Purpose in life in adulthood and older adulthood: integrative review

Ribeiro, Cristina Cristovão; Yassuda, Mônica Sanches; Neri, Anita Liberalesso

doi:10.1590/1413-81232020256.20602018

Cristina Cristovão Ribeiro

Faculdade de Ciências Médicas, Universidade Estadual de Campinas. Cidade Universitária Zeferino Vaz, Barão Geraldo. 13083-970 Campinas SP Brasil. cristinaribeiroft@gmail.com

https://orcid.org/0000-0001-6222-3232

Mônica Sanches Yassuda

Faculdade de Ciências Médicas, Universidade Estadual de Campinas. Cidade Universitária Zeferino Vaz, Barão Geraldo. 13083-970 Campinas SP Brasil. cristinaribeiroft@gmail.com

https://orcid.org/0000-0002-9182-2450

Anita Liberalesso Neri

Faculdade de Ciências Médicas, Universidade Estadual de Campinas. Cidade Universitária Zeferino Vaz, Barão Geraldo. 13083-970 Campinas SP Brasil. cristinaribeiroft@gmail.com

https://orcid.org/0000-0002-6833-7668

Collaborations

CC Ribeiro worked directly on the planning, research, execution and final writing of the manuscript; MS Yassuda contributed to the conception, writing, guidance and final revision of the text and AL Neri contributed to the conception, writing, guidance and final revision of the text

