< Back to Publication

Shu Ling, Tan

A Longitudinal Study of Orthopedic Rehabilitation Patients: Physical Activity, Subjective Health and Positive Affect

Shu Ling, Tana*, Yan Ping, Duanb, Sonia Lippkeac

aDepartment of Psychology and Methods, Health Psychology and Behavioral Medicine, Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany. Email: s.tan@jacobs-university.de

bDepartment of Physical Education, Faculty of Social Sciences, Hong Kong Baptist University, ShekMun Campus, 8 On Muk Street, Shatin, Hong Kong, China

cBremen International Graduate School of Social Sciences (BIGSSS), Jacobs University Bremen, Campus Ring 1, 28759 Bremen, Germany

*Corresponding author.


Every individual – including those with a disability or chronic diseases – has positive assets, like positive affect and positive self-perceived health, that could be essential resources during or after a rehabilitation experience in clinical settings (Nierenberg et al., 2016). The experience of positive affect is essential for healthy functioning and having a better quality of life across one’s lifespan (Scheibe, English, Tsai, & Carstensen, 2013). In addition, positive self-perceived health or subjective health is not only an important outcome measure, but also an influential predictor for different health behaviors that are significant in clinical settings and rehabilitation programs (Doiron, Fiebig, Johar, & Suziedelyte, 2014; Hunt & McEwen, 1980; Liu et al., 2016). For instance, if patients are advised to adopt physical activity, they might feel constrained by their health status, but they might also perceive the benefits when initiating and maintaining a new behavior in terms of their perceived health.

Seligman (2008) outlined the importance of measuring individuals’ subjective health to identify correlates of other health-related constructs, such as physical activity and positive affect. For example, the health benefits of subjective well-being (Diener & Chan, 2011) and positive affect (Fredrickson, 2000; Salovey, Rothman, Detweiler, & Steward, 2000; Schmidt, Ziemer, Piontkowski, & Raque-Bogdan, 2013) have been well reviewed.

However, many rehabilitation patients with orthopedic conditions usually do not meet the recommended level of physical activity (Peiris, Taylor, & Shields, 2013), and their subjective well-being is not always optimal. Therefore, it is important to examine the possible physical and psychological factors that influence the functioning of individuals, especially in clinical settings.

Theoretical Framework – Subjective Well-Being

Subjective well-being is commonly defined by (1) affective components of more positive affect and less negative affect, as well as (2) the cognitive component of satisfaction with different domains of life (Diener, 2006; Diener & Chan, 2011; Kashdan, Biswas-Diener, & King, 2008), such as satisfaction with health. Based on the bottom-up approach of the needs conceptual model of subjective well-being (Gataūlinas & Banceviča, 2014), satisfaction with health forms life satisfaction as a whole, through its role in the satisfaction of needs, since subjective health indicates self-perceived health beyond physical health, including mental health and well-being.

A previous study found significant effects of subjective health on positive affect and negative affect, suggesting that subjective perception of health is a critical indicator for well-being (Cho, Martin, Margrett, MacDonald, & Poon, 2011). Salovey and colleagues (2000) also suggested that positive affect may promote healthy perceptions, beliefs and physical well-being, where the effects include the prevention of illness and the motivation of health behaviors, with greater self-efficacy. Additionally, the positive affect of feeling calm and peace was strongly related to individuals’ physical health (Scheibe et al., 2013).

In addition to the components of subjective well-being, a substantial amount of research has shown that physical activity has sustainable health benefits (McKinney et al., 2016; Rodrigues, Gomes, Tanhoffer, & Leite, 2014). In addition, physical activity improves mood and well-being, with one previous study having discovered that people who were physically inactive were significantly more likely to experience negative emotions than those who were physically active (Lu et al., 2012). A systematic review concluded that physical activity is associated with more positive affect and higher satisfaction with life, although little is known about the associations between physical activity and well-being changes across the lifespan, as well as the underlying mechanisms of this relation (Kanning & Hansen, 2016).

