Physical activity and mental health in India: A narrative review

Type of study

The articles included in this review followed either a cross-sectional or intervention study design. A majority of studies were conducted in clinical settings (Behere et al., 2011; Duraiswamy, Thirthalli, Nagendra, & Gangadhar, 2007; Ganesh & Mishra, 2016; Krishnamurthy & Telles, 2007; Manjunath, Varambally, Thirthalli, Basavaraddi, & Gangadhar, 2013; Ramanathan, Bhavanani, & Trakroo, 2017; Roy, Govindan, & Muralidharan, 2018; Telles, Singh, Yadav, & Balkrishna, 2012; Varambally et al., 2012), while other study settings included educational institutions, community, geriatric centres, a park, and a corporate organization (Deb et al., 2016; Ghosh & Datta, 2012; Ghrouz et al., 2019; Godse, Shejwal, & Godse, 2015; Hallam, Bilsborough, & de Courten, 2018; Mukherjee et al., 2017; Nayak, Mohapatra, & Panda, 2019; Ray et al., 2001; Singh et al., 2018; Telles et al., 2018). The recruitment of clinical patients in many of these studies could prevent results from being generalizable to a broader population of mentally ill individuals. Seven intervention studies utilized a randomized controlled trial (RCT) design with a control group or other intervention group for comparison (Behere et al., 2011; Duraiswamy et al., 2007; Godse et al., 2015; Manjunath et al., 2013; Ramanathan et al., 2017; Ray et al., 2001; Varambally et al., 2012). Use of RCTs for comparison among groups was a more effective way to assess outcomes and odds ratios than allowing participants to choose a group. Studies that used waitlist control groups allowed for a more robust method of comparing outcomes via clinical scores.

There were no longitudinal studies found comparing physical activity for mental disorders. Studies reporting short-term results show differences between groups attributed to the intervention but do not have the capacity to assess the lasting effect of a programme after its completion. Convenience sampling was the most common sampling strategy, as most participants were patients from the hospitals where studies took place or students recruited from institutions.

Instruments used to measure physical activity and mental health/illness

Studies used Diagnostic and Statistical Manual of Mental Disorders (DSM) and International Classification of Diseases (ICD) to diagnose mental illnesses. Self-rated, teacher-rated and clinician-rated questionnaires were used for assessment. Among all tools used, the Positive and Negative Syndrome Scale (PANSS) was mostly commonly administered (Behere et al., 2011; Duraiswamy et al., 2007; Manjunath et al., 2013; Varambally et al., 2012). Majority of studies used more than one tool for assessment, where studies on schizophrenia used PANSS along with Simpson Angus Scale (SAS) or Social and Occupational Functioning Scale (SOFS); studies on depression or anxiety frequently used Hamilton Depression Rating Scale (HDRS) or State Trait Anxiety Inventory (STAI) with Warwick-Edinburgh Mental Well-Being Scale (WEMWBS) (Behere et al., 2011; Duraiswamy et al., 2007; Manjunath et al., 2013; Roy et al., 2018; Telles et al., 2018; Varambally et al., 2012). All tools used in the studies were validated and appropriate for use among Indian patients. No study created their own questionnaires or tools, thereby assuring more reliable and clinically valid results.

There was no standard measurement of physical activity via assessment tools which could have resulted in variations in reporting prevalence. A few studies used validated measure like the Physical Activity Scale for Individuals With Physical Disabilities (PASIPD), the World Health Organization (WHO) Disability Assessment Schedule 2.0 (WHODAS 2.0), and the International Physical Activity Questionnaire–Short Form (IPAQ) (Ganesh & Mishra, 2016; Ghosh & Datta, 2012; Ghrouz et al., 2019; Mukherjee et al., 2017). The remaining studies measured prevalence by designing their own questionnaires asking participants whether they were active or not (Deb et al., 2016; Ghosh & Datta, 2012; Nayak et al., 2019; Singh et al., 2018). Medical data were not collected except in one study which recorded body mass index (BMI) of the participants and determined cut-off values for obesity according to WHO guidelines (Misra et al., 2009; Singh et al., 2018). Using unstandardized tools and questionnaires could have neglected reporting on the nature of the physical activity as well as its intensity and duration. Studies that asked closed yes or no questions may have received subjective and inaccurate responses when no criteria for physical activity were implied.