Authors, year, country	Objectives	Designs and participants
Boyle et al., 2009¹⁶; USA	To evaluate the association between PiL, health conditions, and mortality in the elderly.	RMAP and MARS: 1,238 older persons (µ age = 75 ± 7.8) without dementia at baseline (BL), with follow-up 5 years later.
Boyle et al., 2010²⁵; USA	To test the hypothesis that the higher the level of PiL in older persons, the lower is the risk for MCI and AD.	RMAP: 951 elderly without dementia at BL, with 7 annual follow-up measurements, when the levels of incident MCI and changes in cognitive function were also evaluated.
Boyle et al., 2010²⁶; USA	To test the hypothesis that the higher the level of PiL, the lower is the risk of disability occurring in older persons without dementia.	RMAP: 970 elderly without dementia. Baseline measurements and annual follow-ups.
Boyle et al., 2012²⁷; USA	To test the hypothesis that PiL reduces the deleterious effects of pathological changes caused by AD on cognition in old age.	RMAP: 246 participants without dementia at BL with annual follow-ups.
Yu et al., 2015²⁸; USA	To test the hypothesis that higher levels of PiL are associated with lower risk for stroke.	RMAP: 43 elderly without dementia, with annual clinical exams and postmortem cerebral autopsy.
Wilson et al., 2018²⁹; USA	To test the hypothesis that high levels of PiL in ambulatory patients are associated with a lower risk of subsequent hospitalizations.	RMAP: 805 participants initially with µage = 81.1 ± 6.8, without dementia. Medicare Records (1999-2010).
Kim et al., 2013³⁰; USA	To investigate the extent to which PiL is associated with myocardial infarction in adults and older persons with coronary heart disease.	HRS: 1,546 participants from the 8th (2006) and 9th (2008) waves; 50 years +, coronary disease at baseline.
Kim et al., 2013³¹; USA	To determine whether PiL is related to reduction of the incidence of stroke.	HRS: 6,379 participants of the HRS who have never had a stroke (50 years and over). Data from the 8th (2006), 9th (2008) and 10th wave (2010) and data on the circumstances of death (of the deceased person).
Kim et al., 2014³²; USA	To investigate the associations between PiL and use of preventive services and between PiL and hospitalizations.	HRS: 7,168 participants of the HRS (50 years and +), followed-up for 6 years.
Kim et al., 2015³³; USA	To test the hypothesis that PiL is associated with a reduction in the incidence of sleep disorders.	HRS: 4,144 participants without or with minimal sleep disturbances at baseline (2006) participated in a follow-up study (2010).
Kim et al., 2017³⁴; USA	To evaluate whether a high level of PiL in older patients with good physical functioning is associated with a lower risk for low grip strength and slow gait.	HRS: Follow-up measurements (2006 and 2010); µage men = 70.8 ± 6.5; µage women = 63.0 ± 8.2.
Schaefer et al, 2013³⁵; USA	To test whether PiL is associated with better emotional recovery after exposure to negative stimuli.	MIDUS II: Experimental study with a sub-sample from 2006 (n = 584). Randomized presentation of 90 figures with positive, negative and neutral emotional stimuli and attention task randomly interrupted by distractors.
Burrow et al., 2014³⁶; USA	To test the moderating role of PiL among negative affections in daily life and current, past and future satisfaction.	MIDUS II: Secondary data analysis of 1,745 participants (µage = 56.51 ± 12.08).
Hill, Turiano, 2014³⁷; USA	To investigate whether PiL promotes longevity in adulthood and old age.	MIDUS: 6,163 adults and elderly individuals aged 20 to 75 years (µage = 46.9 ± 12.9), followed-up for 14 years.
Zilioli et al., 2015³⁸; USA	To investigate the prospective associations between PiL and allostatic load over 10 years.	MIDUS II: 985 adults and older persons participating in the Biomarker Sub-Study conducted at the time of the second wave of this study (µage 46.14 ± 11.7).
Hill et al., 2016³⁹; USA	To examine PiL as a predictor of family income and net worth.	MIDUS: 7,108 participants recruited in 1995-1996 and reassessed in 2004 (MIDUS I) and 2006 (MIDUS II) 25 to 74 years of age; µage = 46.5 ± 13.
Lewis et al., 2017⁴⁰; USA	To test the associations between PiL and episodic memory, executive functioning, and global cognition.	MIDUS II: 3,489 participants (µage = 56.3 ± 12.27, varying from 32 to 84 years, with complete records of PiL and cognitive measures.
Hill et al., 2018⁴¹; USA	To examine whether adversities experienced in childhood and adolescence predict a decrease in the sense of PiL in adult life and in old age.	MIDUS II: 3,385 participants (2005-2006) with complete information on sociodemographic data, adversities before age 18, and PiL in MIDUS I.
Hill et al., 2015⁴²; USA	To know the characteristics of men with greater chance to present a high level of PiL and a decrease in PiL over time.	NAS: 587 elderly from the 2003 (µage = 74.1 ± 6.2) and 2006 (n = 386) follow-ups.
Windsor et al., 2015⁴³; Australia	To examine associations between individual differences in the sense of PiL and well-being indicated by health, cognition and depressive symptoms.	ALSA: 1,475 participants (µage = 77.06 ± 6.27) evaluated in 6 moments (1992 to 2010).
Ko et al., 2016⁴⁴; USA	To examine changes in PiL over 5 years.	FLSA: Exploratory study with 163 participants (µ age = 56.37 ± 1.00) (from the 5 waves).
Tomioka et al., 2016⁴⁵; Japão	To investigate the relation between maintenance of pleasurable hobbies and PiL with mortality and declining independence 42 months later.	Prospective data of older persons aged 65 years and over, with an increased risk of death (n = 1853), a decline in BADL (n = 1254) and IADL (n = 1162). Mortality analysis between 2011 and 2014.
Sano, Kyougoku, 2015⁴⁶; Japão	To test the model of structural relations among achievement motive, PiL, social participation and role expectations.	Descriptive cross-sectional study: 281 elderly (µ = 77.1 ± 8.7).
Polenick et al., 2018⁴⁷; USA	To investigate associations between PiL of caregivers and care recipients and the emotional and physical difficulties of caregivers.	Descriptive cross-sectional study: 315 spouses (65 years +) caregivers of partners with functional disability (µ marriage time = 42.75 ± 17.81). NSOC and NHATS.
Hooker, Masters, 2016⁴⁸; USA	To examine the relationship between PiL and level of physical activity measured by an accelerometer.	Descriptive cross-sectional study: convenience sample (n = 104; µage = 35.5 ± 15.5, varying from 18 to 80 years).
Mak, 2011⁵¹; USA	To examine the relationship between goal compliance and PiL, as reported by people with dementia.	Experimental study: 91 elderly (µage = 75.28 ± 9.23); 46 assigned by lot to the experimental group (EG) that carried out an artistic activity oriented to the goal. Participants of the control group (CG) made a creative drawing of their choice.
Burrow et al., 2014⁵²; USA	To explore the association between PiL and comfort of living in contexts with ethnic diversity.	Three studies with white adults, one of them experimental: 1) investigation of the level of PiL and comfort with ethnic diversity; 2) exposure to demographic projections; and 3) essay on demographic projections (EG) or neutral theme (CG) + test of comfort with ethnic heterogeneity.

