Children’s Educational Attainment and Older Parents’ Food Insecurity: Evidence From India

Abstract

Objectives: To examine the relationship between children’s educational attainment and the risk of food insecurity among older parents and to determine if this relationship varies by the sex and residence of the children. Methods: This study utilized data from the Longitudinal Ageing Survey in India (LASI, 2017-18), comprising 25,914 individuals aged 60+ with at least one living child. Logistic regression, interaction analysis, and propensity score matching were employed to achieve the study’s objectives. Results: Having a child with higher education (undergraduate or above) was associated with a 35% reduction in the odds of parents experiencing food insecurity. This negative relationship was significant for both sons’ and daughters’ education, with daughters’ education having higher effect size. Conclusion: In a country like India, characterized by a weak welfare system but a strong collectivist family culture, children’s education may play an important role in ensuring older parents’ food security.

Keywords

food insecurity older adults children’s education LASI India

Introduction

Food insecurity, defined as the inability to consistently obtain sufficient food due to resource constraints (Nord et al., 2005), presents a significant global challenge, affecting approximately 815 million people worldwide (Food and Agriculture Organization, 2023). While it affects individuals across all age groups, older adults are disproportionately vulnerable to its impacts. A complex interplay of factors contributes to this heightened susceptibility. Financial constraints, often resulting from limited incomes and scarce economic opportunities, restrict older adults’ ability to afford food and other essentials (Saha et al., 2021). Furthermore, the loss of social support networks and social isolation, combined with declining mobility and the presence of chronic health conditions and age-related health challenges, exacerbate their vulnerability to food insecurity (Jung et al., 2001; Jih et al., 2018). As individuals age, these interrelated factors converge, substantially increasing the risk of food insecurity during the later stages of life.

The implications of food insecurity on health are profound, particularly for older individuals who are already vulnerable to various health issues. One of the most significant risks associated with food insecurity is undernutrition, as insufficient access to nutritious food can lead to deficiencies in essential nutrients necessary for proper health and functioning (Kandapan et al., 2022). This lower intake of essential micronutrients can contribute to an increased risk of a range of chronic conditions (Laraia, 2013; Leung et al., 2020). Beyond its impact on physical health, food insecurity takes a toll on mental health as well. The psychological stress and anxiety stemming from uncertain access to food can have detrimental effects on the mental health and self-assessed well-being of older adults (Smith et al., 2021). The effects of food insecurity also extend to physical functioning, as inadequate nutrition can impede daily activities and the ability to maintain an active lifestyle (Bishop & Wang, 2018; Janio & Sorkin, 2021). Ultimately, the cumulative effect of food insecurity can result in a significantly reduced quality of life among older adults (Selvamani et al., 2023).

Children’s education can significantly impact parents’ food security through several interconnected pathways. Human capital theory suggests that investing in children’s education can lead to better earnings and well-being in their adult life, thereby enhancing their ability to financially support their aging parents (Bonsang, 2007; Chou, 2010; Jiang & Kaushal, 2020; Margolis & Wright, 2017). This financial support can directly address their parents’ food security. Moreover, children with higher educational attainment often have a better understanding of their aging parents’ needs, leading to increased care and attention, including the promotion of a healthy diet (Jiang & Kaushal, 2020). Furthermore, children’s education can impact parents’ food security by improving their health. Numerous studies have demonstrated that children’s education positively influences the health and well-being of older parents (Zimmer et al., 2002, 2007; Friedman & Mare, 2010; Torssander, 2013; Yahirun et al., 2016; Lee C., 2018). This enhanced physical and mental health in older adults may lead to increased participation in economic activities, thereby improving their financial stability and, consequently, their ability to access adequate food, ensuring food security.

India’s collectivistic family culture, deeply rooted in joint family system and strong familial ties, prioritizes mutual interdependence and filial piety. Children are raised with values of respect and obedience towards elders, fostering a moral obligation to provide care and support to their aging parents (Chadda & Deb, 2013; Esteve & Liu, 2017). In such sociological context, it is reasonable to expect a natural and robust upward transfer of emotional, social, and financial support from children to their aging parents. A few studies have already reported a significant relationship between children’s education and the health and well-being of older parents in India (Ambade et al., 2021; Mustafa et al., 2023; Mustafa & Shekhar, 2024; Thoma et al., 2020). However, the specific relationship between children’s education and older parents’ food insecurity remains unexplored, both in India and elsewhere. Therefore, guided by the aforementioned pathways, this study aims to address this research gap by comprehensively investigating the relationship between offspring’s education and parental food insecurity.

Despite India’s strong family culture, patriarchal norms prevail in many areas, leading to disparities in resource allocation within households (Pande & Malhotra, 2006). This phenomenon is deeply rooted in societal attitudes towards gender roles, where sons are often viewed as future breadwinners and caregivers, while daughters are seen as temporary members of the household who will eventually leave to join their husband’s family after marriage (Dreze & Sen, 1996). As a result, sons are often prioritized for education and other investments due to their expected long-term caregiving roles, while daughters receive less investment in their education as they are expected to leave the parental home after marriage (Katiyar, 2016). This gender bias in resource allocation is further exacerbated by the fact that boys are often viewed as a source of old-age security, leading to a preference for sons over daughters (Das Gupta & Mari Bhat, 1997; Bhat & Zavier, 2003). Considering these prevalent gender biases, in this study, we aim to examine whether the examined relationship is stronger for sons’ education or daughters’ education.

Furthermore, existing literature consistently highlights the significant influence of children’s proximity and contact frequency on the well-being of older parents (Heylen et al., 2012; Tosi & Grundy, 2019). While some studies suggest that co-residence and frequent contact facilitate enhanced support and caregiving, leading to positive outcomes for parents, others report negative consequences, such as strained relationships and diminished parental autonomy (Chen et al., 2021; Guo et al., 2016; White & Rogers, 1997). The relationship between children’s proximity and parental well-being is further complicated by cultural and socioeconomic factors. For example, in some cultures, co-residence with adult children is seen as a normative and desirable arrangement, whereas in others, it may be viewed as a sign of dependency or lack of independence (Katz et al., 2005). Building upon these findings, this study seeks to investigate whether the relationship between children’s educational attainment and parental food insecurity varies as a function of children’s residential status.