Past research has typically considered only one or another part of this interrelation, and only a few studies have tested long-term outcomes. Therefore, the current study investigated the relations among subjective health, positive affect and negative affect together, as well as the associations with physical activity, over the course of eight years. This study hypothesized that among the previous rehabilitation patients, the means of subjective health and positive affect would increase, while means of negative affect would decrease. We also hypothesized that previous rehabilitation patients who were physically active would report greater subjective health and a stronger growth in subjective health over time.



An eight-year longitudinal design was implemented in Germany. This study presents data from a larger interdisciplinary trial, with different aspects related to physical activity, self-regulation and social-cognitive factors. Some data have been published separately (Lippke, Ziegelmann, & Schwarzer, 2004; Paech & Lippke, 2017; Ziegelmann, Lippke, & Schwarzer, 2006), but subjective health and emotion-related questions have not been addressed before, and have not been included in previous analyses.

Participants and Procedures

The study was carried out in an outpatient orthopedic rehabilitation center, where patients were admitted due to different musculoskeletal diseases. All participants were advised to participate in a three-week exercise therapy program on a daily basis during rehabilitation, to enhance their physical functioning and well-being.

Informed consent was obtained and the ethical principles of the American Psychological Association (APA, 1992) were met. After obtaining informed consent, all participants were given paper-and-pencil self-report questionnaires prior to rehabilitation as a baseline measurement (T1, n=640). After three weeks of rehabilitation, patients were discharged and scheduled to fill out a follow-up questionnaire at six months (T2, n=494), at three years (T3, n=330) and eight years (T4, n=224).

The participants were aged between 18 to 80 years (Mage=46.31; SD=11.74), 61.2% (N=389) were female and the average BMI was 26.05kg/m². Among the participants, 54% were working fulltime and 70% were married and/or living with a partner.


Since this study was carried out in Germany, German versions of the questionnaire were used. All items given below are examples items from English versions of the measures.

Demographic variables. Socio-demographic characteristics included: gender, year of birth, marital status, education level and BMI.

Subjective health, positive affect and negative affect. The twelve-item Short-Form Health Survey (SF-12) is a measurement used to evaluate health-related quality of life, predominantly for well-being in physical and emotional dimensions of life (Farivar, Cunningham, & Hays, 2007; Ware, Kosinski, & Keller, 1996). In this study, the item for general health was used to access participants’ subjective health by asking ‘In general, how would you rate your health?,’ ranging from ‘Excellent’ to ‘Poor’. To access positive affect and negative affect, two items were used – ‘…have you felt calm and peaceful?’ to access participants’ positive affect, and ‘…have you felt low and downhearted?’ to access participants’ negative affect – with scale scores ranging from 1 (All of the time) to 6 (None of the time). Past studies have used these two perceived mental health items to access depressive symptoms for screening purposes, targeting treatment and prevention (Vilagut et al., 2013).

Physical activity. The stages algorithm developed by Lippke and colleagues (2010) was used to access participants’ physical activity. Participants were asked ‘Did you engage in physical activity at least three times per week, for at least 30 minutes or more, in such a way you were moderately exhausted?,’ whereby they could answer either ‘Yes’ or ‘No’, the same as the following question: ‘Do you intend to start (new or old activities) soon?’ Participants who implied being active in the past were categorized as ‘Active,’ while those who implied that they had not been active and/or intended to perform physical activity were categorized as ‘Inactive.’ The reliability and validity of this measure was found to be high (Lippke, Fleig, Pomp, & Schwarzer, 2010) and it has been used in other studies (Jackson, Lippke, & Gray, 2011).

Data Analyses

All data analyses were performed with SPSS 24 software. Drop-out analyses were performed with Chi-square (c2) tests and t-tests. A correlation analysis was run to determine the interrelations among the main variables of this study (see supplemental material).