Authors, year, country	Variables and measures
Boyle et al., 2009¹⁶; USA	Inability for BADL; depressive symptoms; neuroticism; self-reported illness; vital status; and mortality.
Boyle et al., 2010²⁵; USA	Identical measures to the previous study plus evaluation of the size of the network of social relationships and frequency of interactions.
Boyle et al. 2010²⁶; USA	Inability for BADL; mobility; health, neurological and neuropsychological measures; cognitive and psychosocial aspects.
Boyle et al.,2012²⁷; USA	Self-report of physical exercises; genotyping for Apo 4; pathological changes associated with AD assessed by postmortem neuropathological examination of brain structures.
Yu et al.,2015²⁸; USA	Current clinical and postmortem neuropathological examination of brain structures made by examiners who were blind to previous information.
Wilson et al., 2018²⁹; USA	Secondary Medicare data on ambulatory care sensitive conditions which can help prevent hospitalizations; annual measures of cardiovascular risk conditions, depressive symptoms and sociodemographic information.
Kim et al., 2013³⁰; USA	Self-reported health; medical records; smoking, physical activity, alcoholism, BMI, obesity and blood pressure; use of medication, high blood pressure, high cholesterol, diabetes, hospitalization; optimism and positive affections, anxiety and hostility, depression.
Kim et al., 2013³¹; USA	Hypertension, diabetes, blood pressure, BMI and coronary diseases; functional capacity for IADL and BADL; depression, anxiety, hostility, negative affection, optimism, positive affection and social participation.
Kim et al., 2014³²; USA	Use of health services; depression, anxiety and negative affection; religiosity and positive affection.
Kim et al. 2015³³; USA	Psychological well-being1-3; Sleep disorders; anxiety and depression; physical exercises; alcoholism and smoking; self-reported illnesses, self-rated health and disabilities.
Kim et al., 2017³⁴; USA	Measurement of grip strength in Kg f measured by a dynamometer; gait speed indicated by time (s) to travel 2.5 m in usual pace.
Schaefer et al., 2013³⁵; USA	Positive and negative affections and subjective well-being; age, number of blinks, and time between the psychological measures and the experiment.
Burrow et al. 2014³⁶; USA	Perceived change between satisfaction 10 years ago and at the present moment, and about the expectation of satisfaction within the next 10 years; daily telephone interviews (8 days) on the experience of 14 negative affections and on the number and intensity of stressors experienced in the previous day.
Hill, Turiano, 2014³⁷; USA	Positive relationships with others1-3; positive and negative affections; mortality and survival time since the beginning of the study.
Zilioli et al., 2015³⁸; USA	Cardiovascular function, lipid and glucose metabolism, chronic inflammation, sympathetic and parasympathetic nervous system activity, hypothalamic-pituitary-adrenal axis; locus of control, positive and negative affections and positive relationships with others1-3.
Hill et al., 2016³⁹; USA	Personality (traits; 5 factors); one item evaluating satisfaction with life; net worth and family income values.
Lewis et al., 2017⁴⁰; USA	Cognitive test battery applied through telephone calls (immediate and delayed recall of word list, fluency, digits in reverse order, numerical series, counting, self-reported health).
Hill et al. 2018⁴¹; USA	Indicators of adversities before age 18 (4 items for physical abuse, 4 for emotional abuse, 3 for poverty, 3 for home arrangement, and 2 for health at age 16).
Hill et al., 2015⁴²; USA.	Personality (traits; 5 factors); questionnaire on health concerns.
Windsor et al. 2015⁴³; Australia.	Mobility, subjective health and depressive symptoms; information processing speed; episodic memory.
Ko et al., 2016⁴⁴; USA.	Personality (traits; 5 factors) and productivity.
Tomioka et al. 2016⁴⁵; Japan	One item evaluating whether older persons had a reason to live and pleasurable hobbies (PiL indicator); independence in BADL and IADL; alcoholism and smoking, number of diseases, cognition and depression.
Sano, Kyougoku, 2015⁴⁶; Japan.	PiL assessed by a scale of commitment to life/ optimism; mastery over self; depressive symptoms; positive affections; self-reported physical exercises. Objective measures of time, intensity and daily caloric expenditure in activities measured by means of an accelerometer, 3 days a week.
Polenick et al., 2018⁴⁷; USA.	PiL indicated by scoring of scale of achievement motive in rehabilitation; perceived economic condition, out-of-home activities and hobbies; social participation; scale of role expectations.
Hooker, Masters, 2016⁴⁸; USA.	One item on PiL applied to caregivers and care recipients; for caregivers: measures of care tasks, available supports, participation in valued activities, age, health, sex and schooling; for care recipients: physical health and cognition.
Mak, 2011⁵¹; USA.	Tenacious Goal Pursuit scale; measure of PiL associated with experimental activity; measure of severity of dementia and cognitive status through telephone interviews.
Burrow et al., 2014⁵²; USA.	Questionnaire about the perceived change between satisfaction 10 years ago and at the present moment, and between this and the expectation of satisfaction within the next 10 years; daily telephone interviews (8 days) on the experience of 14 negative affections and on the number and intensity of stressors experienced in the previous day.