Data and Methods

Data

We used data from the first wave of the Longitudinal Ageing Study in India (LASI) conducted during 2017-18 (IIPS, 2020). LASI is a large-scale, nationally representative survey designed to comprehensively assess the physical, mental, economic, and social well-being of adults aged 45 and above in India. The survey utilized a multistage stratified area probability cluster sampling design to collect data from more than 70,000 older individuals across all states and union territories of India. The survey obtained approval from the Indian Council of Medical Research (ICMR) and the institutional review board at the International Institute for Population Sciences (IIPS), Mumbai, India. For the next 25 years, LASI is planned to be conducted every two years to analyze the dynamics of population aging in India and assess the impacts of policies and interventions. Detailed information about the sample design, data collection process, and methods can be accessed from the LASI Wave-1 report (IIPS, 2020).

The total sample size of the survey was 73,396 adults aged 45 years and above, interviewed from 43,584 households. Since this study focused on older parents, individuals under the age of 60 were excluded from the analysis. Parents with all children under the age of 20 were excluded from the analysis because the primary predictor variable requires children to have attained at least an undergraduate level of education to be categorized as educated, which is unlikely to be achieved before reaching this age. Moreover, individuals with no information on children’s education were also excluded from the analyses. After excluding individuals with missing, incorrectly entered, or invalid observations on the dependent variable, the final sample size of 25,914 older individuals was achieved.

Outcome Variable

The primary outcome variable of the study was “food insecurity.” In this study, food insecurity among older parents was assessed based on four questions: (1) In the last 12 months, did you ever reduce the size of your meals or skip meals because there was not enough food at your household? (2) In the last 12 months, were you hungry but didn’t eat because there was not enough food at your household? (3) In the past 12 months did you ever not eat for a whole day because there was not enough food at your household? (4) Do you think that you have lost weight in the last 12 months because there was not enough food at your household? All the four questions were Yes/No type questions. A new binary variable indicating food insecurity was created, where 1 represented a “YES” response to any of the four questions, and 0 represented a “NO” response to all four questions. The value labels for this variable were 0 = not food insecure and 1 = food insecure.

For sensitivity analysis, we created an ordered variable of food insecurity using the four questions. The variable had five categories as follows: 0 “if the respondent answered NO to all four questions,” 1 “if the respondent answered YES to any one of the four questions,” 2 “if the respondent answered YES to any two of the four questions,” 3 “if the respondent answered YES to any three of the four questions,” and 4 “if the respondent answered YES to all the four questions.” The results of the sensitivity analysis are presented in the supplementary file (Table S1).

Exposure Variable

The primary predictor variable in our study was children’s educational attainment. Previous research has measured offspring’s education in various ways, such as education of the eldest child (Ma et al., 2021; Torssander, 2013; Zimmer et al., 2007), the proportion of children with a college degree (Lee Y., 2018; Yahirun et al., 2016), education of the highest educated child (De Neve & Harling, 2017; Sabater et al., 2020), and children's mean years of schooling (Lee C., 2018; Mustafa & Shekhar, 2024). In the current work, we adopted the education of the highest educated child approach to analyze the relationship between offspring education and parental food security. This is the most commonly used method in the literature related to children's education and parents' health & well-being. We chose this approach as it allows us to test for interactions between a child’s characteristics (e.g., sex and place of residence) and education level when predicting the outcome variable. In case of multiple children having the same (highest) level of education, the eldest child was taken into account for the analysis. For statistical analysis purposes, education of the highest educated child was categorized as a binary variable, coded as “1” if the child had an undergraduate (bachelor’s equivalent) level or higher education, and “0” if the child’s education level was below undergraduate level.

Covariates

The relationship between offspring education and parental food insecurity was controlled for a range of potential socioeconomic and demographic covariates. These variables include age of the respondent (1 = 60–69 years; 2 = 70–79 years; 3 = 80 years or older), sex (1 = male; 2 = female), religion (1 = Hindu, 2 = Muslim; 3 = Other), caste (1 = Scheduled Caste [SC]; 2 = Scheduled Tribe [ST]; 3 = Other Backward Categories [OBC]; 4 = General), education (1 = No education/less than primary; 2 = middle school completed; 3 = secondary level completed; 4 = higher education), marital status (1 = currently married; 2 = widowed; 3 = other), living arrangement (1 = alone; 2 = with spouse and children; 3 = other), place of residence (1 = rural; 2 = urban), and region (central, north, east, west, north-east, and south). The individual’s economic status was assessed based on the household’s monthly per capita expenditure (MPCE). MPCE was divided into five quintiles and coded as 1 = poorest; 2 = poorer; 3 = middle; 4 = rich; 5 = richest. Additionally, two variables related to children’s characteristics were included as covariates: sex of the highest educated child categorized as male or female, and residence of the child categorized as “lives with parents,” “lives within the city/village of parents,” and ‘lives outside the city/village of parents.”

Statistical Analyses

Univariate and bivariate analyses were performed to assess the sample characteristics and the weighted prevalence of food insecurity across various background characteristics. We conducted three sets of multivariate analyses to investigate the relationship between offspring education and parental food insecurity. In the first set of analyses, we employed three multiple logistic regression models to evaluate the odds of food insecurity among older parents in relation to the education of the highest-educated child. The first model was a bivariate model, meaning that the association between the outcome and the primary independent variable was not adjusted for any covariate. In the second model, the association was adjusted for individual-level variables (age, sex, religion, caste, education, place of residence, and marital status). Model 3 further incorporated MPCE quintile, region, living arrangement, and the sex and residence of the highest educated child.

In the second set of analyses, we utilized Propensity Score Matching (PSM) to investigate the proposed relationship. Given the cross-sectional and observational nature of the LASI data, the examined relationship was susceptible to selection bias. It is plausible that parents who face a higher risk of food insecurity may also come from lower socioeconomic status (SES) backgrounds, have pre-existing health issues, engage in risky health behaviors, and possess less cultural capital. These factors not only predispose individuals to food insecurity but may also contribute to lower educational attainment among their children. Consequently, individuals experiencing food insecurity in our sample may already possess fewer resources and human capital, resulting in lower educational attainment among their offspring. On the other hand, individuals with higher socioeconomic status and more resources may be less vulnerable to food insecurity and more inclined to invest in their children’s education. Therefore, this potential selection bias could create a spurious negative relationship between offspring education and parents’ food insecurity in our regression analysis.