Intervention programmes for depression and anxiety

Yoga was also used as an intervention for assessing changes in neurosis in six studies (Godse et al., 2015; Krishnamurthy & Telles, 2007; Ramanathan et al., 2017; Ray et al., 2001; Telles et al., 2018, 2012). Clinical variables like depression, anxiety, somatization of stress, quality of life, sleep quality and self-esteem were studied. All yoga programmes followed similar schedules including asanas, breathing techniques and relaxation albeit from various types and teachings of yoga like Hatha yoga, Patanjali yoga, Bhakti yoga as well as Suryanamaskar, yoga theory and a protocol developed by Centre for Yoga Therapy, Education, and Research (CYTER) (Godse et al., 2015; Ramanathan et al., 2017; Telles et al., 2012). The duration of these programmes ranged from 1 week to 10 months, while programmes had participants practice yoga for as short as 20 minutes to as long as 5 hours. There was no uniformity in the structure, length or type of yoga among any of the programmes. Comparison of these studies’ respective programmes is based on reported outcomes of statistical significance among participants. Although the intensity of various types of yoga programmes differs, they all incorporated elements of relaxation, breathing techniques and meditation in their practice. Inclusion of such techniques could have influenced outcomes of these programmes. Similar incidences have been reported in studies which attribute meditation and relaxation to reducing stress as well as symptoms of anxiety and depression (Ospina et al., 2007; Ramanathan et al., 2017; Shannahoff-Khalsa, 2004; Srivastava, Talukdar, & Lahan, 2011). Yoga breathing or pranayamas was found by one meta-analysis to be one of the most effective techniques for reducing symptoms of anxiety and post-traumatic stress disorder (Cabral, Meyer, & Ames, 2011).

Yoga brought about a statistically significant improvement in scores, thus indicating decreased levels of anxiety and depression, two clinical variables common to all six studies. A reduction in feelings of sleepiness and stress dispositions was found in one study that practised Suryanamaskar, while other studies found significantly reduced somatization of stress and improved self-rated quality of sleep, significant improvement in self-esteem scores and improved total sense of mental well-being. Similar reductions in clinical scores and improvement have been found in other studies which also highlight that yoga can be an inexpensive and effective add-on intervention (Cabral et al., 2011). All studies promoted yoga for neurosis and hypothesized that its benefits in reducing clinical scores could be due to its effect on reducing stress, which is known to exacerbate health problems. Yoga has a calming effect on stress responses in the body, potentially reducing levels of anxiety and depression (Ramanathan et al., 2017). Numerous studies have reported yoga’s beneficial effect of stress reduction among healthy adults (Chong, Tsunaka, Tsang, Chan, & Cheung, 2011). All studies reported that participants experienced increased physical relaxation, mental peace, strength and awareness as well as lower stress dispositions. Studies’ control groups experienced no change on any clinical scale and thus no measured improvement was found (Godse et al., 2015; Krishnamurthy & Telles, 2007; Ramanathan et al., 2017; Ray et al., 2001; Telles et al., 2018, 2012). Improvement in the yoga groups of these studies promotes its use as a treatment for neurotic disorders like anxiety and depression. Further assessments need to be made on other neuroses in order to bring to light the extent of yoga’s benefit for mental disorders.

Two studies used exercise as an intervention for mental health disorders. Roy et al. (2018) utilized a RCT to designate patients to either intervention or control group and conversely Hallam et al. (2018) did not use a comparison group. The lack of randomization and control group could have produced subjective results among the employee participants recruited for the study (Hallam et al., 2018). Furthermore, the narrow representation of participants in these studies makes results appear less generalizable as study participants were from either a hospital or corporate office.

There was no common exercise programme followed by the studies as each utilized participants with varying abilities and settings. Interventions ranged in duration from 10 days to 100 days and involved either an exercise programme with walking or jogging for a minimum of 10,000 steps per day as proposed by Hallam et al. (2018) or an add-on video assisted aerobic exercise programme with warm-ups, low intensity aerobic exercises and cool-down for 20 minutes a day designed by Roy et al. (2018). Short-term programmes may have insufficient amount of time to report clinical changes; however, both studies found significant improvements.