Authors, year, country	Results of interest
Boyle et al., 2009¹⁶; EUA	Mortality: 12.2%. High PiL score was associated with a greater chance of survival (OR = 0.60, 95% CI 0.42-0.87, p < 0.001). Negative correlations between PiL and depressive symptoms, disability and neuroticism; positive correlations between PiL and income.
Boyle et al., 2010²⁵; EUA	In this study, 16.3% developed AD. Regardless of depressive symptoms, neuroticism, social network size and number of chronic diseases, high levels of PiL were associated with a lower risk for AD (OR = 0.48, 95% CI 0.33-0.69, p < 0.001); elderly people with high PiL scores were 2.4-foldmore likely to remain without AD. High PiL score was associated with a lower risk of MCI (OR = 0.71, 95% CI, 0.53-0.95, p = 0.02) and a slower rate of cognitive decline.
Boyle et al., 2010²⁶; EUA	High levels of PiL were associated with a lower risk of disability in BADL (OR = 0.60.95% CI 0.45- 0.81), IADL (OR = 0.56, 95% CI 0.40-0.78) and mobility (OR = 0.61, 95% CI 0.44-0.84), without differences caused by demographic variables, cognition, depression, social networks, neuroticism, fragility, cardiovascular risk and vascular diseases.
Boyle et al., 2012²⁷; EUA	Association of high levels of PiL with better cognitive function, even in the presence of neuropathological alterations and accumulation of beta-amyloid protein plaques. Among people presenting AD and high levels of PiL, the disease progressed more slowly. Significant interactions between level of PiL and depressive symptoms, social network size, number of diseases, level of physical activity, number of people with Apoε4, time since onset of AD, and death.
Yu et al., 2015²⁸; EUA	Macroscopic infarctions in 154 (34%) and microscopic infarctions in 128 (28.3%) older persons. The higher the PiL level, the lower the risk of having one or more macroscopic infarctions (OR = 0.780, 95% CI = 0.495-1.229).
Wilson et al., 2018²⁹; EUA	In 4 years, 2,403 hospitalizations occurred (442 with primary diagnosis, 1,322 with secondary diagnosis, 279 with no diagnosis for outpatient care, and 233 were not hospitalized; high baseline PiL and during follow-up were associated with a 25.3% lower risk of hospitalization than a moderate level of PiL, an effect not observed among patients not recommended for outpatient care. Chronic diseases and socioeconomic status reduced, but did not eliminate the association of PiL with hospitalization.
Kim et al., 2013³⁰; EUA	Higher PiL levels were associated with a lower risk of myocardial infarction in the subsequent 2-year follow-up (OR 0.73 95% CI 0.57-0.93, p = 0.01). The association remained even with adjustment for the severity of the coronary disease, self-assessment of health, sex, alcohol use, BMI and blood pressure.
Kim et al., 2013³¹; EUA	The higher the PiL score at baseline, the lower is the probability of stroke during the follow-up. At each standard deviation of increase in PiL, an adjusted risk ratio of 0.78 for stroke (95% CI, 0.67-0.91, p = .002) was recorded.
Kim et al., 2014³²; EUA	Each point in PiL was associated with a higher probability of performing cholesterol tests [OR = 1.18, 95% CI = 1.08-1.29] or colonoscopy [OR = 1.06, 95% CI = 0.99-1.14]; mammography (OR = 1.27, 95% CI = 1.16-1.39); Pap smear examination [OR = 1.16, 95% CI = 1.06-1.28]; and prostate examination [OR = 1.31, 95% CI = 1.18-1.45]. Each additional point in PiL led to 17% less nights in the hospital [RP = 0.83, 95% CI = 0.77-0.89].
Kim et al., 2015³³; EUA	A total of 1326 participants developed sleep disorders between 2006 and 2010. In the multivariate analysis, each additional point in PiL was associated with an OR = 0.84 for sleep disorders (95% CI = 0.77-0.92), regardless of sociodemographic variables. The other variables attenuated, but did not eliminate the influence of PiL.