The abovementioned issue can be addressed by matching individuals based on relevant covariates in the treated and untreated groups. Matches can be found easily if there are only one or two covariates, but it becomes increasingly challenging as the number of covariates increases. This dimensionality problem can be solved by reducing relevant factors into a single score (propensity score). In PSM, the propensity score represents the predicted probability of sample participants receiving the treatment. In our study, having a child with undergraduate or higher education is considered the treatment. Thus, the treatment group consists of parents with at least one child having undergraduate or higher education, while the untreated group comprises parents whose no child possesses undergraduate-level education. In order to estimate the propensity scores, the treatment variable is regressed on the selected socioeconomic and demographic variables using logistic regression, and the predicted probabilities (propensity scores) are stored. Subsequently, individuals with similar propensity scores from the treated and untreated groups can be compared. A similar methodology was used in a study that investigated the influence of children’s education on parents’ health in the United States (Dennison & Lee, 2021). Using PSM, we calculated the Average Treatment Effect (ATE), the Average Treatment Effect on the Treated (ATT), and the Average Treatment Effect on the Untreated (ATU). ATE is the expected average change in the outcome (food insecurity) if the entire population were moved from the untreated group to the treated group. ATE is estimated as follows:

ATE = E [Y 1 - Y 0] = E [Y 1] - E [Y 0]

where E[Y1] is the expected value of the outcome in the treatment group and E[Y0] is the expected value of Y in the untreated group (Pan & Bai, 2015).

ATT is the estimated average effect of treatment on those subjects who ultimately received the treatment. Using the counterfactual model, the ATT is estimated as follows:

ATT = E [Y 1 - Y 0 | Z = 1] = E [Y 1 | Z = 1] - E [Y 0 | Z = 1]

where E[Y1|Z = 1] is the food insecurity among parents who have a child with higher education and E[Y0 |Z = 1] is the expected food insecurity if the same parents did not have any child with higher education. Similarly, ATU is estimated as follows:

ATU = E [Y 1 - Y 0 | Z = 0] = E [Y 1 | Z = 0] - E [Y 0 | Z = 0]

where Z = 0 indicates the untreated subjects. All the relevant covariates were included in the model for matching. The ATE, ATT, and ATU were estimated using the psmatch2 package in STATA (Leuven & Sianesi, 2013). The balance in the matched samples was assessed using the pstest command and density balance plots. The overlap assumption was tested by visually inspecting the overlap plot. Figure 1 shows that after matching, both the treated and untreated groups were well-balanced in terms of the selected covariates. Figure 2 illustrates the overlap plot, showing a lack of excessive concentration around 0 and 1, which indicates a substantial area of common support for matching. Therefore, our analysis satisfies both the covariate balance and overlap assumption. Detailed descriptions of the PSM analysis can be accessed elsewhere (Caliendo & Kopeinig, 2008; Pan & Bai, 2015; Rosenbaum & Rubin, 1983).

In the third set of analyses, we conducted interaction analyses to explore if the relationship between the education of the highest educated child and parental food insecurity differs based on the residence and sex of the child. To explore this, we introduced two interaction terms (education_of_the_highest_educated_child##residence_of_the_child and education_of_the_highest_educated_child##sex_of_the_child) separately into model 3 of the first set of analyses. Additionally, we performed subgroup analyses as part of sensitivity analyses to test the robustness of the results found in the interaction analysis. Results of sensitivity analysis are presented in the supplementary file.

Furthermore, as additional analyses, we also investigated whether the examined relationship varies across different age, sex, economic, and education groups of the parents using interaction analyses. The findings of these additional analyses are presented in the supplementary file. All the statistical analyses were performed using STATA 16.0.

Results

Table 1 presents the background characteristics of the sample population alongside the prevalence of food insecurity by socioeconomic and demographic characteristics. The results show that approximately 59.4% of the older individuals were aged between 60 and 69 years. A significant portion of the older adults, 66.9%, had either no education or education below the primary level. Out of all respondents, 63.8% were currently married. Only 16.8% of the respondents had a child with undergraduate or higher-level education. The observed prevalence of food insecurity was found to be 8.7% overall. This prevalence was relatively higher among females (8.9%), Muslims (11.2%), SC (Scheduled Castes) individuals (12.0%), those with the lowest economic status (11.3%), rural residents (10.1%), individuals living alone (16.9%), and residents of the central region (13.1%). The prevalence of food insecurity was 9.2% among older parents with no children having higher education; conversely, it was 4.5% among those with at least one child with higher education.

Table 1.

Background Characteristics of the Sample Population and Prevalence of Food Insecurity.

	n (%)	Weighted prevalence of food insecurity (%)
Age (years)
60–69	15,391 (59.39)	8.57
70–79	7,777 (30.01)	9.12
80+	2,746 (10.60)	8.02
Sex
Male	12,058 (46.53)	8.40
Female	13,856 (53.47)	8.92
Religion
Hindu	19,265 (74.34)	8.43
Muslim	3,027 (11.68)	11.23
Others	3,621 (13.97)	7.73
Caste
Scheduled Caste (SC)	4,391 (17.5)	12.01
Scheduled Tribe (ST)	3,928 (15.87)	10.84
OBC	9,948 (39.65)	8.03
General	6,771 (29.98)	6.22
Education
No education/less than primary	17,340 (66.92)	10.35
Primary/middle	4,953 (19.11)	5.87
Secondary	1,880 (7.26)	3.75
Higher	1,740 (6.71)	3.12
MPCE quintile
Poorest	5,225 (20.16)	11.28
Poorer	5,357 (20.67)	8.40
Middle	5,297 (20.44)	7.89
Richer	5,152 (19.88)	8.16
Richest	4,883 (18.84)	7.20
Place of residence
Rural	17,446 (67.32)	10.13
Urban	8,468 (32.68)	4.99
Marital status
Currently married	16,537 (63.81)	8.21
Widowed	9,010 (34.77)	9.40
Others	367 (1.42)	11.25
Region
Central	4,186 (16.15)	13.14
North	4,407 (17.01)	4.12
East	4,839 (18.67)	9.77
West	3,104 (11.98)	6.96
North-east	2,833 (10.93)	6.53
South	6,545 (25.26)	7.21
Living arrangement
Alone	1,342 (5.18)	16.95
With spouse and children	10,970 (42.33)	7.48
Other	13,602 (52.49)	8.73
Highest educated child’s age
20–30	13,502 (52.10)	8.58
30–40	10,003 (38.60)	9.12
40+	2,409 (9.30)	7.43
Highest educated child’s sex
Male	13,732 (53.00)	9.04
Female	12,182 (47.00)	8.53
Highest educated child’s residence
Lives with parents	6,595 (25.45)	8.91
Lives within the city/village of parents	10,061 (38.82)	9.08
Lives outside the city/village of parents	9,258 (35.73)	8.08
Highest educated child’s education
Less than undergraduate	21,555 (83.18)	9.23
Undergraduate or higher	4,359 (16.82)	4.49
Total	25,914	8.68