Even though drop-out is a common phenomenon in longitudinal studies within health-related studies (Bhamra, Tinker, Mein, Ashcroft, & Askham, 2008), Linear Mixed Models (LMM) with the maximum likelihood method was used. This is mainly due to LMM being well-adapted for an unbalanced sample size across different points in time, which might be due to drop-out or missing values (Seltman, 2016).

For the main analyses of longitudinal data, LMM was used to examine changes in human behaviors over time. The application of the individual growth curve (IGC) model was used to develop different types of polynomial growth curve models to examine individual growth and use non-linear growth curves to describe between- and within-subject changes over time (Field, 2009; Shek & Ma, 2011). Intra-class correlation coefficient (ICC) were identified for variance of the random effects, which were also used as effect size measures (Field, 2009; Rosnow, Rosenthal, & Rubin, 2000; Shek & Ma, 2011).


LMM was carried out and Fsubjective_health(3, 13828.02) = 226.34, p < .001 indicates significant time differences on subjective health, as well as on positive affect and negative affect, with Fpositive_affect(3, 13806.77) = 166.22, p < .001 and Fnegative_affect(3, 13790.50) = 148.78, p < .001, respectively. The means of the main variables, pairwise comparison, and the growth models are reported in Table 1. Based on the values of the ICC, the random intercepts of the participants accounted for 56% of the total variance for subjective health, about 49% of the total variance for positive affect and 50% of the total variance for negative affect.

Table 1.

Summary of the Main Variables in this Study.


Note. Sub. Health = Subjective Health; CI = Confidence Interval. T1 = Baseline; T2 = At 6-month follow-up;

T3 = At 3-year follow-up; T4 = At 8-year follow-up. **p< .01. * p< .05.

In Table 2, the study participants were categorized as being ‘Active’ and ‘Inactive.’ Significant differences in subjective health were found within both groups over time (Ftime(3, 13790.57) = 153.01, p < .001, and Fphysical_activity(1, 14126.36) = 211.83, p < .001, respectively). Means and standard errors for the four time points are reported in Table 2, as well as standardized beta coefficient values, ICC and the growth models. Participants who were physically inactive showed significantly lower levels of subjective health compared to those who were physically active, with b = -0.14, SE = 0.01, t = -9.87, p < .001, 95% CI [-0.16, -0.11].

Table 2.

Model of Subjective Health and Physical Activity

Note. T1 = Baseline; T2 = At 6-month follow-up; T3 = At 3-year follow-up; T4 = At 8-year follow-up; ICC=Intra-class Correlation Coefficient (effect size). **p < .01. * p < .05.


Orthopedic rehabilitation patients with physical limitations quite often report low levels of well-being (Ayers, Franklin, & Ring, 2013) and do not meet the recommended level of physical activity during and after the rehabilitation (Peiris et al., 2013). The findings of this longitudinal study showed that subjective health was significantly positively associated with positive affect, but negatively associated with negative affect, across all the measurement points from baseline to up to eight years of follow-ups. In line with previous studies, significant effects of subjective health were found on positive affect and negative affect (Cho et al., 2011). This finding strengthened the theoretical formation of subjective well-being, which we identified through the associations explaining the relationships of the paradigms of affective component (i.e., positive affect and negative affect) and the cognitive component of subjective health (Diener, 2006; Diener & Chan, 2011; Gataūlinas & Banceviča, 2014; Kashdan et al., 2008). This finding is consistent with the existing literature.

As hypothesized, participants self-reported greater subjective health and positive affect compared to baseline, and showed increased linear trends during follow-up at six months and three years. Although there were signs of slightly reduced subjective health at the eight-year follow-up, it always remained higher than at the baseline. Similar trends in the opposite direction were displayed for negative affect. These findings consolidate previous research claiming that participants experienced more positive affect over time with the increasing age (Scheibe et al., 2013), and experienced less negative affect, since positive affect tends to last longer and negative affect fades quicker at a later age (Carstensen, Mayr, Pasupathi, & Nesselroade, 2000). As Fredrickson (2000) suggested, the experience of positive affect broadens an individual’s typical way of thinking and builds their personal resources for coping, which has an undoing effect on problems initiated by negative affect, like depression, being overweight and health problems.

