Corrigendum to “Digital exclusion and functional dependence in older people: findings from five longitudinal cohort studies” eClinical Medicine 54 (2022) 101708

In the original published version of this article, the author has identified some errors in the classification of dependency. They have made the corrections and re-analysis the data, and the results and conclusions continued to be consistent. The code in the published version does not accurately represent the definition of functional dependencies (Table 1). Specifically, "low dependency" is defined as "difficulty in bathing, managing money, shopping for groceries, using the telephone, or cleaning the house, as well as no difficulty in any of the items defined as medium and high dependency". Inadvertently, the analysis failed to exclude individuals who did not report difficulties with the items covered by medium and high dependency criteria. Also, there is a discrepancy in the definition of "Medium dependency." The analysis did not account for participants who did not display difficulties in items of high dependency. As a result, the proportion of high dependency cases is currently lower than anticipated.Table 1Functional dependency by interval-of-need dependency categorization.CategoriesDefinitionHigh dependencyDifficulty in eating, dressing, getting in/out of bed, using the toilet, or walkingMedium dependencyDifficulty in preparing hot meals or taking medications, and no-difficulty in the items defined in high dependencyLow dependencyDifficulty in bathing, managing money, shopping for groceries, using the telephone, or cleaning the house, and no-difficulty in the items defined in medium and high dependencyIndependentNo-difficulty in the items above Open table in a new tab Revised Table 2Descriptive statistics in HRS, ELSA, SHARE, CHARLS, and MHAS.HRS (N = 49,583)ELSA (N = 27,338)SHARE (N = 96,184)CHARLS (N = 23,342)MHASaFor MHAS, the question on retirement was unavailable, so labour force status was recoded into currently working and currently not working. (N = 26,968)Age, median (Q1–Q3)72 (65–78)69 (64–76)70 (12–77)67 (63–72)69 (65–76)Male gender20,469 (41.3)12,853 (47.0)42,160 (43.8)11,261 (48.2)11,854 (44.0)Labour force status Currently not working36,479 (73.6)21,670 (79.3)78,390 (81.5)11,214 (48.0)19,162 (71.1) Currently working without retirement10,404 (21.0)4846 (17.7)10,737 (11.2)11,095 (47.5)7806 (28.9) Currently working after retirement2700 (5.45)822 (3.01)7057 (7.34)1033 (4.43)Education Less than upper secondary9381 (18.9)8003 (29.3)42,587 (44.3)21,649 (92.7)23,949 (88.8) Upper secondary and vocational training28,989 (58.5)13,955 (51.0)33,529 (34.9)1363 (5.84)691 (2.56) Tertiary11,213 (22.6)5380 (19.7)20,068 (20.9)330 (1.41)2328 (8.63)Household wealth Low tertile14,998 (30.2)7464 (27.3)28,993 (30.1)7253 (31.1)10,825 (40.1) Medium tertile14,974 (30.2)9423 (34.5)33,028 (34.3)9231 (39.5)7322 (27.2) High tertile19,611 (39.6)10,451 (38.2)34,163 (35.5)6858 (29.4)8821 (32.7)Married or partnered29,735 (60.0)19,171 (70.1)69,294 (72.0)18,341 (78.6)17,000 (63.0)Co-residence with children7908 (15.9)224 (0.82)13,692 (14.2)9270 (39.7)18,302 (67.9)Smoking5181 (10.4)2462 (9.01)13,667 (14.2)6294 (27.0)2689 (9.97)Alcohol drinking25,587 (51.6)23,557 (86.2)47,555 (49.4)7245 (31.0)5905 (21.9)Ever had hypertension32,657 (65.9)12,854 (47.0)52,886 (55.0)9456 (40.5)16,907 (62.7)Ever had stroke4957 (10.0)1341 (4.91)7320 (7.61)1365 (5.85)1192 (4.42)Ever had cancer8946 (18.0)3621 (11.9)10,595 (11.0)378 (1.62)1101 (4.08)Depressive symptom10,314 (20.8)5057 (18.5)26,091 (27.1)8931 (38.3)8760 (32.5)Cognitive impairment2275 (4.59)1047 (3.83)5117 (5.32)327 (1.40)1652 (6.13)Digital exclusion26,394 (53.2)8316 (30.4)55,201 (57.4)22,622 (96.9)17,653 (65.5)BADL, median (Q1–Q3)0 (0–0)0 (0–0)0 (0–0)0 (0–1)0 (0–0)IADL, median (Q1–Q3)0 (0–0)0 (0–0)0 (0–0)0 (0–1)0 (0–0)Difficulty in BADL0.42 (1.07)0.33 (0.90)0.26 (0.85)0.54 (1.16)0.45 (1.10)Difficulty in IADL0.30 (0.82)0.31 (0.81)0.33 (0.95)0.72 (1.29)0.26 (0.73)Functional dependency Independent38,780 (71.2)26,429 (72.7)85,775 (77.6)20,843 (57.0)20,666 (70.7) Low dependency2674 (4.9)3002 (8.3)8886 (8.0)4201 (11.5)919 (3.2) Medium dependency1639 (3.0)577 (1.6)2208 (2.0)2191 (6.0)568 (1.9) High dependency11,412 (20.9)6351 (17.5)13,655 (12.4)9358 (25.6)7066 (24.2)BADL: basic activities of daily living; CHARLS: China Health and Retirement Longitudinal Study; ELSA: English Longitudinal Study of Ageing; HRS: Health and Retirement Study; IADL: instrumental activities of daily living; MHAS: Mexican Health and Aging Study; SHARE: Survey of Health, Ageing and Retirement in Europe.Data are N (%) for categorical variables or mean (SD) or median (Q1–Q3) for continuous variables.a For MHAS, the question on retirement was unavailable, so labour force status was recoded into currently working and currently not working. Open table in a new tab BADL: basic activities of daily living; CHARLS: China Health and Retirement Longitudinal Study; ELSA: English Longitudinal Study of Ageing; HRS: Health and Retirement Study; IADL: instrumental activities of daily living; MHAS: Mexican Health and Aging Study; SHARE: Survey of Health, Ageing and Retirement in Europe. Data are N (%) for categorical variables or mean (SD) or median (Q1–Q3) for continuous variables. Below, we take part of the Stata code for example to show the differences of generation of functional dependency variable. The variable r`wv'dependency was defined as functional dependency, where the values are categorized as follows: 0 for independence, 1 for low dependency, 2 for medium dependency, and 3 for high dependency. Variables including r`wv'dressa, r`wv'eata, r`wv'beda, r`wv'toilta, r`wv'walk100a, r`wv'medsa, r`wv'mealsa, r`wv'moneya, r`wv'shopa, r`wv'housewka, r`wv'phonea, r`wv'batha refers to items contributing to the determination of functional dependency. r`wv'adl and r`wv'iadl represent the count of difficulties in items related to activities of daily living (ADL) and instrumental activities of daily living (IADL), respectively. ADL and IADL variables include all the above variables contributing to the definition of functional dependency. Original Stata code forvalues wv = 2/4{ gen r`wv'dependency = . replace r`wv'dependency = 3 if (r`wv'dressa == 1 | r`wv'eata == 1 | r`wv'beda == 1 | r`wv'toilta == 1 | r`wv'walk100a == 1) replace r`wv'dependency = 2 if (r`wv'medsa == 1 | r`wv'mealsa == 1) replace r`wv'dependency = 1 if (r`wv'moneya == 1 | r`wv'shopa == 1 | r`wv'housewka == 1 | r`wv'phonea == 1 | r`wv'batha == 1) replace r`wv'dependency = 0 if (r`wv'adl == 0 & r`wv'iadl == 0) } Revised Stata code forvalues wv = 2/4{ gen r`wv'dependency = . replace r`wv'dependency = 0 if (r`wv'adl == 0 & r`wv'iadl == 0) replace r`wv'dependency = 1 if (r`wv'moneya == 1 | r`wv'shopa == 1 | r`wv'housewka == 1 | r`wv'phonea == 1 | r`wv'batha == 1) replace r`wv'dependency = 2 if (r`wv'medsa == 1 | r`wv'mealsa == 1) replace r`wv'dependency = 3 if (r`wv'dressa == 1 | r`wv'eata == 1 | r`wv'beda == 1 | r`wv'toilta == 1 | r`wv'walk100a == 1) } Adjustments were made to the relevant results, including one figure and two tables in the main text, as well as one table in the appendix. In general, there has been an increase in the proportion of individuals with high dependency while the proportions of other functional statuses have shown relative changes. In Table 4, we now observe the statistical significance of medium dependency in Health and Retirement Study (HRS) and high dependency in Mexican Health and Aging Study (MHAS).Revised Table 4Association between digital exclusion and functional dependency.HRSELSASHARECHARLSMHASRRR95% CIp valueRRR95% CIp valueRRR95% CIp