Table 2 presents the results of the logistic regression analyses. In Model 1, which was unadjusted, the results indicate that older parents with a child having higher education had 57% lower odds of food insecurity (unadjusted odds ratio [UOR]: 0.43; 95% CI: 0.36–0.51) than those whose no child attained higher education. Upon controlling for individual-level background characteristics of the respondents in Model 2, the odds ratio increased to 0.69 (95% CI: 0.57–0.84). Model 3 further took into account additional factors, including household and geographic characteristics of the respondents, as well as characteristics of the highest-educated child. According to Model 3, older parents with an undergraduate or higher-level educated child had 35% lower odds (adjusted odds ratio [AOR]: 0.75; 95% CI: 0.61–0.91) of facing food insecurity than their counterparts. Furthermore, the study identified significant association between food insecurity and factors such as place of residence, economic status (measured by MPCE quintile), region, living arrangement, and education of the respondents. These findings were consistent with the results of sensitivity analysis (see supplementary file Table S1). Notably, the sex of the highest educated child was found to be insignificant, while the child’s residence exhibited a weak association with parents’ food insecurity.

Table 2.

Results of Logistic Regression Analysis Examining the Relationship Between Children’s Education and Parents’ Food Insecurity.

	Model 1		Model 2		Model 3
	Odds ratio	95% CI	AOR	95% CI	AOR	95% CI
Highest educated child’s education
Less than undergraduate	1.00	-	1.00	-	1.00	-
Undergraduate or higher	0.43***	0.36 - 0.51	0.69***	0.57 - 0.84	0.75***	0.61 - 0.91
Age (years)
60–69			1.00	-	1.00	-
70–79			1.02ns	0.91 - 1.14	0.99ns	0.88 - 1.11
80+			0.89ns	0.75 - 1.06	0.83ns	0.70 - 1.01
Sex
Male			1.00	-	1.00	-
Female			090**	0.81 - 0.99	0.87**	0.77 - 0.97
Religion
Hindu			1.00	-	1.00	-
Muslim			1.04ns	0.90 - 1.20	1.21**	1.04 - 1.42
Others			0.66***	0.56 - 0.77	1.06ns	0.87 - 1.28
Caste
Scheduled Caste (SC)			1.00	-	1.00	-
Scheduled Tribe (ST)			0.74***	0.62 - 0.88	0.73***	0.59 - 0.89
OBC			0.81***	0.71 - 0.92	1.47**	1.08 - 2.01
General			0.63***	0.53 - 0.73	0.73ns	0.35 - 1.52
Education
No education/less than primary			1.00	-	1.00	-
Primary/middle			0.63***	0.54 - 0.73	0.68***	0.58 - 0.78
Secondary			0.44***	0.34 - 0.58	0.52***	0.39 - 0.68
Higher			0.39***	0.28 - 0.54	0.42***	0.30 - 0.59
Place of residence
Rural			1.00	-	1.00	-
Urban			0.65***	0.58 - 0.74	0.67***	0.58 - 0.76
Marital status
Currently married			1.00	-	1.00	-
Widowed			1.22***	1.09 - 1.37	1.31ns	0.57 - 2.46
Others			1.51**	1.04 - 2.19	0.99ns	0.65 - 1.49
MPCE quintile
Poorest					1.00	-
Poorer					0.78***	0.68 - 0.90
Middle					0.67***	0.58 - 0.79
Richer					0.66***	0.56 - 0.77
Richest					0.59***	0.49 - 0.70
Region
Central					1.00	-
North					1.96***	1.62 - 2.37
East					2.32***	1.92 - 2.81
West					1.30***	1.04 - 1.63
North-east					1.08ns	0.83 - 1.40
South					1.31**	1.08 - 1.59
Living arrangement
Alone					1.00	-
With spouse and children					0.32***	0.25 - 0.40
Other					0.43***	0.36 - 0.52
Highest educated child’s sex
Male					1.00	-
Female					0.95ns	0.85 - 1.06
Highest educated child’s residence
Lives with parents					1.00	-
Lives within the city/village of parents					0.93ns	0.81 - 106
Lives outside the city/village of parents					0.80*	0.68 - 0.95

95% CI: 95% confidence interval.

AOR: adjusted odds ratio.

Significance levels: *p < .05; **p < .01; ***p < .001, ns: not significant.

Table 3 presents the results of the PSM analysis. The Average Treatment Effect (ATE) was found to be −0.04, signifying that if all respondents had a child with at least undergraduate-level education, the prevalence of food insecurity would have been reduced by 4% in the population. In other words, having an undergraduate or higher educated child was associated with a 4% decrease in the risk of experiencing food insecurity among older parents. The estimate of the average treatment effect on the untreated (ATU) suggests that among parents whose no child had undergraduate-level education (untreated), the risk of food insecurity was expected to decrease by 6.4% if they had a child with higher-level education.

Table 3.

Results of Propensity Matching Score (PSM) Analysis Examining the Relationship Between Offspring Education and Older Parents’ Food Insecurity.

	Treated	Untreated	Difference	SE	p-Value
Unmatched	0.034	0.077	−0.042	0.005	0.00
ATT	0.034	0.023	−0.011	0.004	0.007
ATU	0.076	0.012	−0.064
ATE			−0.041

SE: standard error; ATT: average treatment effect on treated; ATU: average treatment effect on untreated; ATE: average treatment effect.

