Social isolation as a risk factor for language disorders in 4-year-old children

INTRODUCTION

Social distancing has been implemented as a public health policy during specific events of national or international scope as a protective measure against natural phenomena, such as epidemics. Specifically, in Mexico, this policy was in effect from 2020 to 2023. The implemented measures included the use of face masks, restrictions on interpersonal contact outside the household, and the closure of schools and public gathering places. On the other hand, social isolation is considered an environmental factor that may have a negative impact on language development by limiting opportunities for verbal interaction, social play, and exposure to diverse linguistic models.^1-4

Language development during childhood is a dynamic and multifactorial process. It involves the interplay of neurobiological, psychosocial, and environmental factors in the child. Language acquisition and consolidation involve the maturation of the central nervous system, as well as exposure to varied linguistic input through social interactions with caregivers and peers. Research has shown that frequent, high-quality social interaction is a key factor in early language development, as it facilitates exposure to verbal language and responsive feedback, and promotes speech and pragmatic skills.^5-9

The Child Development Evaluation test (EDI, for the Spanish acronym of Evaluación del Desarrollo Infantil) was developed and validated in Mexico for screening of developmental delays in children aged 1 month to 5 years and 11 months. The assessment is structured in five axes: neurodevelopment (with assessment of skills in five key areas: language, fine motor, gross motor, social, cognitive), biological risk factors, warning signs, alarm signs, and neurologic examination. It allows classification of children into three categories (normal development, developmental delay, and risk of developmental delay), facilitating early intervention and appropriate referral to specialized care. However, we must underscore that it is not a diagnostic tool. The early identification of neurodevelopmental disorders in childhood is key in family medicine and primary care practice due to its immediate and long-term impact on the health, educational and social integration of individuals.¹⁰

There is evidence that greater degrees of social isolation are associated with decreased receptive and expressive language skills in preschoolers. In addition, there are permanent changes in neurodevelopment, which is cause for concern from the perspective of public health. Language is not only essential for communication but is also associated with academic performance, emotional regulation, and social integration throughout the lifespan. The prevalence of language disorders in preschoolers is 15%.^11-16

In this context, we conducted a study with the aim of determining whether social isolation is a risk factor for language disorders in children aged 4 years.

MATERIAL AND METHODS

Natural quasi-experimental study in children aged 4 years affiliated to a social services institution in the city of Querétaro, Mexico, from year 2018 to year 2025.

The exposure consisted of official social distancing measures implemented as public health policy from 2020 to 2023. These measures included home confinement/lockdown orders; consistent mask-wearing both inside and outside the home; suspension of group activities; suspension of in-person schooling at the preschool, elementary, middle school, and high school levels; and suspension of recreational activities, family gatherings, social gatherings, and religious group meetings; restrictions to social activities, access to public parks, theaters, shopping centers and restaurants; and, in within the household, a distance of 1.5 m between individuals.

The control group consisted of children aged 4 years prior to the start of the lockdown assessed with the EDI test. The exposed group consisted of children born in 2020-2021 who were exposed to social distancing measures for 3 years and who, at the time of the survey, were 4 years old. The study included children aged 4 years from the same community whose legal guardians consented to their participation and who willing to complete the EDI test. We excluded children from other communities, with a prior diagnosis of a neurodevelopmental disorder (autism; attention-deficit/hyperactivity disorder [ADHD] including hyperactive, inattentive, or combined subtypes), Down syndrome, sensorineural hearing loss, or intellectual disability, with a personal history of traumatic brain injury, perinatal asphyxia o prematurity, or with a family history of language disorders.

We calculated the minimum sample size with the quasi-experimental design formula for a 95% level of confidence for the rejection region (Z_α = 1.64), and a power of 80% (Z_β = 0.84), with an expected proportion of language disorders of 15% in the unexposed group (assessed before the social distancing period) (p₀ = 0.15) and of 35% (p₁ = 0.35) in the exposed group (assessed after the social distancing period). The necessary sample size was estimated at 57 children per group, and the actual sample included 87 children in the unexposed group and 62 in the exposed group.