In these current findings, given that the changes of these elements of subjective well-being happened after the participants were discharged from rehabilitation, the findings could indicate that participants’ well-being increased after the rehabilitation, which is aligned with previous studies (Kanimozhi & Karupaaiah, 2014). A past study suggested that subjective well-being is associated with the perceptions of the new circumstances of life after rehabilitation, which could influence the accountability towards rehabilitation treatments (Fanciullacci, Straudi, Basaglia, & Chisari, 2017). For example, self-perceived health will be improved after rehabilitation. A similar study suggested a cognitive component –in this case, subjective health –whereby the changes in health status could predict subjective well-being (Realo, Johannson, & Schmidt, 2017).

During rehabilitation, daily exercise therapy was introduced to enhance patients’ physical functioning and encourage patients to be more physically active, which may improve their physical and mental health. This is mainly due to most orthopedic rehabilitation patients not only reporting low levels of physical activity, but also self-perceived poor health, thus affecting their motivation to engage in physical activity (Peiris et al., 2013). Consequently, it is imperative to examine subjective health between being physically active and inactive in the longterm. This current finding reveals that individuals who are physically active are more likely to perceive their health as better in the long run, compared to individuals who are physically inactive. The changes of subjective health increased at follow-ups after the rehabilitation where exercise therapy was introduced; thus, the findings may indicate that physical activity can increase the subjective health and well-being of the orthopedic rehabilitation patients. This is aligned with a past study concluding that physical activity is strongly associated with self-perceived health status (Pino et al., 2013). Kull (2002) previously discovered that physically-active individuals experience not only better subjective health, but also better mental health compared with physically-inactive individuals. A systematic review explored a lifespan perspective of physical activity and the components of subjective well-being, emphasizing the importance of their associations and the need to translate these findings into people’s lives (Hyde, Maher, & Elavsky, 2013).

There are limitations that should be acknowledged. These include (1) high drop-out rates and an unequal sample size at different measurement points, and (2) compared to those who dropped out during the follow-ups, participants who completed all time measurements showed higher means of subjective health and positive affect, which may indicate that they were more interested and motivated in improving health and physical conditions after orthopedic rehabilitation. Moreover, (3) the interrelationships of the main study variables are insufficient to determine causal effect, although the findings of the current study have contributed to the understanding of the associations of the main study variables.

Based on our findings, future research should be directed towards uncovering other possible factors like employment status, coping mechanisms, and characteristics of medical rehabilitation, by considering other health behaviors like restorative sleep and healthy eating, which are also critical for maintaining physical well-being and quality of life (Schmidt et al., 2013). Additionally, it is also worth exploring age-group differences, since everyone experiences emotions differently with age, as suggested by Robbins and colleagues (2010).

In sum, the findings are potentially informative for practical behavioral interventions for rehabilitation, and constructive suggestions throughout the human lifespan targeted toward changeable elements. The understanding of body-mind relationships is important in clinical settings. Any positive development or constructive changes across time could be beneficial in alleviating difficulties during or after rehabilitation, as well as improving rehabilitation patients’ quality of life and well-being. The findings shed light on acknowledging the notion that health behaviors like physical activity, and the experience of positive affect may influence individuals’ perceptions of their health. This study not only provides empirical evidence for the subjective well-being theory, but also adds new knowledge of the notion that orthopedic patients can obtain long-lasting subjective well-being if they maintain an active lifestyle and perceive more positive affect. These seem to be key aspects in buffering against negative affective states in their daily living.


The authors wish to thank Dr. Jochen P. Ziegelmann, for his support and his valuable comments. This work is based on a project funded by the German Federal Ministry of Education and Research (Bundesministerium für Bildung und Forschung, BMBF), Grant number: 01HH12003.