Kim et al., 2017³⁴; EUA	After adjusting for sociodemographic variables, each additional point in PiL was associated with a 13% decrease in the risk of developing low grip strength (95% CI = 1% -23%) and 14% decrease in the risk of developing slow gait (95% CI = 8% -20%). Associations with gait time were maintained for all co-variation models, but associations with grip strength did not reach an acceptable level of significance after adjusting for health factors, depressive symptoms, and baseline health behaviors (RR = 0.91, 95% CI = 0.80-1.04).
Schaefer et al., 2013³⁵; EUA	High PiL level assessed in the previous 2 years predicted a better recovery from exposure to negative stimuli indicated by the number of blinks, even with adjustments for initial reactivity to the stimulus during presentation of figures, gender, age, positive and negative affections, and subjective well-being.
Burrow et al., 2014³⁶; EUA	Hierarchical analysis of conglomerates resulted in three profiles: 1. Continuously high satisfaction (n = 439); 2. Rising satisfaction (n = 1,033) and 3. Declining satisfaction (n = 274), with participants with the highest levels of negative affections in daily life. Older persons in conglomerates 1 and 2 had lower levels of negative affections than those of the 3.
Hill, Turiano, 2014³⁷; EUA	The participants who died (569) were older, male, the less educated, with low PiL level and less positive relationships with others. At each standard deviation of increase in PiL, the risk of dying in the next 14 years decreased by 15%.
Zilioli et al., 2015³⁸; EUA	Higher PiL levels at baseline predicted lower levels of allostatic load 10 years later, even with adjustment of other psychological factors potentially associated with allostatic load. PiL was a strong predictor of individual differences in health-related control locus, which was partially related to allostatic loading.
Hill et al., 2016³⁹; EUA	Participants with higher levels of PiL had higher initial levels of family income and net worth, and were more likely to improve these conditions in the 9-year follow-up, moderated by age rather than sex.
Lewis et al., 2017⁴⁰; EUA	High PiL score was associated with high scores in memory, executive functioning, and global cognition, regardless of sociodemographic variables. Age did not moderate the relationship between PiL and cognition.
Hill et al., 2018⁴¹; EUA	PiL correlated negatively with all indicators of adversity, with the exception of family structure and health adversity. Adversities before age 18 were associated with lower levels of PiL in adult life, regardless of sociodemographic factors. Age did not protect against the negative effects of exposure to adversity.
Hill et al., 2015⁴²; EUA	Fourteen percent (14%) of men declined more than one standard deviation in PiL. PiL levels in 2003 correlated negatively with age and positively with self-reported health and with 5 personality factors, but none predicted changes in PiL in 2006.
Windsor et al., 2015⁴³; Austrália	Individuals with a high sense of PiL had lower levels of disability, better performance in speed and memory tests, better self-reported health, and fewer depressive symptoms. Each additional point in PiL led to 4% decrease in the risk of death (OR 0.96, 95% CI [0.94, 0.97]). Age was associated with increased risk of death (OR = 0.69.95% CI [0.62, 0.78]).
Ko et al., 2016⁴⁴; EUA	The majority had high and stable scores in PiL throughout the 5 years; low scores were also stable. Personality and black race predicted high levels of PiL (effect size = 0.42 and 0.25, p = 0.008 and 0.003).
Tomioka et al., 2016⁴⁵; Japão	A total of 248 elderly died, 119 declined in BADL and 178 in IADL. Absence of pleasurable hobbies or PiL were associated with increased risk of death (OR 2.08; 95% CI 1.47-2.94), decline in BADL (OR 2.74; 95% CI 1.44-5.21 ), and decline in IADL (OR 1.89; 95% CI 1.01-3.55).

