How Did the Elderly in Tamil Nadu Weather the COVID-19 ......Compared to 12% of the state as a...

The Tamil Nadu Panel Elderly Well-Being at Baseline (2019) COVID-19 Data Collection Well-Being in the COVID-19 Lockdown Counseling Intervention

How Did the Elderly in Tamil Nadu Weather the COVID-19 Lockdown?Evidence from the Tamil Nadu Aging Panel

September 6, 2020

Tamil Nadu’s Elderly and COVID-19 September 6, 2020 1 / 35


Outline

• The Tamil Nadu Aging Panel and Government Partnership

• What we know about the lives of the elderly (Baseline: Jan - Jun 2019)

• How the elderly are coping with the coronavirus pandemic

• Improving mental health and economic well-being



The Tamil Nadu Aging Panel Government Partnership

• Challenge: Demographic transition, with elderly population (55 years or above) in developingcountries slated to grow more than three-fold between 2011 and 2050

• Tamil Nadu: from 7.2% in 2011 to 22%+ by 2050• Insufficient data to inform policy to address this changing context

• Solution: create a high-quality 7-year-long panel dataset

• Inform design and implementation of social safety schemes (pensions, cash transfers, publicdistribution), health policy, mental health interventions

• Launch an RCT with an intervention to fight loneliness among the elderly living alone

• Collaborating with Government of Tamil Nadu to collect data, part of the larger J-PALinstitutional partnership since 2014

• The Department of Economics and Statistics collected self-reported survey data; medical staff at theDirectorate of Public Health collected health measurements;

• J-PAL provided technical guidance and research support; principal investigators designed and pilotedproject and interventions



A representative sample of TN’s elderly

• Census: 61,954 households (HHs) from 5 state-representative districts, further stratified intodistrict-representative Primary Sampling Units, were surveyed to identify eligible households (thosewith at least one elderly)

• Panel sample: 5,000 HHs with at least one elderly member (55 years or over)• Subgroups of interest

• Elderly living alone (ELA), i.e., single-member HHs (1530)• Elderly potentially eligible for but not receiving Old Age Pensions (OAP) (1279)

• Stratified random sampling, by village/town, identified a panel sample with three subgroups fromthe census listing: a random sample, the ELA, and the OAP



Panel Survey Waves and COVID-19 Update

• Baseline survey: January-June 2019. Future waves: 2021, 2023, 2025.• Economic well-being (consumption, income, food security)• Health (mobility, diagnosed diseases, health-seeking behavior and utilization)• Mental health (depression, loneliness)• Social interaction (community activities, family)• Health measures (blood pressure, diabetes, mobility)

• COVID update: Two short phone surveys. April and July 2020.• Awareness of symptoms/lockdown, prevalence of symptoms• Coping: access to the government’s COVID welfare measures, food security, economic situation,

physical and mental health.



The Elderly Living Alone (ELA) are predominantly female

0

5

10

15

Perc

ent o

f Eld

erly

that

are

ELA

55 - 59 60 - 64 65 - 69 70 - 74 75 - 79 80+Age

Female and ELA Male and ELA

• TN census exercise with 61,954 households

• 13% of households with elderly aresingle-member households, i.e. elderly livingalone.

• 87% of the ELA are female; written anotherway, 15% of elderly females are ELA.



Poverty is very high among the elderly

Extreme poverty:38 Rs or $1.9/day

Low-middleincome poverty:

64 Rs or $3.2/day

0

5

10

15

Perc

ent

0 100 200 300Daily per-capita household expenses (Rs.)

Elderly Living with Others

Extreme poverty:38 Rs or $1.9/day

Low-middleincome poverty:

64 Rs or $3.2/day

0

5

10

15

Perc

ent

0 100 200 300Daily per-capita household expenses (Rs.)

Elderly Living Alone

Notes: Extreme and low-middle income poverty lines are taken from the World Bank. USD coversions are PPP.

• 25% of the elderly livingwith others fall under theextreme poverty line, $1.90per person per day.

• 28% of the elderly livingalone are below the extremepoverty line

• Compared to 12% of thestate as a whole.



Self-described financial situations are worse for the ELA

0

10

20

30

Perc

ent

2 4 6 8 10Self-rated financial health (1-10)

Elderly Living with Others

0

10

20

30

Perc

ent

2 4 6 8 10Self-rated financial health (1-10)

Elderly Living Alone • The average elderly householdrated their financial situation as“difficult,” a 3 on a 1-10 scale.