The control (unexposed) group included all children who had been evaluated with the EDI test for whom records were available in the database. The exposed group was selected by convenience, including children consecutively as they visited the clinic.

The child-related variables studied included the sex, enrolment in early education, and whether or not the child had siblings. We also took into account maternal age.

To screen for language impairment, we used the language domain of the EDI test specific to Group 13 (37 to 48 months and 29 days of age). The EDI was administered by a single researcher specifically trained in its application.

The statistical analysis included calculation of percentages and means, the c² test, calculation of relative risks with their corresponding confidence intervals, the t-test, multiple logistic regression, and the calculation of the probability of the event.

The protocol was registered with the Research and Ethics Committee of the social security institution under institutional registration number R-2024-2201-219. We obtained written informed consent in every case, and the study was conducted in adherence to the principles of the Declaration of Helsinki.

RESULTS

In the group exposed to social isolation conditions, 62.9% of children were female, compared to y 70.1% in the control group (p = 0.356). Groups were also comparable in terms of enrolment in day care or preschool (p = 0.374) and being the first-born child (p = 0.511), as there were no statistically significant differences in these variables (Table 1).

Table 1. Social isolation and characteristics of children aged 4 years
Social isolation	Characteristic		χ2	P	RR	95 CI
						Lower	Upper
Only child
	Yes	No
Yes (n = 62)*	17.7	82.3	0.43	0.511	1.04	0.90	1.20
No (n = 87)*	13.8	86.2
Schooling
	Kindergarten Year 1	Kindergarten Year 2
Yes (n = 62)*	3.2	96.8	0.79	0.374	1.02	1.07	1.97
No (n = 87)*	1.1	98.9
Sex
	Female	Male
Yes (n = 62)*	62.9	37.1	0.85	0.356	1.11	0.88	1.41
No (n = 87)*	70.1	29.9
*Expressed as percentage. CI: confidence interval; RR: relative risk.

Maternal age was slightly older (p <0.001) in the social isolation group: 29.52 years compared to 24.09 years in the group without social isolation (Table 2).

Table 2. Social isolation in children aged 4 years and maternal age
Social isolation	Maternal age (years)		Difference	t	P
	Mean	Standard deviation
Yes (n = 62)	29.52	5.79	5.42	6.92	<0.001
No (n = 87)	24.09	3.75

In the group with social isolation, language impairment was detected in 97.0% of children, compared to a prevalence of 17.2% in the group without social isolation, a difference that was statistically significant (p <0.001) (Table 3).

Table 3. Social isolation and language delays in children aged 4 years
Social isolation	Language		χ2	P	RR	95 CI
	Delay	Normal				Lower	Upper
Yes (n = 62)*	79.0	21.0	56.40	<0.001	3.94	1.41	6.46
No (n = 87)*	17.2	82.8
CI: confidence interval; RR: relative risk. *Expressed as percentage.

The multiple regression equation in the model explaining language impairment included social isolation and maternal age y = −4.027+2.491 (social isolation) +0.100 (maternal age). Table 4 presents these data. For maternal age is 29 years, the probability of language delay in children aged 4 years is 24.5% in absence of social isolation, compared to 79.6% when there is social isolation. For maternal age 35 years, the probability of language impairment is 37.1% in absence of social isolation and increases to 87.7% with social isolation. Table 5 shows the probabilities calculated for different scenarios.

Table 4. Multiple logistic regression model to explain language delay in children aged 4 years
Cox and Snell R2	Nagelkerke R2	χ2	P
0.353	0.474	64.89	<0.001
Variable	Coefficient	Statistic	P
Social isolation	2.491	30.91	<0.001
Maternal age	0.100	4.72	0.030
Constant	-4.027

Table 5. Probability of language delay in children aged 4 years based on social isolation and maternal age
Maternal age	Social isolation
	No	Yes
	Probability of language delay*	Probability of language delay*
20	11.6	61..4
21	12.7	63.7
22	13.9	66.0
23	15.1	68.2
24	16.4	70.3
25	17.8	72.4
26	19.4	74.3
27	21.0	76.2
28	22.7	78.0
29	24.5	79,6
30	26.4	81.2
31	28.4	82.7
32	30.4	84.1
33	32.6	85.4
34	34.8	86.6
35	37.1	87.7
36	39.5	88.7
37	41.9	89.7
38	44.3	90.6
39	46.8	91.4
40	49.3	92.2
*Values expressed as percentages.

