This dissertation is organized in a traditional format while including three essays that address specific research questions. Essay 1 examined the relationship between physical activity and community engagement and their effect on mental well-being among older men and women. Data from National Health and Aging Trends Study (NHATS) from 2018 to 2020 were explored and the posited relationships were tested. This essay provides empirical support that older adults who are reasonably active and involved in the community have greater mental well-being than those who isolate themselves. Essay 2 provides insight into older adults' motivation to improve their physical activity through the use of a fitness tracker. The key finding from this study is that wearables, especially fitness trackers, can significantly facilitate increased physical activity. Essay 3 is a mixed-methods study to understand older adults' perception of the usefulness of fitness trackers and interaction with such devices. Findings suggest that to increase the adoption of fitness trackers among older adults, makers could improve the esthetics and quality of the wristband in addition to the battery life of the tracker.

Background: Objectively tracking sedentary behavior (SB) and physical activity (PA) is becoming increasingly important as research continues to show the negative effects with increasing SB and decreasing PA. Objectives: The purpose of this study was to evaluate three commercial activity trackers with young adults regarding how they accurately measure SB and PA behaviors in a laboratory setting. Methods: 50 college-aged participants wore three wrist-based activity trackers (Fitbit Surge, Apple Watch, and Basis Peak) and two ActiGraph accelerometer devices during a series of SB and PA behaviors for five-minute intervals in a laboratory setting. The activity trackers were evaluated against direct observation and the ActiGraph devices, placed on the hip and wrist, which are consistent with the National Health and Nutrition Examination Survey (NHANES) standards of measure. Results: Overall accuracy during the SBs compared to direct observation was high, with Apple (99.0%), Basis (99.0%), and Fitbit (94.9%) performing similar to the Hip ActiGraph (95.1%) and markedly better than the Wrist ActiGraph (58.6%). Overall significant correlations (p ≤ 0.05) during the PAs were higher with the Wrist ActiGraph (66.7%) than with the Hip ActiGraph (8.3%). The Wrist and Hip ActiGraphs significantly correlated in three out of four SBs, but not in any PA behaviors.Discussion: Activity trackers are reliable when determining sedentary behavior, tend to overestimate step count during light walking, and underestimate activity level when biking. Also,the Wrist ActiGraph consistently underestimated both SB and PA step count compared to the Hip ActiGraph. While some variability is seen in the validity of the activity trackers, each activity tracker studied has its strengths and weaknesses. Understanding these strengths and limitations helps healthcare professionals more accurately interpret recorded data based on the patient specific device.

The purpose of the study was to evaluate the effects of goal setting and fitness trackers to increase daily step counts in adolescents with intellectual disability. An A-B-A-B withdrawal research design was implemented to evaluate the effectiveness of the intervention. The results indicated that goal setting and fitness trackers were effective in increasing daily step counts for two out of three participants.

No description available.