• 1 reflects an extremely difficultfinancial situation; 10 reflectsextremely comfortable.

• Despite an only slightly higherELA poverty rate, nearly 60% ofthe ELA rated their financialsituation as “extremely or verydifficult.”



Objective disease prevalence is high yet not self-reported

0 20 40 60Percent of all elderly

Hypertension

Diabetes

Cataract

Hearing loss

Kidney disease

Lung or heart disease

Arthritis

Objectively measured prevalence Self-reported prevalence

• Very high disease prevalence among theelderly when medically examined

• Awareness gap: much lower self-reportedprevalence of most diseases.



Disease prevalence increases in age

1.4

1.6

1.8

2

2.2

2.4

55 60 65 70 75 80 85Age

Avg. no. of common diseases - Males

1.4

1.6

1.8

2

2.2

2.4

55 60 65 70 75 80 85Age

Avg. no. of common diseases - Females

Non-ELA ELA

• Of the seven commondiseases listed above,prevalence is increasing inage.

• Particularly steep increaseamong males, as femaleshave higher diseaseprevalence early on,especially the female ELA



Significant functional impairment, especially for the ELA

0 10 20 30 40 50Percent with severe difficulty completing

Climbing a flight of stairsWalking long distances (1km)

Stooping or kneelingStanding for long periods

Carrying thingsLearning a new task

Standing up from sittingUsing transportation

Getting up from lying downSitting for long periods

Walking 100 metersGetting out of home

Joining community activitiesConcentrating for 10 minutesLifting arms above shouldersTending to household dutiesPicking items up with fingers

Getting to and using toiletEating

Washing whole bodyGetting dressed

Moving around at home

Elderly Living with Others Elderly Living Alone

• ELA considerably more likely tohave difficulty completingvarious activities of daily living;

• Disparity grows with activitydifficulty.



Functional impairment increases steeply with age

1

2

3

4

5

6

7

55 60 65 70 75 80 85Age

No. of deficient ADLs - Males

1

2

3

4

5

6

7

55 60 65 70 75 80 85Age

No. of deficient ADLs - Females

Non-ELA ELA

• “Functional impairment” is thenumber of ADLs (of 22 listedpreviously) which respondents haveat least severe difficulty completing.

• Females and the female ELA aremore likely to be functionallyimpaired.

• Functional impairment growssharply with age.



Cognitive impairment is very high for females

20

40

60

80

55 60 65 70 75 80 85Age

Pct. with cognitive impairment - Males

20

40

60

80

55 60 65 70 75 80 85Age

Pct. with cognitive impairment - Females

Non-ELA ELA

• High prevalence of cognitiveimpairment, as measured by theMini Mental State Examination

• Higher for the ELA and far higherfor females, across ages.

• 80% of females over 80 demonstratemild or severe cognitive impairment.



The ELA are far more likely to demonstrate depression

UK Elderly over 65

ELA over 60 in Shanghai

Elderly Prisoners over 60 in the UK

20

30

40

50

60

55 60 65 70 75 80 85Age

Pct. suggestive of depression - Males

UK Elderly over 65

ELA over 60 in Shanghai

Elderly Prisoners over 60 in the UK

20

30

40

50

60

55 60 65 70 75 80 85Age

Pct. suggestive of depression Females

Non-ELA ELA

• High absolute rates ofdepression, as has commonlybeen seen for the elderly.

• More females and many moreELA show symptoms ofdepression depression.

Notes: Depression scores from 15-item Geriatric Depression Scale: scores above 5 (out of 15) are suggestive of depression and scores above 9 are almost always indicative of depression. Values forelderly prisoners from O’Hara et al. 2016, ELA in Shanghai from Chen and While 2018, UK elderly from D’Ath et al. 1994.



We measure loneliness in two ways

• Directly-reported loneliness: “Do you often feel lonely?”

• UCLA Loneliness Scale:• Short-form (four-item) version;

• “I feel in tune with the people around me”• “No one really knows me well”• “I can find companionship when I want it”• “People are around me but not with me”

• Responses on 3-point Likert scale. Items are reverse or forward-scored, depending on direction.• Not validated in the Indian context



Among males, the ELA are especially lonely

10

30

50

70

90

55 60 65 70 75 80Age

Pct. Who Often Feel Lonely - Males

10

30

50

70

90

55 60 65 70 75 80Age

Pct. Lonely from UCLA Scale - Males

Male living with others Male living alone

• Loneliness among males differssignificantly between directreports and that from 4-itemUCLA scale, but considerable rateof loneliness persists.