Table 4 displays the results of the logistic regression analysis examining the interaction between the education of the highest educated child and their residence in predicting parents’ food insecurity. The findings indicate that having an educated child residing within the village/city of the parents was associated with an additional 32% reduction in the odds of parental food insecurity (AOR: 0.68; 95% CI: 0.63–0.99) compared to having an educated child living with the parents in the same household. However, the interaction term was not significant for cases where the child was living outside the village/city of the parents (AOR: 1.02; 95% CI: 0.65–1.58), suggesting that the association in this group was similar and not significantly different from those whose highest educated child was living with the parents. Similar results were observed in the sensitivity analysis (refer to supplementary file Table S2)

Table 4.

Interaction Between Highest Educated Child’s Education and Residence of the Child in Predicting Parents’ Food Insecurity.

	AOR	95% CI
Highest educated child’s education
(1) Less than undergraduate	1.00	-
(2) Undergraduate or higher	0.81*	0.63 - 0.99
Residence of the highest educated child
Lives with parent in the household	1.00	-
Lives within the village/city	1.20**	1.01 - 1.40
Lives outside the village/city	0.90ns	0.77 - 1.04
Highest_educated_child’s_education##residence_of_the_child
(2)#Lives with parent in the household	1.00	-
(2)#Lives within the village/city of parents	0.68*	0.40 - 0.99
(2)#Lives outside the village/city of parents	1.02ns	0.65 - 1.58

AOR: adjusted odds ratio. Significance levels: ***p-value <.001; **p-value <.01; *p-value <.05; ns: not significant. The results are adjusted for all the relevant covariates.

Table 5 displays the results of the logistic regression analysis investigating the interaction between the education of the highest educated child and their sex in predicting parents’ food insecurity. The findings reveal that daughters’ education was associated with a 19% (AOR: 0.81; 95% CI: 0.67–0.96) additional reduction in the odds of food insecurity compared to sons’ education. This suggests that daughters’ education had a stronger association with parents’ food insecurity than sons’ education. Sensitivity analysis, involving subgroup analysis, also yielded similar results (see supplementary file Table S3).

Table 5.

Interaction Between Highest Educated Child’s Education and Sex of the Child in Predicting Parents’ Food Insecurity.

	AOR	95% CI
Highest educated child’s education
(1) Less than undergraduate	1.00	-
(2) Undergraduate or higher	0.80**	0.63 - 0.99
Sex of the highest educated child
Son	1.00	-
Daughter	0.93ns	0.82 - 1.03
Highest_educated_child’s_education ##sex_of_the_child
(2)##Son	1.00	-
(2)##Daughter	0.81**	0.67 - 0.96

AOR: adjusted odds ratio. Significance values: ***p-value <.001; **p-value <.01; *p-value <.05; ns: not significant. The results are adjusted for all the relevant covariates.

Discussion

In the past few years, a growing body of research has explored the relationship between children’s educational attainment and various aspects of older parents’ well-being, including depression, functional health, cognitive function, self-rated health, and longevity (De Neve & Kawachi, 2017). This study stands as the pioneer in investigating the relationship between offspring education and older parents’ food insecurity. The study utilized data from the first wave of the LASI to investigate the proposed research questions. The findings underscore that offspring educational attainment has a negative relationship with the risk of food insecurity among their older parents. Even after addressing potential selection biases through propensity score matching and accounting for a range of socioeconomic and demographic factors, the negative association between offspring education and parental food insecurity remained robust. Furthermore, the results show that having an educated daughter was associated with a greater reduction in the odds of parental food insecurity than having educated sons. These findings highlight the long-term benefits of investing in children’s education, demonstrating that such investments in human capital can positively influence the well-being of older adults in later years.

Food security has been a serious concern in India. According to the Nutrition and Food Security Report 2018, approximately 195 million Indians were undernourished (Food and Agriculture Organization, 2018). While the Indian government has implemented numerous programs and policies to address poor nutritional status among children and adolescents, food security among older adults has largely remained overlooked. Currently, the government has only one major scheme that directly focuses on older adults’ food security—the Annapoorna Scheme. Under this initiative, destitute individuals aged 65 years and above are provided 10 kg of food grains every month free of cost. However, the eligibility criteria for this program are rather stringent, as beneficiaries must not only be destitute but also not recipients of any national or state pension scheme. While the Annapoorna Scheme appears promising on paper, the data does not show satisfactory results. According to LASI, only 12% of older adults were aware of the scheme, and merely 1.3% were benefiting from it in 2016-17 (IIPS, 2020). Additionally, it has been documented that a significant portion of India’s older population is deprived of sufficient social security or old-age pension (Goli et al., 2019). Even among those who do receive benefits from pension schemes, the amount they receive is often insufficient (Goli et al., 2019), highlighting the inadequacies of India’s social welfare system for older adults. Moreover, a majority of senior citizens in India have inadequate savings for their later life due to the low per capita income. In such circumstances, where social welfare systems are weak, and older individuals lack sufficient savings, the transfer of human capital from children to older parents becomes crucial. With the elderly population projected to experience significant growth in India in the coming years, the transfer of educational benefits from children to older parents could play a vital role in addressing emerging concerns such as food insecurity among older adults. In this context, the findings of this study stimulate meaningful discussions and underscore the significance of its results.

In this study, we also conducted interaction and subgroup analyses to investigate whether the association between children’s education and parental food insecurity varies based on children’s residence. It was expected that the education of the child living with parents in the household would have the strongest relationship with parental food security due to their geographical proximity to the parents. However, the results showed an insignificant association between offspring education and parental food insecurity for children living with parents in the household (supplementary file Table S2). Interestingly, among the three categories of child’s residence, the highest effect size was observed when the highest educated child resided within the same village or city as their parents. This finding lacks direct support in existing literature, but a plausible explanation could be linked to the employment scenario of educated youth in India. Given the notable unemployment rate, unemployed children often continue to live with their parents in the same household, which may lead to minimal resource contributions to the parents. On the other hand, employed children residing in close proximity to their parents’ city or village may have better opportunities to check on their parents’ needs and well-being regularly. While these findings suggest that children’s residence could also matter in how children’s education relates to their parents’ well-being, understanding the complex relationship between education, residence, and how they work together for the well-being of parents needs further investigation.