Health monitoring devices, such as Fitbit, are gaining popularity both as wellness tools and as a source of information for healthcare decisions. Predicting such wellness goals accurately is critical for the users to make informed lifestyle choices. The core objective of this thesis is to design and implement such a system that takes energy consumption and privacy into account. This research is modelled as a time-series forecasting problem that makes use of Spiking Neural Networks (SNNs) due to their proven energy-saving capabilities. Thanks to their design that closely mimics natural neural networks (such as the brain), SNNs have the potential to significantly outperform classic Artificial Neural Networks in terms of energy consumption and robustness. In order to prove our hypotheses, a previous research by Sonia et al. [1] in the same domain and with the same dataset is used as our starting point, where a private forecasting system using Long short-term memory (LSTM) is designed and implemented. Their study also implements and evaluates a clustering federated learning approach, which fits well the highly distributed data. The results obtained in their research act as a baseline to compare our results in terms of accuracy, training time, model size and estimated energy consumed. Our experiments show that Spiking Neural Networks trades off accuracy (2.19x, 1.19x, 4.13x, 1.16x greater Root Mean Square Error (RMSE) for macronutrients, calories burned, resting heart rate, and active minutes respectively), to grant a smaller model (19% less parameters an 77% lighter in memory) and a 43% faster training. Our model is estimated to consume 3.36μJ per inference, which is much lighter than traditional Artificial Neural Networks (ANNs) [2]. The data recorded by health monitoring devices is vastly distributed in the real-world. Moreover, with such sensitive recorded information, there are many possible implications to consider. For these reasons, we apply the clustering federated learning implementation [1] to our use-case. However, it can be challenging to adopt such techniques since it can be difficult to learn from data sequences that are non-regular. We use a two-step streaming clustering approach to classify customers based on their eating and exercise habits. It has been shown that training different models for each group of users is useful, particularly in terms of training time; however this is strongly dependent on the cluster size. Our experiments conclude that there is a decrease in error and training time if the clusters contain enough data to train the models. Finally, this study addresses the issue of data privacy by using state of-the-art differential privacy. We apply e-differential privacy to both our baseline model (trained on the whole dataset) and our federated learning based approach. With a differential privacy of ∈= 0.1 our experiments report an increase in the measured average error (RMSE) of only 25%. Specifically, +23.13%, 25.71%, +29.87%, 21.57% for macronutrients (grams), calories burned (kCal), resting heart rate (beats per minute (bpm), and minutes (minutes) respectively. / Hälsoövervakningsenheter, som Fitbit, blir allt populärare både som friskvårdsverktyg och som informationskälla för vårdbeslut. Att förutsäga sådana välbefinnandemål korrekt är avgörande för att användarna ska kunna göra välgrundade livsstilsval. Kärnmålet med denna avhandling är att designa och implementera ett sådant system som tar hänsyn till energiförbrukning och integritet. Denna forskning är modellerad som ett tidsserieprognosproblem som använder sig av SNNs på grund av deras bevisade energibesparingsförmåga. Tack vare deras design som nära efterliknar naturliga neurala nätverk (som hjärnan) har SNNs potentialen att avsevärt överträffa klassiska artificiella neurala nätverk när det gäller energiförbrukning och robusthet. För att bevisa våra hypoteser har en tidigare forskning av Sonia et al. [1] i samma domän och med samma dataset används som utgångspunkt, där ett privat prognossystem som använder LSTM designas och implementeras. Deras studie implementerar och utvärderar också en klustringsstrategi för federerad inlärning, som passar väl in på den mycket distribuerade data. Resultaten som erhållits i deras forskning fungerar som en baslinje för att jämföra våra resultat vad gäller noggrannhet, träningstid, modellstorlek och uppskattad energiförbrukning. Våra experiment visar att Spiking Neural Networks byter ut precision (2,19x, 1,19x, 4,13x, 1,16x större RMSE för makronäringsämnen, förbrända kalorier, vilopuls respektive aktiva minuter), för att ge en mindre modell ( 19% mindre parametrar, 77% lättare i minnet) och 43% snabbare träning. Vår modell beräknas förbruka 3, 36μJ, vilket är mycket lättare än traditionella ANNs [2]. Data som registreras av hälsoövervakningsenheter är enormt spridda i den verkliga världen. Dessutom, med sådan känslig registrerad information finns det många möjliga konsekvenser att överväga. Av dessa skäl tillämpar vi klustringsimplementeringen för federerad inlärning [1] på vårt användningsfall. Det kan dock vara utmanande att använda sådana tekniker eftersom det kan vara svårt att lära sig av datasekvenser som är oregelbundna. Vi använder en tvåstegs streaming-klustringsmetod för att klassificera kunder baserat på deras mat- och träningsvanor. Det har visat sig att det är användbart att träna olika modeller för varje grupp av användare, särskilt när det gäller utbildningstid; detta är dock starkt beroende av klustrets storlek. Våra experiment drar slutsatsen att det finns en minskning av fel och träningstid om klustren innehåller tillräckligt med data för att träna modellerna. Slutligen tar denna studie upp frågan om datasekretess genom att använda den senaste differentiell integritet. Vi tillämpar e-differentiell integritet på både vår baslinjemodell (utbildad på hela datasetet) och vår federerade inlärningsbaserade metod. Med en differentiell integritet på ∈= 0.1 rapporterar våra experiment en ökning av det uppmätta medelfelet (RMSE) på endast 25%. Specifikt +23,13%, 25,71%, +29,87%, 21,57% för makronäringsämnen (gram), förbrända kalorier (kCal), vilopuls (bpm och minuter (minuter).