American Psychological Association. (1992). Ethical principles of psychologists and code of conduct. American Psychologist47(12), 1597-1611. doi:10.1037//0003-066x.47.12.1597
Ayers, D. C., Franklin, P. D., & Ring, D.C. (2013). The role of emotional health in functional outcomes after orthopaedic surgery: Extending the biopsychosocial model of orthopaedics: AOA critical issues. The Journal of Bone & Joint Surgery, 95(21), e165(1)-e165(7). doi:10.2106/JBJS.L.00799
Bhamra, S., Tinker, A., Mein, G., Ashcroft, R., & Askham, J. (2008). The retention of older people in longitudinal studies: A review of the literature. Quality in Aging and Older Adults9(4), 27-35. doi:10.1108/14717794200800025
Carstensen, L. L., Mayr, U., Pasupathi, M., & Nesselroade, J. R. (2000). Emotional experience in everyday life across the adult life span. Journal of Personality & Social Psychology79(4), 644-655. doi:10.1037//0022-3514.79.4.644
Cho, J., Martin, P., Margrett, J., MacDonald, M., & Poon, L. (2011). The relationship between physical health and psychological well-being among oldest-old adults. Journal of Aging Research2011, 1-8. doi:10.4061/2011/605041
Diener, E. (2006). Guidelines for national indicators of subjective well-being and ill-being. Applied Research in Quality of Life, 1, 151-157. doi:10.1007/s11482-006-9007-x
Diener, E., & Chan, M. (2011). Happy people live longer: Subjective well-being contributes to health and longevity. Applied Psychology: Health and Well-Being, 3(1), 1-43. doi:10.1111/j.1758-0854.2010.01045.x
Doiron, D., Fiebig, D., Johar, M., & Suziedelyte, A. (2014). Does self-assessed health measure health? Applied Economics, 47(2), 180-194. doi:10.1080/00036846.2014.967382.
Farivar, S., Cunningham, W., & Hays, R. (2007). Correlated physical and mental health summary scores for the SF-36 and SF-12 Health Survey, V.1. Health and Quality of Life Outcomes, 5(1), 54. doi:10.1186/1477-7525-5-54
Field, A.P. (2009). Discovering statistics using SPSS (3rd ed.). London: Sage.
Fredrickson, B. (2000). Cultivating positive emotions to optimize health and well-being. Prevention & Treatment, 3(1). doi:10.1037/1522-3736.3.1.31a
Gataūlinas, A., & Banceviča, M. (2014). Subjective health and subjective well-being (the case of EU countries). Advances in Applied Sociology, 4, 212-223. doi:10.4236/aasoci.2014.49026
Hunt, S., & McEwen, J. (1980). The development of a subjective health indicator. Sociology of Health & Illness, 2(3), 231-246. doi:10.1111/1467-9566.ep11340686
Hyde, A., Maher, J., & Elavsky, S. (2013). Enhancing our understanding of physical activity and wellbeing with a lifespan perspective. International Journal of Wellbeing3(1), 98-115. doi:10.5502/ijw.v3i1.6
Jackson, J., Lippke, S., & Gray, C. D. (2011). Stage-specific prediction of physical activity in orthopaedic patients after rehabilitation treatment. International Journal of Sports Psychology, 42, 1-24. doi:10.1016/j.psychsport.2004.11.002
Kanimozhi, A., & Karupaaiah. (2014). Effectiveness of orthopedic rehabilitative nursing care on wellbeing of patients with lower limb fractures in selected hospitals at Puducherry. International Journal of Basic and Applied Medical Sciences, 4(3), 87-91. Available at http://www.cibtech.org/jms.htm
Kanning, M., & Hansen, S. (2016). Need satisfaction moderates the association between physical activity and affective states in adults aged 50+: An activity-triggered ambulatory assessment. Annals of Behavioral Medicine51(1), 18-29. doi:10.1007/s12160-016-9824-6
Kashdan, T., Biswas-Diener, R., & King, L. (2008). Reconsidering happiness: The costs of distinguishing between hedonics and eudaimonia. The Journal of Positive Psychology, 3(4), 219-233. doi:10.1080/17439760802303044
Kull, M. (2002). The relationships between physical activity, health status and psychological well-being of fertility-aged women. Scandinavian Journal of Medicine & Science in Sports, 12(4), 241-247. doi:10.1034/j.1600-0838.2002.00341.x
Lippke, S., Fleig, L., Pomp, S., & Schwarzer, R. (2010). Validity of a stage algorithm for physical activity in participants recruited from orthopedic and cardiac rehabilitation clinics. Rehabilitation Psychology, 55(4), 398-408. doi:10.1037/a0021563
Lippke, S., Ziegelmann, J. P., & Schwarzer, R. (2004). Behavioral intentions and action plans promote physical exercise: A longitudinal study with orthopedic rehabilitation patients. Journal of Sport & Exercise Psychology, 26, 470-483. doi:10.1123/jsep.26.3.470