In the third set of analyses, we further examined whether daughters’ education is more beneficial for parental food security than sons’ education. Considering the patriarchal context, we expected that sons’ education would have a stronger association with parental food insecurity. While both sons’ and daughters’ educations were found to be significantly associated with parental food insecurity, contrary to our expectations, daughters’ education displayed a more pronounced association. One plausible explanation could be that daughters are often perceived as more adept at caring for aging parents than sons (American Sociological Association, 2014), and the parent–daughter relationship is often regarded as warmer than the parent–son relationship. Angelina Grigoryeva, in her study, concluded that “not only do daughters provide more care than sons to their older parents, but daughters’ caregiving is also more elastic with respect to their own and their parents’ attributes than is sons’ caregiving” (Grigoryeva, 2017). Perhaps the same could be the reason for the stronger association between offspring education and parental food security for daughters than sons. Similar results were reported in a recent study conducted in India, where daughters’ education was found to be more prominent for parents’ healthcare utilization than sons’ education (Ambade et al., 2021). However, it is worth noting that the effect heterogeneity by child’s sex has shown inconsistency in the existing literature. While some studies suggest that son’s education holds more significance for parents’ welfare (Liu et al., 2022), others emphasize the importance of daughters’ education (Ambade et al., 2021). Additionally, some studies have found no noticeable gender differences in this regard (Jiang & Kaushal, 2020). This variability may be attributed to regional and contextual differences, underscoring that the moderating influence of offspring’s sex on the transfer of educational benefits from children to parents is not fixed but conditional on multiple factors.

To further examine the heterogeneity of effects based on parents’ characteristics, we conducted additional interaction analysis (refer to supplementary file, Table S4, S5, S6, and S7). These analyses aimed to determine whether the relationship between offspring education and parental food insecurity is influenced by factors such as age, sex, economic status, and education of the parents. In other words, the goal was to determine if certain socioeconomic groups demonstrate a stronger association between offspring education and parents’ food security. Our analysis did not reveal any statistically significant effect heterogeneity in the relationship under investigation based on the parents’ age, sex, economic status, or education. However, this does not conclusively establish the absence of actual effect heterogeneity. Instead, it indicates that the available data does not provide sufficient evidence to support the notion that the association between offspring education and parents’ food insecurity varies across these specific socioeconomic and demographic groups. For instance, when considering the effect heterogeneity by the economic status of parents (see supplementary file Table S6), we observed that the interaction term (odds ratio) was higher for the middle and higher economic classes than for the lower class. This suggests that the strength of the negative association between offspring education and parents’ food insecurity was higher in the lower economic class. However, this relationship was not statistically significant, as indicated by the confidence interval and p-value. It is important to note that interaction terms have a higher variance than the main effect (Leon & Heo, 2009), leading to higher standard errors and, consequently, higher p-values. Although our sample size was sufficient, the number of food insecurity cases was relatively low. Therefore, the limited number of cases might have impacted the statistical power needed to detect significant interaction effects in the analysis. Considering this, we recommend further in-depth investigation in this area. Perhaps, with a larger sample size and a more balanced distribution of cases, the results may differ.

In their study, Friedman and Mare (Friedman & Mare, 2014) suggested that policies targeting one generation of the family can result in improved well-being for preceding generations, succeeding generations, and the entire family. Extending the findings of Friedman and Mare, our study provides further evidence that investing in the education of the younger generation not only benefits the youth themselves but may also have positive implications for the well-being of their older parents. Hence, this study contradicts the idea of “generational equality,” which posits that programs benefiting the older population can only be implemented at the expense of the young and vice versa (Preston, 1984). The outcomes of our study advocate for a perspective shift, emphasizing that encouraging the younger generation to pursue higher education could lead to a mutually beneficial scenario for both older and younger generations. As a practical implication of our findings, we propose that incorporating lessons on caregiving for aging parents into early educational curricula could lay a strong foundation for children to comprehend and fulfill their obligations towards their parents in the future. This could foster a culture of intergenerational care and support, ensuring that both the older and younger generations benefit from such an approach.

While this study has utilized population-based data and employed robust statistical methods, it is important to acknowledge certain limitations. The study design is cross-sectional, which means that causality cannot be inferred from the results. Although we controlled for several potential confounding factors and minimized potential selection bias using PSM, there may still be unobserved variables that could introduce bias. For instance, parental intelligence or awareness may influence both parental food security and their children’s education, which could lead to endogeneity issues in the analyses. Nevertheless, we examined the risk of endogeneity in the regression models using the Hausman test and found no evidence of it. Furthermore, the analysis does not rule out the possibility of reverse causality. It is plausible that parents who experienced food insecurity earlier in life may have lacked sufficient resources to provide education for their children, potentially influencing their educational outcomes as a result. However, this phenomenon is very unlikely to play a significant role in our analysis, as food security in early adulthood significantly differs from food security in later life. Additionally, the analysis did not take into consideration the employment status of children due to data limitations, a dimension that could be explored in future studies. Despite these limitations, this study provides further evidence of the significance of offspring educational resources for the well-being of older parents, especially in low- and middle-income countries where public support systems are often limited in scope.

Supplemental Material

Supplemental Material - Children’s Educational Attainment and Older Parents’ Food Insecurity: Evidence From India

Supplemental Material for Children’s Educational Attainment and Older Parents’ Food Insecurity: Evidence From India by Akif Mustafa and Chander Shekhar in Journal of Aging and Health.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical Statement

ORCID iD

Akif Mustafa

Data Availability Statement

Data may be obtained from a third party and are not publicly available. The study uses secondary data of LASI which are accessible on reasonable request through .

Supplemental Material

Supplemental material for this article is available online.

References

Ambade

Joe

Subramanian

S. V.

(2021). Children’s education and level of health care utilization among parents: A pooled cross-sectional study from national population-based survey in India. GeroScience, 43(5), 2497–2514. https://doi.org/10.1007/s11357-021-00415-0

American Sociological Association . (2014). Daughters provide as much elderly parent care as they can, sons do as little as possible. Science Daily. Retrieved from. https://www.sciencedaily.com/releases/2014/08/140819082912.htm

Bhat

P. N. M.

Zavier

A. J. F.

(2003). Fertility decline and gender bias in northern India. Demography, 40(4), 637–657.