• Loneliness among the male ELA isextremely high, around 60%.



Among females as well, the ELA demonstrate outsized loneliness

20

30

40

50

60

70

55 60 65 70 75 80Age

Pct. Who Often Feel Lonely - Females

20

30

40

50

60

70

55 60 65 70 75 80Age

Pct. Lonely from UCLA Scale - Females

Female living with others Female living alone

• Among females, the high rate ofloneliness roughly doubled for thefemale ELA.

• Mirrors male loneliness, as doesthe discrepancy between directreports and UCLA loneliness scale.



Social interaction with individuals and community

0 20 40 60 80 100Percent of Respondents

Has someone to call in emergency

Has someone to borrow money from

Talks on phone at least weekly

Can make phone call

Was visited in last week

Visited someone in last week

Had conversation in last day

Interaction with Individuals0 20 40 60 80 100

Percent of Respondents

Feels respected in community

Trusts neighbors considerably

Currently works

Attends cultural events

Meets with community leaders

Community Interaction


• Elderly report having sourcesof support in times of troubleand feel respected in theircommunity.

• Little in-person and phoneinteraction, as well as lowcommunity engagement.

• For ELA, all forms ofcommunity interaction slightlyless likely, as are phone useand having sources ofemergency support.



The ELA own far fewer assets enabling connection

(1) (2)Living with others Living alone

mean mean

Bicycle 30 3Scooter 47 2Car 4 1Phone 84 36Computer 5 1Internet connection 11 1

Observations 4760 1534

Table: Asset ownership (percentage who own each asset)

• ELA far less likely to ownassets enabling both in-personand remote connection

• Raises concerns about reachingthe ELA in times of crisis



Two waves of COVID-19 phone surveys

• Two waves of phone surveys were planned to understand the effects of COVID-19 and thelockdown on the elderly, specifically:

• Access to government COVID welfare measures, food security, economic situation, physical andmental health.

• The role of state pensions and welfare measures in alleviating the pandemic’s effects over time.

• Target sample: 4,929 elderly with phone access, from the 6,294 baseline sample.



Receipt of government schemes

0

20

40

60

80

Free rations (April) Cash transfer Free rations (July)

Percent of card holders receiving scheme


Notes: During Wave 1 (April), the ELA were surveyed one week earlier than all others.

• The state of Tamil Nadu launched two schemes forration card holders

(1) Free rations from April to July, doubling therice entitlement

(2) Cash transfer of Rs. 1000 (about USD 14)per ration card for April and May

• The schemes reached most, although not all,intended recipients.



Health challenges and food insecurity at the beginning of the lockdown

0 10 20 30 40 50Percent

Has no resources for food next week

More worried than usual about having food

Had to skip meals

Needed medication but could not buy

Needed to visit doctor but could not

Living with Others April Living Alone April

Notes: During Round 1, the ELA were surveyed one week earlier than all others.

• At the start of the lockdown, theelderly faced significant health andfood-security challenges.

• Nearly 50% reported not havingenough resources for food in thenext week.



Health challenges and food insecurity later in the lockdown

0 20 40 60Percent


Had to skip meals



Living with Others April Living Alone AprilLiving with Others July Living Alone July

• The number of elderly unable tobuy medicine or see a doctor sinceApril has fallen.

• Food insecurity, however, has risenand is high, particularly among theelderly living alone.



Among the pension-eligible, recipients and non-recipients are equally well-off in April

0 20 40 60Percent



Had to skip meals


Assigned OAP Control April Assigned OAP Treatment April

• Not all eligible for the Old Age pensionreceive it.

• In the census exercise, a group of 1,124individuals were identified as OAPeligible but not receiving it. Theseindividuals were randomized into atreatment group or a control group forreceiving the pension. 42% of thetreatment group, compared to 8% ofthe control group, now receive thepension.

• Those in the treatment and controlgroups were of equal financial stabilityin April, when pension payouts weredelayed.



Those randomized to receive the pension fare better as the lockdown progresses (July)

0 20 40 60 80Percent



Had to skip meals


Assigned OAP Control April Assigned OAP Treatment AprilAssigned OAP Control July Assigned OAP Treatment July

• By the July survey, Old Agepension payments had resumed fora number of weeks.

• The treatment group, relative tocontrol, was now much less likelyto report having not having enoughresources for food in the next week.