Major findings from the two reviewed studies state that exercise as an intervention can significantly reduce stress, depression and anxiety (Hallam et al., 2018; Roy et al., 2018). Roy et al. (2018) found no significant reduction in somatic symptoms as well as no improvement in the control group. The significant changes reported by these studies stand out where other studies and reviews globally reported that exercise programmes brought about only moderate or little change in measured outcomes (Pearsall, Smith, Pelosi, & Geddes, 2014). These findings could imply that exercise programmes of this calibre have more of an impact on Indian participants than those in other countries. Overall, these studies are consistent with literature found globally on the use of exercise for reducing symptom scores in patients with depression and anxiety (Callaghan, 2004; Rimer et al., 2012). Only a small number of studies conducted in India reported an association between exercise and mental disorders like anxiety and depression. These report the beneficial use of exercise for improvement of clinical scores. However, in order to derive a conclusion more research is required from India.

Intervention programmes for psychosis

Four studies have assessed the effect of physical activity on symptoms of psychosis, using a sample of patients with schizophrenia. Researchers had assessed patients’ clinical state prior to the study to determine their ability to take part and give consent. All studies’ participants were either inpatients or outpatients and were monitored throughout or by caregivers at home recording progress in a log book.

All studies compared yoga and exercise; however, only two studies used waitlist control groups receiving no add-on treatment for comparison along with yoga and exercise groups (Behere et al., 2011; Varambally et al., 2012). All studies were RCTs and single blinded to evaluators; no studies were double-blinded. The utilization of RCT for intervention and waitlist control groups allowed for a systematic comparison of the effectiveness of each group and their improvement on clinical scales. Stabilization of medication for all patients recruited in these studies facilitated a more ethical and reliable assessment. Three out of four study samples were on antipsychotic medication for several months without any changes in medication dosage for a period of at least 4 weeks prior to the intervention and during the study period in order to assess results attributed to the programme and not due to effects of medication (Behere et al., 2011; Duraiswamy et al., 2007; Varambally et al., 2012).

All yoga programmes followed similar schedules adapted from the integrated yoga treatment developed by Swami Vivekananda Yoga Anusandhana Samsthana (SVYASA) (R. Nagarathna & H. R. Nagendra, 2004). The use of a single SVYASA programme for yoga among all studies on patients with schizophrenia allowed a more systematic method of assessing and reviewing reported outcomes. Yoga sessions lasted around an hour and were held five times a week. All studies’ programmes consisted of (a) Sithilikarana Vyayama, (b) asanas, (c) breathing practice and (d) relaxation techniques. Practice of Vyayama, asanas and breathing techniques were in equal measure (R. Nagarathna & H. R. Nagendra, 2004). Meditation was not included in any studies’ yoga regime. Similarly, a structured programme was followed for exercise groups. All exercise programmes were adapted from the National Fitness Corps: Handbook for Middle High and Higher Secondary Schools (Ministry of Education, Government of India, 1965). The regime included brisk walking, jogging and exercises in standing and sitting postures as well as relaxation. All sessions were around an hour and held five times a week. The yoga and exercise programmes utilized by all studies ranged in duration from 6 weeks to 4 months. All studies followed the same structured and well-established programmes, a feature which adds legitimacy and aids in comparison of results.

Minor differences were found in the method of monitoring patients participating in these programmes. The continuation of practice at home and logging participation and duration of yoga or exercise had drawbacks, where many patients either did not stick to the schedule of the programme or caregivers had not consistently logged their progress (Duraiswamy et al., 2007; Varambally et al., 2012). The studies found that allowing patients or caregivers to log progress was not a reliable method of confirming adherence to the programme (Behere et al., 2011; Duraiswamy et al., 2007; Manjunath et al., 2013; Varambally et al., 2012). Absence of a third party monitoring progress may have impacted the credibility of the results, especially those reporting significant improvements as a result of physical activity. This led to the exclusion of some participants who had not followed the programme as instructed. These dropouts ranged from 16 to 28 and were common in all studies.