Bishop

N. J.

Wang

(2018). Food insecurity, comorbidity, and mobility limitations among older U.S. adults: Findings from the health and Retirement study and health care and nutrition study. Preventive Medicine, 114(2), 180–187. https://doi.org/10.1016/j.ypmed.2018.07.001

Bonsang

(2007). How do middle-aged children allocate time and money transfers to their older parents in Europe? Empirica, 34(2), 171–188. https://doi.org/10.1007/s10663-007-9034-3

Caliendo

Kopeinig

(2008). Some practical guidance for the implementation of propensity score matching. Journal of Economic Surveys, 22(1), 31–72. https://doi.org/10.1111/j.1467-6419.2007.00527.x

Chadda

R. K.

Deb

K. S.

(2013). Indian family systems, collectivistic society and psychotherapy. Indian Journal of Psychiatry, 55(Suppl 2), S299–S309. https://doi.org/10.4103/0019-5545.105555

Chen

Shen

Ruan

(2021). The mixed blessing of living together or close by: Parent–child relationship quality and life satisfaction of older adults in China. Demographic Research, 44(2), 563–594. https://doi.org/10.4054/demres.2021.44.24

Chou

K. L.

(2010). Number of children and upstream intergenerational financial transfers: Evidence from Hong Kong. The Journals of Gerontology: Social Sciences, 65B(2), 227–235. https://doi.org/10.1093/geronb/gbp103

10.

Das Gupta

Mari Bhat

P. N.

(1997). Fertility decline and increased manifestation of sex bias in India. Population studies, 51(3), 307–315. https://doi.org/10.1080/0032472031000150076

11.

De Neve

J. W.

Harling

(2017). Offspring schooling associated with increased parental survival in rural KwaZulu-Natal, South Africa. Social Science & Medicine, 176(2), 149–157. https://doi.org/10.1016/j.socscimed.2017.01.015

12.

De Neve

J. W.

Kawachi

(2017). Spillovers between siblings and from offspring to parents are understudied: A review and future directions for research. Social Science & Medicine, 183, 56–61. https://doi.org/10.1016/j.socscimed.2017.04.010

13.

Dennison

C. R.

Lee

K. S.

(2021). Adult children's educational attainment and parent health in Mid- and later-life. The Journals of Gerontology: Social Sciences, 76(9), 1857–1869. https://doi.org/10.1093/geronb/gbab109

14.

Dreze

Sen

(1996). India economic development and social opportunity. Oxford University Press.

15.

Esteve

Liu

(2017). Family and household composition in Asia. In Zhao

Hayes

A. C.

(Eds.), Routledge Handbook of Asian Demography. Routledge. https://doi.org/10.4324/9781315148458

16.

Food and Agriculture Organization . (2018). The state of food security and nutrition in the World 2023. Available online at. https://www.fao.org/3/I9553EN/i9553en.pdf

17.

Food and Agriculture Organization . (2023). The state of food security and nutrition in the World 2023. Available online at. https://www.fao.org/documents/card/en/c/cc3017en

18.

Friedman

E. M.

Mare

(2010). Education of children and Differential mortality of parents: Do parents benefit from their children’s attainments? UCLA CCPR Population Working Papers. Available at. https://papers.ccpr.ucla.edu/index.php/pwp/article/view/PWP-CCPR-2010-011

19.

Friedman

E. M.

Mare

(2014). The schooling of offspring and the survival of parents. Demography, 51(4), 1271–1293. https://doi.org/10.1007/s13524-014-0303-z

20.

Goli

Reddy

James

K. S.

Srinivasan

(2019). Economic independence and social security among India’s elderly. Economic and Political Weekly, 54(39), 52–60.

21.

Grigoryeva

(2017). Own gender, Sibling’s gender, parent’s gender: The division of elderly parent care among adult children. American Sociological Review, 82(1), 116–146. https://doi.org/10.1177/0003122416686521

22.

Guo

M. A. N.

Liu

Mao

Chi

(2016). Parent–child relationships among older Chinese immigrants: The influence of co-residence, frequent contact, intergenerational support and sense of children's deference. Ageing and Society, 36(7), 1459–1482. https://doi.org/10.1017/s0144686x15000446

23.

Heylen

Mortelmans

Hermans

Boudiny

(2012). The intermediate effect of geographic proximity on intergenerational support: A comparison of France and Bulgaria. Demographic Research, 27(7), 455–486. https://doi.org/10.4054/demres.2012.27.17

24.

IIPS . (2020). Longitudinal ageing study in India (LASI) wave 1, 2017–18, India report. International Institute for Population Sciences (IIPS).

25.

Janio

E. A.

Sorkin

D. H.

(2021). Food insecurity and healthcare access, utilization, and quality among middle and later life adults in California. Journal of Aging and Health, 33(3-4), 171–186. https://doi.org/10.1177/0898264320967563

26.

Jiang

Kaushal

(2020). How children’s education affects caregiving: Evidence from parent’s last years of life. Economics and Human Biology, 38(7), 100875. https://doi.org/10.1016/j.ehb.2020.100875

27.

Jih

Stijacic-Cenzer

Seligman

Boscardin

Nguyen

Ritchie

(2018). Chronic disease burden predicts food insecurity among older adults. Public Health Nutrition, 21(9), 1737–1742. https://doi.org/10.1017/S1368980017004062

28.

Jung Sun

Frongillo

D. (2001). Factors associated with food insecurity among U.S. Elderly Persons: Importance of functional Impairments. The Journals of Gerontology, 56(7), S94–S99. https://doi.org/10.1093/geronb/56.2.S94

29.

Kandapan

Pradhan

(2022). Food insecurity and Malnutrition among Indian older adults: Findings from Longitudinal ageing study in India, 2017-18. Journal of Population Ageing, 3(1), 1–21. https://doi.org/10.1007/s12062-022-09378-2

30.

Katiyar

S. P.

(2016). Gender disparity in literacy in India. Social Change, 46(1), 46–69. https://doi.org/10.1177/0049085715618558

31.

Katz

Lowenstein

Phillips

Daatland

S. O.

(2005). Theorizing intergenerational family relations: Solidarity, conflict, and ambivalence in cross-national contexts. Sourcebook of family theory and research, 6(1), 393–421.

