those of humans, even more so than mice. In nearly all vertebrate species, direct input from the autonomic nervous system tightly controls cardiac contractility and excitability. Although there is an abundant amount of research on the autonomic control of cardiac contractility and excitability in numerous mammalian species, the characterization of pathophysiological mechanisms is still difficult to obtain for humans specifically. This is in part due to humans having strikingly dissimilar AP characteristics and electrocardiographic morphology in comparison to commonly used animal models such as mice, rats, and rabbits. Fish, on the other hand, are the largest and most diverse group of vertebrates, and as such, their autonomic nervous system regulation can often deviate from the classical vertebrate models used to study autonomic control of cardiac contractility and excitability. Ventricular APs, electrocardiograms, and Ca2+ transients recorded from the goldfish intact heart showed perfusion with either 100 nM isoproterenol (sympathetic agonist) or 5 µM carbamylcholine (parasympathetic agonist), was enough to stimulate the sympathetic branch or parasympathetic branch, respectively. Interestingly, our results indicate stimulation of the goldfish autonomic nervous system by these commonly used agonists resulted in a corresponding change in cardiac dromotropism, chronotropism, ionotropism, and lusitropism in a similar manner observed in humans. The data obtained from our experiments have led us to propose the goldfish heart as an excellent model for performing physiological experiments at the intact-heart level. Moreover, its shared ionic and electrical similarities with larger mammals open a new avenue for goldfish hearts to be used as a model to study human physiology.

Despite the potential negative implications of avoiding information about one’s health, research finds that people sometimes avoid this information. Current theorizing on avoidance implies that information is avoided to the extent that it is unwanted, however, no research has measured this construct of unwantedness nor examined its potential role in avoidance behavior. Further, although past work examines links between individual factors and avoidance, it generally fails to consider how these factors may work in tandem to predict avoidance. The present work proposes and defines a novel construct which can summarize multiple factors that contribute to health information avoidance: unwantedness of information, or the extent to which one does not desire information.In Study 1, participants completed a disease risk calculator, faced a decision to learn or forego their disease-risk feedback, and completed three measures of unwantedness. The results revealed that a brief 2-item measure of unwantedness accounted for as much variance in avoidance as more extended measures. Further, unwantedness of disease-risk feedback was associated with a greater likelihood of avoiding disease-risk feedback. In Study 2, participants completed a disease risk calculator, the 2-item measure of unwantedness, and then faced a decision to learn or forego their disease-risk feedback. The results revealed that the 2-item measure of unwantedness demonstrated convergent and discriminant validity. Moreover, disutility of learning information and lack of resources for learning information were both associated with greater unwantedness of information which, in turn, was associated with a greater likelihood of information avoidance. In Study 3, participants were assigned to a behavioral obligation manipulation, a manipulation emphasizing lack of treatment, or a control condition before completing a risk calculator for TAA deficiency (a fictitious disease). Next, they completed the 2-item measure of unwantedness and faced a decision to learn or forego their disease-risk feedback. The results suggested that the behavioral obligation manipulation increased unwantedness of disease-risk feedback and, in turn, avoidance. Collectively, these findings support the role of unwantedness of information in decisions to avoid health information and suggest that unwantedness is a promising target for future research aimed at reducing avoidance of health information.

Objective: Theoretical work has proposed that acceptance, attention monitoring, decentering, self-compassion, and nonreactivity may be mechanisms that explain beneficial effects of mindfulness. Yet, our understanding of how mindfulness mechanisms operate in everyday life is limited. The goal of this dissertation is to examine the extent to which these mechanisms naturally vary in everyday life and following a smartphone app-based mindfulness intervention. Methods: A large-scale university-wide randomized controlled trial was conducted to test the effects of app-based mindfulness intervention (Headspace) on well- being of university employees (n = 143). The study utilized ecological momentary assessment design to measure mindfulness mechanisms in participants’ natural environment. Study 1 examined whether mindfulness mechanisms represent independent constructs that naturally vary within a person over time, and whether naturally occurring fluctuations in mindfulness mechanisms differentially predict negative emotions. Study 2 investigated how mindfulness mechanisms change over the course of an 8-week app- based mindfulness intervention, with mindfulness mechanisms assessed at week 0 (pre- treatment), week 2, week 5, and week 8. Results: For Study 1, multilevel factor analysis revealed four distinct mechanisms – acceptance-attention, decentering, self-compassion, and nonreactivity – that exhibit substantial moment-to-moment variation. Greater acceptance-attention, self-compassion, and nonreactivity were associated with lower negative emotions, while greater decentering predicted higher negative emotions when examined concurrently with the other mechanisms. Study 2 found that improvements in acceptance-attention, decentering, and nonreactivity occurred after two weeks of the intervention. These mechanisms continued improving steadily over the course of the 8-week intervention. Conclusions: These findings highlight that not all mindfulness mechanisms may similarly relate to the same health outcomes. Although significant improvements in mindfulness mechanisms occur after two weeks of mindfulness intervention, these effects compound over time suggesting that longer intervention format can lead to greater improvements in mindfulness mechanisms compared to shorter programs. Finally, the findings demonstrate that app-based mindfulness interventions can effectively teach mindfulness mechanisms and provide a suitable alternative for people who cannot access traditional in-person mindfulness interventions.

This thesis explores the effects that skin color, discrimination, and religious coping have on Black Americans’ mental health. Using colorism as a theoretical framework I ask two research questions, How does everyday discrimination and skin color influence the mental health of young Black Americans? How do gender and religion mediate the impact of discrimination and skin color on mental health? This thesis utilizes data from the National Longitudinal Study of Adolescent to Adult Health (Add Health), 1994-2008. The dependent variables used in this thesis are the mental health measures depression diagnosis and feeling disliked. The primary independent variable in this thesis is interviewer reported skin color. My findings indicate that there is skin color variation among mental health outcomes. I find that the two mental health outcomes depression diagnosis and feeling disliked showed different outcomes. Lighter skin respondents reported more experiences of depression diagnosis and darker skin respondents reported more experiences of feeling disliked. The findings in this thesis show how colorism can influence skin color differences in mental health. I also show how skin color variation can be influenced by the way that mental health is being measured for Black Americans.