Major findings of these studies show that yoga has significant benefits in reducing mean clinical scores in patients with schizophrenia. Studies that planned interventions longer than 2 weeks showed significantly lower mean scores in patients on the Clinical Global Impression Severity (CGIS), PANSS general psychopathology subscore, and HDRS than studies conducted for a period of 2 weeks or less. Manjunath et al. (2013) compared results at 2 weeks and at 6 weeks and found improvement only in the latter assessment. Benefits of both programmes have been recorded by all studies; however, significantly more patients in the yoga group had improved PANSS and socio-occupational functioning scores compared with the exercise or waitlist group (Behere et al., 2011; Duraiswamy et al., 2007; Varambally et al., 2012). In addition, yoga improved facial emotion recognition deficit, improved quality of life, and lowered scores on HDRS and CGIS (Behere et al., 2011; Duraiswamy et al., 2007; Manjunath et al., 2013). All studies highlight the advantage of yoga over exercise in reducing clinical mean scores and improving social-occupational functioning in patients with schizophrenia. The likelihood of improvement in the yoga group according to odds ratio analysis for negative symptoms was five times greater than either the exercise or waitlist groups in one study (Varambally et al., 2012). Furthermore, Behere et al. (2011) found no significant differences in mean scores in the exercise or waitlist group on any scales. These results are consistent with global studies reporting little or no effect of exercise on severe mental disorders (Pearsall et al., 2014; Vancampfort et al., 2012). Such results have been hypothesized to be due to insufficient intensity of exercise or short duration of its practice (Pearsall et al., 2014).

Improvement of negative symptoms in schizophrenia has been found in all studies after add-on yoga treatment (Behere et al., 2011; Duraiswamy et al., 2007; Manjunath et al., 2013; Varambally et al., 2012). Interestingly, Varambally et al. (2012) found no improvement in positive symptoms from any intervention group (yogasana, exercise, or waitlist) but significant reduction in negative symptoms. Existing low scores for positive symptoms on PANSS due to stabilized antipsychotic treatment could be a possible explanation for no significant change in positive symptom scores for some studies (Duraiswamy et al., 2007).

Studies’ outcomes imply that yoga could be a beneficial add-on treatment for negative symptoms in antipsychotic stabilized patients with schizophrenia (Behere et al., 2011). A small number of participants required changes in their medication dosage during the studies’ duration, suggesting that yoga could be used as an intervention by patients with schizophrenia without aggravating psychotic experiences (Duraiswamy et al., 2007; Manjunath et al., 2013). Yoga programmes implemented breathing exercises like diaphragmatic breathing during various postures and an attentive component practised through meditation and relaxation. These could have an added benefit over physical movement alone. Such features of yoga are even attributed to improving social functioning (Bangalore & Varambally, 2012; Varambally et al., 2012). These studies emphasize that implementation of yoga for functional non-organic psychosis is an effective and practical approach.

Limitations of these studies involved implementation of brief yet intensive intervention designs, single-blind methodology, small sample sizes, drop-outs due to lack of interest or travel difficulties, and inconsistently kept logs by patients or caretakers (Behere et al., 2011; Duraiswamy et al., 2007; Manjunath et al., 2013; Varambally et al., 2012). Studies have found that benefits of yoga therapy are not long term after intervention is ceased and continuous treatment would be required to maintain the effects (Duraiswamy et al., 2007). In addition, patients were not deterred from participating in moderate exercise or games conducted in hospital wards and this could have affected the progress of the groups (Manjunath et al., 2013). Normal routines that involve physical activity of a nature different than that of yoga or exercise could have affected results. Since these studies could not monitor or record possible physical exertion that could be classified as physical activity, the possibility of such an influence cannot be ignored. Finally, no studies reported the effect of yoga on drug-induced extrapyramidal side effects (Behere et al., 2011; Duraiswamy et al., 2007; Manjunath et al., 2013; Varambally et al., 2012). Since antipsychotic side effects are difficult to control, this would be a significant finding if studied among Indian participants.