Sano, Kyougoku, 2015⁴⁶; Japan	Positive correlations between roles of friend, lover and participant of an organization and achievement motive, social participation, and role expectations. Direct effect of achievement motive on PiL (0.445; p < 0.001); indirect effect mediated by social participation or role expectations (0.170; p < 0.001); total effect on PiL = 0.615.
Polenick et al. 2018⁴⁷; USA	The higher the PiL of caregivers, the less are the physical and emotional difficulties of care. PiL of the care recipients was not associated with the emotional difficulties of the caregivers, but the association between the PiL of caregivers was moderated by the PiL of care recipients. The PiL of caregivers was responsible for reducing their emotional difficulties. The PiL of care recipients did not influence the level of physical difficulties of caregivers. The gender of the caregivers did not affect the associations.
Hooker, Masters, 2016⁴⁸; USA	PiL was positively associated with movements objectively measured by an accelerometer, regular practice of moderate and vigorous physical activity, self-report of physical activity; the associations were not dependent on sociodemographic variables. PiL was positively correlated with optimism, mastery over self, and positive affections, and negatively correlated with depressive symptoms. Absence of relations with the two measures of physical activity.
Schaefer et al., 2013³⁵; USA ^() ()** Deliberately included in Box 3 as a longitudinal study and in Box 4 as an experimental study.	High level of PiL assessed in the 2 previous years revealed a better recovery from exposure to negative stimuli, indicated by the number of blinks, even with adjustments for the initial reactivity to the stimulus during the presentation of figures, sex, age, positive and negative affections and subjective well-being.
Mak, 2011⁵¹; USA	Both bivariate and hierarchical regression analyses showed statistically significant relationships between PiL measurements, which were not moderated by the level of severity of dementia. Those who completed the goal-oriented activity perceived more PiL in the activity than those who completed the non-goal-oriented activity.
Burrow et al., 2014⁵²; USA	In all three studies, high PiL levels had a positive effect on comfort with diversity and attenuated the feelings of threat resulting from exposure to projections of increased ethnic diversity. The effects persisted even when adjusted for positive affections and perceived connections with different ethnic groups.

[TFN2] ^(*)
Deliberately included in Box 3 as a longitudinal study and in Box 4 as an experimental study.

Purpose in life in adulthood and older adulthood: integrative review

Abstract