DISCUSSION

Social distancing measures implemented among other unplanned and temporary public health policy interventions in the wake of the COVID pandemic allowed investigation of the effects of a decrease in social interaction on different areas of child development, without random allocation to social isolation conditions. The study of communication barriers is important to contribute evidence for the primary care setting and an area in which there is still limited data on the impact of social isolation on language development.

In this context, we conducted a quasi-experimental study, a design characterized by the assessment of exposure occurring independently of the researcher’s will and determined by an exogenous agent, such as a law, a policy in a given sector, or a natural phenomenon, and which allows comparison of exposed and unexposed groups. It can be used to predict the effects of exposure and evaluate associations in different populations or subsets of the population, and it offers an adequate external validity. This approach is appropriate for the analysis of social determinants of health—in this case, of child development.^17,18

The assessment of developmental language risk in children, in addition to the customary clinical and demographic variables, should also include social determinants and the characteristics of the close environment, particularly the time the child spends socializing, the quality of stimulation by the family, and screen time, aspects that, not being covered in the study, could constitute a limitation.¹⁹

In our sample, maternal age was identified as a risk factor for language delay in preschoolers: the older the mother, the higher the probability of language delay (p <0.001). These findings are significant because, at present, there is an increasing trend in the age at which women are having their first child, with the average age for a first pregnancy ranging between 30 and 31 years, which is consistent with the average maternal age in our study, reinforcing the validity of its findings. The association between child development and maternal age points at relevant social determinants and behavioral factors, as older mothers tend to have different experiences with various parenting styles. Provided the child is not their first-born, these mothers tend to have support networks and more experience with language-stimulation practices; however, they may also have a heavier burden of work-related responsibilities, which would reduce the available time for face-to-face interactions with the child. There is evidence of a nonlinear association between maternal age and language development outcomes, with findings indicating that children of mothers of intermediate age tend to exhibit better language profiles—a condition attributed to a combination of experience, emotional stability, and better socioeconomic conditions.^20-22

The social context of young mothers live includes stronger support networks and greater opportunities for family and community interaction, which contribute to lower levels of social isolation among their children; in contrast, children of older mothers—who may have more demanding schedules, a heavier family burden, or be less involved in the community—may not be exposed to social situations as frequently.²³

The results indicate that social isolation is a risk factor for language disorders in preschool-aged children, which is consistent with the scientific evidence. Specifically, social distancing, school closures, and limited face-to-face interaction over a three-year period were associated with language delay stemming from the failure to form neural networks, lack of exposure to diverse linguistic models, and lack of natural pragmatic feedback—due to the inability to actively imitate facial expressions and gestures, a situation exacerbated by the constant use of masks.^24-26

Language development in early childhood relies on key neurobiological processes: myelination and language networks in frontotemporal regions (specifically, Broca's area and Wernicke's area), experience-dependent synaptic strengthening mediated by rich and repeated environmental input, and heightened neural plasticity during the sensitive period from ages 2 to 5 years. Similarly, social factors have been identified as the most influential during this stage. Social stimulation provides structured language input (including intonation, conversational turn-taking, and vocabulary expansion) that strengthens language networks. Thus, social isolation limits rich conversational interactions, resulting in reduced linguistic input, diminished activation of neural circuits involved in speech processing, and reduced consolidation of grammatical and pragmatic rules.^27-33

This multiple logistic regression analysis provides a more comprehensive understanding of language impairment and suggests that social isolation and older maternal age may be associated with language impairment in preschool-aged children.

CONCLUSIONS

Social isolation and older maternal age may be risk factors for language delay in children aged 4 years.

CONFLICTS OF INTEREST

The authors have no conflicts of interest to declare in relation to the preparation and publication of this article.