valueRRR95% CIp valueRRR95% CIp valueIndependentRefRefRefRefRefLow dependency1.59(1.25–2.04)<0.0010.97(0.75–1.26)0.8121.06(0.84–1.33)0.6391.31(0.52–3.35)0.5670.78(0.57–1.08)0.131Medium dependency1.85(1.28–2.66)0.0011.52(0.79–2.93)0.2121.19(0.69–2.06)0.5271.14(0.38–3.42)0.8200.73(0.47–1.12)0.151High dependency1.33(1.11–1.59)0.0021.05(0.81–1.36)0.7171.09(0.87–1.37)0.4611.58(0.69–3.58)0.2770.80(0.68–0.95)0.009CHARLS: China Health and Retirement Longitudinal Study; CI: confidence interval; ELSA: English Longitudinal Study of Ageing; HRS: Health and Retirement Study; MHAS: Mexican Health and Aging Study; RRR: relative-risk ratio; SHARE: Survey of Health, Ageing and Retirement in Europe.Models were adjusted for the minimal sufficient adjustment set (MSAS) identified using a causal directed acyclic graph (DAG) including gender, age, education, labour force status, marital status, household wealth, and co-residence with children. Open table in a new tab Revised Table A1Distribution of digital exclusion, difficulties in BADL/IADL, and functional dependency of study participants by geographic regions.Digital exclusionDifficulty in BADLDifficulty in IADLFunctional dependencyIndependentLow dependencyMedium dependencyHigh dependencyHRS United States53.219.616.371.24.93.020.9ELSA England35.117.517.372.78.31.617.5SHARE Austria59.311.217.077.98.91.911.2 Germany52.111.713.680.86.51.511.3 Sweden28.68.510.983.76.21.28.9 Netherlands30.57.012.683.37.81.87.2 Spain79.913.017.275.77.32.814.2 Italy79.512.313.779.76.31.612.5 France51.314.417.774.99.81.414.0 Denmark23.88.513.181.56.92.39.4 Greece83.08.916.081.99.11.27.8 Switzerland37.96.78.787.05.71.16.2 Belgium47.116.820.970.412.02.515.0 Israel52.913.925.068.310.44.716.6 Czech Republic60.713.716.676.28.12.113.6 Poland84.617.424.172.69.71.716.0 Luxembourg49.511.015.179.46.62.111.9 Portugal82.225.521.268.26.02.723.1 Slovenia72.111.914.879.56.22.012.3 Estonia67.417.821.671.510.02.416.1 Croatia80.011.512.282.15.31.711.0CHARLS China96.925.733.157.011.56.025.6MHAS Mexico65.521.515.070.73.21.924.2BADL: basic activities of daily living; CHARLS: China Health and Retirement Longitudinal Study; ELSA: English Longitudinal Study of Ageing; HRS: Health and Retirement Study; IADL: instrumental activities of daily living; MHAS: Mexican Health and Aging Study; SHARE: Survey of Health, Ageing and Retirement in Europe. Open table in a new tab CHARLS: China Health and Retirement Longitudinal Study; CI: confidence interval; ELSA: English Longitudinal Study of Ageing; HRS: Health and Retirement Study; MHAS: Mexican Health and Aging Study; RRR: relative-risk ratio; SHARE: Survey of Health, Ageing and Retirement in Europe. Models were adjusted for the minimal sufficient adjustment set (MSAS) identified using a causal directed acyclic graph (DAG) including gender, age, education, labour force status, marital status, household wealth, and co-residence with children. BADL: basic activities of daily living; CHARLS: China Health and Retirement Longitudinal Study; ELSA: English Longitudinal Study of Ageing; HRS: Health and Retirement Study; IADL: instrumental activities of daily living; MHAS: Mexican Health and Aging Study; SHARE: Survey of Health, Ageing and Retirement in Europe. Digital exclusion and functional dependence in older people: Findings from five longitudinal cohort studiesThere is a substantial proportion of older adults who are excluded from the Internet, especially those in LMIC. Older people excluded from the Internet regardless of whether they live in HICs or LMICs are more likely to develop functional dependency. It should be made a priority to remove barriers to Internet access in order to assist older people in maintaining their independence and, consequently, to reduce the care burden associated with the ageing population worldwide. Full-Text PDF Open Access