Bakgrund Ländryggssmärta är mycket vanligt förekommande i västvärlden. Det innebär stort lidande för individen och stora kostnader för samhället. Idag behandlas ryggsmärta med information om att vara aktiv och vid behov rådgivning om smärtstillande läkemedel. Mer kunskap behövs gällande fysisk träning som prevention och behandling av akut- och subakut ländryggssmärta. Syfte Syftet med studien var att, för patienter med akut- och subakut ländryggsmärta, beskriva den fysiska aktivitetsnivån samt eventuella skillnader i smärtintensitet och funktionsnedsättning mellan grupper med olika aktivitetsnivåer. Syftet var även att undersöka samband mellan aktivitetsnivå och smärtintensitet respektive aktivitetsnivå och funktionsnedsättning. Metod Studien var en enkätstudie med tvärsnittsdesign. Deltagarna var 15 patienter, 9 kvinnor och 6 män, medelålder 49,2 år, som sökt vård för akuta eller subakuta ländryggsbesvär till 4 olika primärvårdsenheter. Fysisk aktivitetsnivå skattades via Socialstyrelsens indikatorfrågor för fysisk aktivitet, smärtan med numerisk skattningsskala 0-10 (NRS) och Roland Morris Disability Questionnaire (RMDQ) besvarades. Data sammanställdes med deskriptiv statistik, skillnader testades med Mann-Whitney U-test och samband analyserades med Spearmans korrelationskoefficient. Resultat Åtta av 15 deltagare uppnådde Världshälsoorganisationens (WHO) rekommendationer för fysisk aktivitet (> 150 minuter i veckan). De som ägnade sig åt fysisk träning minst 90 minuter per vecka hade medianvärde NRS 5,5 och RMDQ 8, för de som tränade mindre var motsvarande värden NRS 7,5 (p=0,153) och RMDQ 11,5 (p=0,175). Ett svagt negativt samband identifierades mellan NRS (r=-0,316,) och nivå av fysisk aktivitet medans sådant samband mellan RMDQ och fysisk aktivitetsnivå var negligerbart (r=-0,158). Slutsats Det var ingen statistiskt signifikant skillnad mellan grupperna som tränade minst 90 minuter per vecka och de som tränade mindre gällande varken smärtintensitet eller funktionsnedsättning. Ett svagt negativt men ej statistiskt signifikant samband kan ses mellan fysisk aktivitetsnivå och smärtintensitet. / Background Lower back pain is very common in the western world. It results in a great suffering for the person and large economic costs for the society. Today lower back pain is treated with information to stay active and medication for pain relief. There is a lack of knowledge with regards to what effect physical training has as prevention and treatment for acute and subacute lower back pain. Aim The aim of this study was to, among patients with acute and subacute non-specific lower back pain, describe their level of physical activity and evaluate differences between groups with different levels of activity. The aim was also to explore the association between pain intensity, disability and level of physical activity. Method The study was conducted as a survey. The participants was 15 patients, 6 men and 9 women with the mean age of 49,2 years old, who had sought care for acute and subacute lower back pain in 6 different primary care clinics. The level of physical activity were estimated using the indicator questions for physical activity by Socialstyrelsen, the pain intensity was measured with the Numeric Rating Scale and the Roland Morris Disability Questionnaire was answered. Data was analyzed with descriptive statistics, differences were tested with Mann-Whitney U-test and correlations analyzed with Spearman correlations coefficient. Results Eight out of 15 participants reached the WHO recommendations of physical activity (> 150 min/week). Those who participated in physical training minimum 90 mins/week had a median value of NRS 5,5 and RMDQ 8. For those who trained less the median values were for NRS 7,5 (p=0,153) and RMDQ 11,5 (p=0,175). A week correlation between NRS (r=-0,136) and level of physical activity was noted while such correlation between RMDQ was negligible (r=-0,158). Conclusion There were no statically significant difference between the groups who trained at least 90 minutes per week and those who trained less neither in regards to pain intensity or disability. A weak but not statistically significant correlation was observed between physical activity and pain intensity.