AUTHORSHIP

All authors contributed equally to the development of the published manuscript.

ABREVIATIONS

EDI: Evaluación del Desarrollo Infantil (Child Development Evaluation test).

REFERENCES

1. Cabrera NJ, Moon UJ, Fagan J, West J, Aldoney D. Cognitive stimulation at home and in child care and children’s preacademic skills in two-parent families. Child Dev. 2020;91(5):1709-17. https://doi.org/10.1111/cdev.13380
2. Donnellan E, Bannard C, McGillion ML, Slocombe KE, Matthews D. Infants’ intentionally communicative vocalizations elicit responses from caregivers and predict the transition to language. Dev Sci. 2020;23:e12843. https://doi.org/10.1111/desc.12843
3. Mani N, Plunkett K. Phonological priming and cohort effects in toddlers. Cognition. 2011;121(2):196-206. https://doi.org/10.1016/j.cognition.2011.06.013
4. Donnelly S, Kidd E. The longitudinal relationship between conversational turn-taking and vocabulary growth in early language development. Child Dev. 2021;92(2):609-25. https://doi.org/10.1111/cdev.13511
5. Golinkoff RM, Hoff E, Rowe ML, Tamis-LeMonda CS, Hirsh-Pasek K. Language matters: denying the existence of the 30-million-word gap has serious consequences. Child Dev. 2019;90(3):985-92. https://doi.org/10.1111/cdev.13128
6. Anderson NJ, Graham SA, Prime H, Jenkins JM, Madigan S. Linking quality and quantity of parental linguistic input to child language skills: a meta-analysis. Child Dev. 2021;92(2):484-501. https://doi.org/10.1111/cdev.13508
7. Rowe ML, Snow CE. Analyzing input quality along three dimensions: interactive, linguistic, and conceptual. J Child Lang. 2020;47(1):5-21. https://doi.org/10.1017/S0305000919000655
8. Boerma T, ter Haar S, Ganga R. What risk factors for developmental language disorder can tell us about the neurobiological mechanisms of language development. Neurosci Biobehav Rev. 2023;154:105398. https://doi.org/10.1016/j.neubiorev.2023.105398
9. Hessami K, Norooznezhad AH, Monteiro S. COVID-19 pandemic and infant neurodevelopmental impairment: a systematic review and meta-analysis. JAMA Netw Open. 2022;5(10):e2238941. https://doi.org/10.1001/jamanetworkopen.2022.38941
10. Centro Nacional para la Salud de la Infancia y Adolescencia (CeNSIA). Manual para la aplicación de la prueba EDI [online] [accessed 27/01/2026]. Available at www.gob.mx/salud/censia/articulos/manual-para-la-aplicacion-de-la-prueba-edi/
11. Pejovic J, Severino C, Vigário M, Frota S. Prolonged COVID-19-related effects on early language development: a longitudinal study. Early Hum Dev. 2024;195:106081. https://doi.org/10.1016/j.earlhumdev.2024.106081
12. Feijoo S, Amadó A, Sidera F, Aguilar-Mediavilla E, Serrat E. Language acquisition in a post-pandemic context: impact of COVID-19 measures on early language development. Front Psychol. 2023;14:1205294. https://doi.org/10.3389/fpsyg.2023.1205294
13. Lee CS, Hwang S. Association between COVID-19 exposure duration on receptive and expressive language development in preschool children. Children (Basel). 2025;12(12):1637. https://doi.org/10.3390/children12121637
14. Kyvrakidou E, Kyvrakidis G, Stefanaki AS. The impact of COVID-19 on the language skills of preschool children: data from a school screening project in Greece. Children (Basel). 2025;12(3):376. https://doi.org/10.3390/children12030376
15. Cassinat JR, Whiteman SD, Serang S. Changes in family chaos and family relationships during the COVID-19 pandemic. Dev Psychol. 2021;57(10):1597-1610. https://doi.org/10.1037/dev0001217
16. Cristia A, Lavechin M, Scaff C. A thorough evaluation of the Language ENvironment Analysis (LENA) system. Behav Res Methods. 2021;53:467-86. https://doi.org/10.3758/s13428-020-01393-5