Liu, B., Floud, S., Pirie, K., Green, J., Peto, R., & Beral, V. (2016). Does happiness itself directly affect mortality? The prospective UK Million women study. The Lancet, 387(10021), 874-881. doi:10.1016/s0140-6736(15)01087-9
Lu, E., Dayalu, R., Diop, H., Harvey, E., Manning, S., & Uzogara, S. (2012). Weight and mental health status in Massachusetts, National Survey of Children’s Health, 2007. Maternal and Child Health Journal, 16(S2), 278-286. doi:10.1007/s10995-012-1145-1
McKinney, J., Lithwick, D.J., Morrison, B.N., Nazzari, H., Isserow, S.H., Heibron, B., & Krahn, A.D. (2016). The health benefits of physical activity and cardiorespiratory fitness. British Columbia Medical Journal, 58(3), 131-137.
Nierenberg, B., Mayersohn, G., Serpa, S., Holovatyk, A., Smith, E., & Cooper, S. (2016). Application of well-being therapy to people with disability and chronic illness. Rehabilitation Psychology61(1), 32-43. doi:10.1037/rep0000060
Paech, J., & Lippke, S. (2017). Social-cognitive factors of long-term physical exercise 7 years after orthopedic treatment. Rehabilitation Psychology62(2), 89-99. doi:10.1037/rep0000136
Peiris, C. L., Taylor, N. F., & Shields, N. (2013). Patients receiving inpatient rehabilitation for lower limb orthopaedic conditions do much less physical activity than recommended in guidelines for healthy older adults: An observational study. Journal of Physiotherapy, 59(1), 39-44. doi:10.1016/S1836-9553(13)70145-0
Pino, L., González-Vélez, A., Prieto-Flores, M., Ayala, A., Fernandez-Mayoralas, G., Rojo-Perez, F…. João Forjaz, M. (2013). Self-perceived health and quality of life by activity status in community-dwelling older adults. Geriatrics & Gerontology International, 14(2), 464-473. doi:10.1111/ggi.12119
Realo, A., Johannson, J., & Schmidt, M. (2017). Subjective well-being and self-reported health in osteoarthritis patients before and after arthroplasty. Journal of Happiness Study, 18, 1191-1206. doi:10.1007/s10902-016-9769-2
Robbins, S. P., Judge, T. A., & Campbell, T. (2010). Organizational behaviour. Harlow, UK: Pearson Prentice Hall.
Rodrigues, E. V., Gomes, A. R. S., Tanhoffer, A. I. P., & Leite, N. (2014). Effects of exercise on pain of musculoskeletal disorders: A systematic review. Acta Ortopédica Brasileira, 22(6), 334-338. doi:10.1590/1413-78522014220601004
Rosnow, R. L., Rosenthal, R., & Rubin, D. B. (2000). Contrasts and correlations in effect-size estimation. Psychological Science, 11(6), 446-453. doi:10.1111/1467-9280.00287
Salovey, P., Rothman, A. J., Detweiler, J. B., & Steward, W. T. (2000). Emotional states and physical health. American Psychologist, 55(1), 110-121. doi:10A037//0003-O66X.55.1.110
Scheibe, S., English, T., Tsai, J., & Carstensen, L. (2013). Striving to feel good: Ideal affect, actual affect, and their correspondence across adulthood. Psychology and Aging28(1), 160-171. doi:10.1037/a0030561
Schmidt, C. K., Ziemer, K. S., Piontkowski, S., & Raque-Bogdan, T. L. (2013). The history and future directions of positive health psychology. In Sinnott, J. (Ed.), Positive psychology: Advances in understanding adult motivation (pp.207-228). New York, NY: Springer.
Seligman, M. (2008). Positive health. Applied Psychology, 57, 3-18. doi: 10.1111/j.1464-0597.2008.00351.x
Seltman, H. J. (2015). Chapter 15: Mixed models: A flexible approach to correlated data. In Experimental design and analysis. Retrieved from http://www.stat.cmu.edu/~hseltman/309/Book/Book.pdf.
Shek, D., & Ma, C. (2011). Longitudinal data analyses using linear mixed models in SPSS: Concepts, procedures and illustrations. The Scientific World Journal, 11, 42-76. doi:10.1100/tsw.2011.2
Vilagut, G., Forero, C., Pinto-Meza, A., Haro, J., de Graaf, R., Bruffaerts, R., …& Alonso, J. (2013). The mental component of the short-form 12 health survey (SF-12) as a measure of depressive disorders in the general population: Results with three alternative scoring methods. Value in Health, 16(4), 564-573. doi:10.1016/j.jval.2013.01.006
Ware, J. E., Kosinski, M., & Keller, S. D. (1996). A 12-item short-form health survey: Construction of scales and preliminary tests of reliability and validity. Medical Care, 34(3), 220-233. Retrieved from: http://www.jstor.org/stable/3766749.
Ziegelmann, J. P., Lippke, S., & Schwarzer, R. (2006). Subjective residual life expectancy in health self-regulation. Journals of Gerontology: Psychological Sciences, 61, 195-201. doi:10.1093/geronb/61.4.