32.

Laraia

B. A.

(2013). Food insecurity and chronic disease. Advances in Nutrition, 4(2), 203–212. https://doi.org/10.3945/AN.112.003277

33.

Lee

(2018). Adult children's education and Physiological Dysregulation among older parents. The Journals of Gerontology: Social Sciences, 73(6), 1143–1154. https://doi.org/10.1093/geronb/gbx044

34.

Lee

(2018). Adult children's educational attainment and the cognitive trajectories of older parents in South Korea. Social Science & Medicine, 209(6), 76–85. https://doi.org/10.1016/j.socscimed.2018.05.026

35.

Leon

A. C.

Heo

(2009). Sample sizes required to detect interactions between two binary fixed-effects in a mixed-effects Linear regression model. Computational Statistics & Data Analysis, 53(3), 603–608. https://doi.org/10.1016/j.csda.2008.06.010

36.

Leung

C. W.

Kullgren

J. T.

Malani

P. N.

Singer

D. C.

Kirch

Solway

Wolfson

J. A.

(2020). Food insecurity is associated with multiple chronic conditions and physical health status among older US adults. Preventive Medicine Reports, 20(3), 101211. https://doi.org/10.1016/j.pmedr.2020.101211

37.

Leuven

Sianesi

(2013). PSMATCH2: Stata module to perform full Mahalanobis and propensity score matching, common support graphing, and covariate imbalance testing. Statistical Software Components, 2(3), S432001. Available online at. https://ideas.repec.org/c/boc/bocode/s432001.html#author-abstract

38.

Liu

Smith

J. P.

(2022). Adult children's education and older parents' chronic Illnesses in aging China. Demography, 59(2), 535–562. https://doi.org/10.1215/00703370-9766973

39.

Yahirun

Saenz

Sheehan

(2021). Offspring educational attainment and older parents’ cognition in Mexico. Demography, 58(1), 75–109. https://doi.org/10.1215/00703370-8931725

40.

Margolis

Wright

(2017). Older adults with three generations of Kin: Prevalence, Correlates, and transfers. The Journals of Gerontology: Series B, 72(6), 1067–1072. https://doi.org/10.1093/geronb/gbv158

41.

Mustafa

Jungari

Shekhar

(2023). Relationship between children’s educational attainment and parental depression among Indian elderly parents. Asian Population Studies.

42.

Mustafa

Shekhar

(2024). Adult children’s educational attainment and its relationship with older parents’ life satisfaction in India. Educational Gerontology, 50(9), 775–792. https://doi.org/10.1080/03601277.2024.2339597

43.

Nord

Andrews

Carlson

(2005). Household food security in the United States, 2005. United States Department of Agriculture. Available at. https://www.ers.usda.gov/

44.

Pan

Bai

(Eds.), (2015). Propensity score analysis: Fundamentals and developments. Guilford Press.

45.

Pande

Malhotra

(2006). Son preference and daughter Neglect in India: What Happens to living Girls?. ICRW. Available at. https://www.unfpa.org/sites/default/files/resource-pdf/UNFPA_Publication-39764.pdf

46.

Preston

S. H.

(1984). Children and the elderly: Divergent paths for America’s dependents. Demography, 21(4), 435–457.

47.

Rosenbaum

P. R.

Rubin

D. B.

(1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70(1), 41–55. https://doi.org/10.2307/2335942

48.

Sabater

Graham

Marshall

(2020). Does having highly educated adult children reduce mortality risks for parents with low educational attainment in Europe? Ageing and Society, 40(12), 2635–2670. https://doi.org/10.1017/S0144686X19000795

49.

Saha

Behnke

Oldewage-Theron

Mubtasim

Miller

(2021). Prevalence and factors associated with food insecurity among older adults in Sub-Saharan Africa: A Systematic review. Journal of Nutrition in Gerontology and Geriatics, 40(4), 171–196. https://doi.org/10.1080/21551197.2021.1988027

50.

Selvamani

Arokiasamy

Chaudhary

(2023). Association between food insecurity and quality of life among older adults (60+) in six low and middle-income countries. Archives of Gerontology and Geriatrics, 114(4), 105079. https://doi.org/10.1016/j.archger.2023.105079

51.

Smith

Il Shin

McDermott

Jacob

Barnett

López-Sánchez

G. F.

Veronese

Yang

Soysal

Grabovac

Koyanagi

(2021). Association between food insecurity and depression among older adults from low- and middle-income countries. Depression and Anxiety, 38(4), 439–446. https://doi.org/10.1002/da.23147

52.

Thoma

Sudharsanan

Karlsson

Joe

Subramanian

S. V.

De Neve

J. W.

(2020). Children’s education and parental old-age health: Evidence from a population-based, nationally representative study in India. Population Studies, 75(1), 51–66. https://doi.org/10.1080/00324728.2020.1775873

53.

Torssander

(2013). From child to parent? The significance of children’s education for their parents’ longevity. Demography, 50(2), 637–659. https://doi.org/10.1007/s13524-012-0155-3

54.

Tosi

Grundy

(2019). Intergenerational contacts and depressive symptoms among older parents in Eastern Europe. Aging & Mental Health, 23(6), 686–692. https://doi.org/10.1080/13607863.2018.1442412

55.

White

L. K.

Rogers

S. J.

(1997). Strong support but uneasy relationships: Coresidence and adult children's relationships with their parents. Journal of Marriage and Family, 59(1), 62–76. https://doi.org/10.2307/353662

56.

Yahirun

J. J.

Sheehan

C. M.

Hayward

M. D.

(2016). Adult children’s education and parents’ functional limitations in Mexico. Research on Aging, 38(3), 322–345. https://doi.org/10.1177/0164027515620240

57.

Zimmer

Hermalin

A. I.

Lin

H. S.

(2002). Whose education Counts? The added impact of adult-child education on physical functioning of older Taiwanese. The Journals of Gerontology, 57(1), S23–S32. https://doi.org/10.1093/geronb/57.1.s23

58.

Zimmer

Martin

L. G.

Ofstedal

M. B.

Chuang

Y. L.

(2007). Education of adult children and mortality of their elderly parents in Taiwan. Demography, 44(2), 289–305. https://doi.org/10.1353/dem.2007.0020

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.20 MB