17. Göppert F, Bhatia K, Meyen S, Franz VH. Realistic expectations for replications: expecting too little is as bad as expecting too much. Adv Methods Pract Psychol Sci. 2025;8(3):1-8. https://doi.org/10.1177/25152459251369846
18. Nass Kunstmann L, Merino JM. El experimento natural como un nuevo diseño cuasi-experimental en investigación social y de salud [online] [accessed 27/01/2026]. Available at www.scielo.cl/scielo.php?script=sci_arttext&pid=s0717-95532008000200002&lng=es
19. Weng W, Liu M, Wang S. The impact of lockdown on child adjustment: a propensity score matched análisis. BMC Psychol. 2024;12:409. https://doi.org/10.1186/s40359-024-01894-4
20. Şengül-İnal G, Borgen NT, Skopek J, Nærde A, Zachrisson HD. Maternal education, early language skills, and mother-child interactions across three welfare states. J Marriage Fam. 2025;87(4):1549-70. https://doi.org/10.1111/jomf.13087
21. Madigan S, Plamondon A, Jenkins JM. Association between maternal sensitivity and infant receptive language development: quasi-causal evidence using a sibling-comparison design. Dev Psychol. 2023;59(12):2265-76. https://doi.org/10.1037/dev0001604
22. Ahmad M, Sechi C, Vismara L. Advanced maternal age: an exploratory review of psychological impact on mothers, infants, and their relationship. Behav Sci (Basel). 2024;14(3):147. https://doi.org/10.3390/bs14030147
23. Falster K, Hanly M, Banks E. Maternal age and offspring developmental vulnerability at age five: a population-based cohort study. PLoS Med. 2018;15(4):e1002558. https://doi.org/10.1371/journal.pmed.1002558
24. Singh L, Tan A, Quinn PC. Infants recognize words spoken through opaque masks but not through transparent masks. Dev Sci. 2021;24:e13117. https://doi.org/10.1111/desc.13117
25. Chen Y, Cabrera NJ, Reich SM. Mother–child and father–child serve-and-return interactions at 9 months: associations with children’s language skills at 18 and 24 months. Infant Behav Dev. 2023;73:101894. https://doi.org/10.1016/j.infbeh.2023.101894
26. Lalonde K, Buss E, Miller MK, Leibold LJ. Face masks impact auditory and audiovisual consonant recognition in children with and without hearing loss. Front Psychol. 2022;13:874345. https://doi.org/10.3389/fpsyg.2022.874345
27. Frausel RR, Richland LE, Levine SC, Goldin-Meadow S. Personal narrative as a breeding ground for higher-order thinking talk in early parent–child interactions. Dev Psychol. 2021;57(4):519-34. https://doi.org/10.1037/dev0001166
28. Winstone LK, Benitez VL, van Huisstede L. Patterns of maternal interactive behaviors and dual vocabulary development in Mexican American children. Dev Psychol. 2021;57(11):1866-79. https://doi.org/10.1037/dev0001024
29. Donnelly S, Kidd E. Individual differences in lexical processing efficiency and vocabulary in toddlers: a longitudinal investigation. J Exp Child Psychol. 2020;192:104781. https://doi.org/10.1016/j.jecp.2019.104781
30. Kidd E, Donnelly S, Christiansen MH. Individual differences in language acquisition and processing. Trends Cogn Sci. 2018;22(2):154-69. https://doi.org/10.1016/j.tics.2017.11.006
31. Kuhl PK. Is speech learning gated by the social brain? Dev Sci. 2007;10:110-20. https://doi.org/10.1111/j.1467-7687.2007.00572.x
32. Kidd E, Donnelly S. Individual differences in first language acquisition. Annu Rev Linguist. 2020;6:319-40. https://doi.org/10.1146/annurev-linguistics-011619-030326
33. Borovsky A. Lexical–semantic structure of vocabulary and its relation to lexical processing in early childhood and language outcomes at age three. Dev Psychol. 2022;58(4):607-30. https://doi.org/10.1037/dev0001291

Comments

This article has no comments yet.