Supplemental Materials

Table 3.

Means, Standard Deviations and Inter-relations (Pearson’s, r) of the Major Study Variables Across Time.

Note. BMI = Body Mass Index; Sub. Health = Subjective Health; Calm & Peace = Positive Affect; Low & Down = Negative Affect; P.A. = Physical Activity.

T1 = Baseline; T2 = After 6 months; T3 = After 3 years; T4 = After 8 years; Physical Activity, dummy variable, with 0 = Inactive, 1 = Active

**p < .01. *p < .05,  two-tailed significance levels of correlations.


Table 4.

Model of Predictors of Subjective Health

Note. BMI = Body Mass Index; Calm & Peace = Positive Affect; Low & Down = Negative Affect; CI = Confidence Interval; ICC=Intra-class Correlation Coefficient (effect size).

T1 = Baseline; T2 = At 6-month follow-up; T3 = At 3-year follow-up; T4 = At 8-year follow-up. Intercept value represents the value for T4 based on the SPSS default settings.


Figure 1. Flow of participants through each measurement points.

Dropout Analysis

Significant results were discovered for age differences, whereby those who dropped out tended to be younger than those who completed all measurement points, particularly at the three-year follow-up (T3: t(325) = -2.124, p = .034) and eight-year follow-up (T4: t(220) = -2.887, p = .004). There was no significant difference for BMI, with all p values being larger than .05. Participants who dropped out at six month (T2, t(483) = -2.926, p = .004) and three year(T3, t(320) = -2.428, p = .015) follow-ups reported a significantly lower level of subjective health compared to those who had completed all measurement points.

Similarly, people who dropped out at six month, three year, and eight year follow-ups appeared to show less positive affect of feeling calm and peaceful, with T2: t(463) = -2.284, p = .023; T3: t(308) = -2.737, p = .006; and T4: t(208) = -2.223, p = .027. A Chi-square test was performed to examine the relation between drop-out and categorical variables of gender and physical activity stages, whereby no significant difference was found, with all p values being larger than .05.