Most elderly with COVID symptoms did nothing or self-medicated

0 10 20 30 40Actions taken by elderly with COVID-19 symptoms (%)

Self-medicated

Did nothing

Saw government health provider

Saw private doctor

Called government helpline• 1.4% of the elderly reported having COVID-19

symptoms during the April phone surveys.

• Of the elderly reporting symptoms, nearly 80% didnothing or self-medicated.

• No one reported calling the government helpline inApril; only 0.1% reported having called thegovernment helpline in July.



COVID-19 awareness

0 20 40 60 80 100Percent

Knew COVID-19 precautions

Knew COVID-19 symptoms

Knew how COVID-19 spread



• Awareness of COVID-19 spread,symptoms, and precautions ismoderate.

• Awareness has grown among the elderlyliving alone, presumably as informationpercolates through networks.



Social connection was very low early lockdown (April)

0 10 20 30 40 50Percent

Had regular phone calls in last week

Had regular visits in last week



• At the beginning of the lockdown, socialinteraction was low for the elderly, both in-personand remote.

• Early on, the ELA were less likely to have regularphone calls.



Social connection increased mildly throughout lockdown (July)

0 20 40 60Percent

Had regular phone calls in last week

Had regular visits in last week


• By July, both regular in-person visits and calls hadpicked up moderately for the elderly living aloneand elderly living with others.



The ELA demonstrate outsized loneliness and depression early lockdown (April)

0 5 10 15 20Percent

Often felt lonely

Felt their situation was hopeless

Afraid bad thing would happen to them



• At the beginning of the lockdown, the elderlyliving alone were about twice as likely toindicate feelings of loneliness and depression.

• These indicators reflect lower loneliness anddepression than those at the baseline survey.This may be due to nature of the COVIDsurvey - shorter and via the phone - beingless conducive to rapport-building betweensurveyors and respondents.



Loneliness and depression spike significantly during the lockdown (July)

0 10 20 30Percent

Often felt lonely

Felt their situation was hopeless

Afraid bad thing would happen to them


• As the lockdown progressed, all elderlybecame far more likely to expressfeelings of loneliness and depression.

• This was especially true for the ELA,32% of whom reported often feelinglonely, when asked in July.



Improving mental health, food security, and health access: background

• We were on the cusp of starting an in-person therapy and group counseling intervention whenCOVID-19 crisis began.

• But, it is obviously no longer safe to have older people meet together and to have young peopletravel from village to village to spend extended periods of time with them.

• The elderly are most vulnerable to COVID-19, and those living alone are completely left alone fromany support system: starkly more likely to report being lonely (32% vs 15%) and having to skipmeals in last week (20% vs 15%)

• Suggestive indication that pension receipt mitigates food insecurity and health access.



Improving mental health, food security, and health access: proposed intervention

Cross-randomized RCT with a therapy and a cash component for 1,530 participants across 5 districts

• Therapy: Delivered over the phone for 6 weeks (meticulous IRB-approved protocol to ensure safetyof the elderly and interviewer)

• Counseling to address depression• Companionship to address loneliness• Self-efficacy and problem-solving for addressing financial and/or health concerns• Follows non-specialist therapy approach that was found highly effective in low-resource settings

(Singla et al. 2017)

• Cash transfer: One-time, Rs. 1000, equivalent one-time government COVID-19 cash transfer forration-card holders

• Enable elderly to meet immediate needs: health, food• Delivered after therapy sessions• Considerable evidence that cash and other transfers improve mental health (Ridley et al. 2020)



Main intervention outcomes

• Depression: geriatric depression scale

• Food security: Skipping meals in the last week or month

• Mobility and health (access and adherence)

• Social interaction with family members and other community members



Conclusion

• The elderly living alone are not assumed to exist in India; they constitute a blind spot in economicpolicy-making.

• The COVID-19 crisis has given new urgency to monitoring what is happening to the elderly andELA and helping them, as they are particularly vulnerable both in terms of their health and thefinancial situations.

• At the same time, COVID-19 makes any intervention for this group more difficult to implement; itis a unique opportunity to see what might work.


The Tamil Nadu PanelElderly Well-Being at Baseline (2019)COVID-19 Data CollectionWell-Being in the COVID-19 LockdownCounseling Intervention

How Did the Elderly in Tamil Nadu Weather the COVID-19 ......Compared to 12% of the state as a...

Documents

Transcript of How Did the Elderly in Tamil Nadu Weather the COVID-19 ......Compared to 12% of the state as a...