How to Evaluate Mental Wellbeing Using Digital Health Data

Introduction

Mental well-being can be viewed through two primary dimensions: psychological well-being and subjective well-being. Psychological well-being, as described in Carol Ryff’s Six-Factor Model of Psychological Well-Being, is concerned with factors that facilitate a fulfilled and purposeful life. These factors include:

• Autonomy: The sense of being self-directed and independent in your decisions.
• Environmental mastery: The ability to manage your life and the surrounding world effectively.
• Personal growth: The continuous effort to develop and realize your potential.
• Positive relations with others: The capacity to form meaningful, fulfilling relationships.
• Purpose in life: A sense of direction and having goals that bring meaning to your life.
• Self-acceptance: Accepting yourself and acknowledging both strengths and areas for improvement.

Subjective well-being concerns emotional states, life satisfaction, and happiness:

• General emotional health: Emotional experiences such as happiness, contentment, sadness, and stress that influence overall mood.
• Life satisfaction: The cognitive evaluation of your life, including achievements, relationships, and experiences.
• Eudaimonia: Living a life of meaning and purpose.

While some factors influencing mental well-being may be outside our immediate control, certain lifestyle habits can have a significant impact on emotional health. These habits can enhance our sense of autonomy, increase mastery, and ultimately improve mental well-being. Here, we focus specifically on depression levels, characterized by symptoms such as persistent low mood, loss of interest or pleasure in activities, and negative self-perception.

Various studies have consistently shown that individuals with higher levels of depressive symptoms tend to have lower overall activity levels, a higher proportion of activity occurring at night, and a later, more irregular sleep schedule. These findings suggest that it is theoretically feasible to infer elements of a user’s mental well-being via their activity and sleep patterns.

However, these findings are based on data derived from wearable activity trackers, and users of wearable activity trackers still represent a relatively small percentage of the population.

While modern smartphones are equipped with increasingly sophisticated gyroscopes and accelerometers that can measure activity levels with surprising accuracy, discrepancies between measurements made by smartphones and wearable trackers have been observed.

This brings us to a key question: can we observe the same association between activity patterns and depression levels using smartphone-derived activity data?

In our research conducted in collaboration with Prof Nathan Berg of the University of Otago, we have validated that these trends are also observable in smartphone-derived activity data in two separate studies.

1. Using PHQ-9 (Patient Health Questionnaire-9) instrument, an instrument for measuring levels of depressive symptoms, we found that
2. Using the DASS-21 (Depression, Anxiety, Stress Scales - 21) instrument, an instrument for measuring depression along with anxiety and stress levels, we found that

The Sahha Mental Well-Being Score: A Data-Driven Measure of Mental Wellbeing

By combining the latest research findings and data collected in our research studies, we developed the Sahha Mental Well-Being Score. Using the user’s activity and sleep pattern data, the Sahha Mental Well-Being Score aims to capture how well the user’s activity and sleep patterns support positive mental well-being.

The Sahha Mental Well-Being Score provides a sophisticated data-driven approach to connect users’ mental well-being to their activity and sleep patterns. The Sahha Mental Well-Being Score is negatively correlated with DASS depression scores (Pearson’s correlation = -0.284, p = 2.88e-07), meaning that higher DASS scores are associated with lower mental well-being scores. Our data illustrates this:

Conclusion

This score is not diagnostic but instead indicates how well a user’s lifestyle supports positive mental well-being. It empowers organizations with actionable insights into employee or user wellness trends and offers a non-intrusive way to assess emotional health using daily behavioral data.