The impact of poor air quality on neurocognitive health is a global concern. This impact was recently quantified by the Global Burden of Disease Study in India, attributing 1.67 million deaths in 2019 to air pollution-related causes with an overall national economic loss of $36.8 billion (Pandey et al., 2021). Such economic losses are compounded by evidence that poor air quality impacts neurocognitive health in childhood as economic losses accumulate over years of reduced productivity (Heckman and Mosso, 2014). Studies have reported that poor air quality and proximity to roadways is associated with reduced general cognitive functioning in childhood (Freire et al., 2010; Wang et al., 2009; Harris et al., 2015; Suglia et al., 2008) and slower growth in working memory (Sunyer et al., 2015). Exposure to poor air quality is also a risk factor for child emotional and behavioural problems which can have severe impacts on families (Midouhas et al., 2019).

But how early do these effects emerge? This is an important question as animal studies show profound impacts on the brain in early development, with inhalation of diesel exhaust and ultrafine particles resulting in elevated cytokine expression and oxidative stress in the brain (Gerlofs-Nijland et al., 2010; Bos et al., 2012), as well as altered neurogenesis (Patten et al., 2020). Very small particulate fragments (particulate matter with diameter of <2.5 μm; PM_2.5) are of major concern as they can move from the respiratory tract into the circulatory system reaching the brain. The brain may also be particularly sensitive during infancy due an immature detoxification response (Grandjean and Landrigan, 2014). Several large-scale studies have looked at the effects of prenatal exposure to nitrous oxide and PM_2.5 on human cognition early in development (Guxens et al., 2012; Guxens et al., 2014). Results show reduced psychomotor functioning at 1–6 years of age, but no associations with early cognition, although three studies have reported slower cognitive growth and emotional / conduct problems in children exposed to poorer indoor air quality as assessed via questionnaires (Midouhas et al., 2019; Gonzalez-Casanova et al., 2018; Vrijheid et al., 2012).

Critically, no studies have looked at the relationship between poor air quality and cognition in the first year of life when brain size doubles and may be particularly sensitive to toxins. This may reflect the challenge of assessing cognition in infancy. While standardized measures exist (e.g. Mullen Scales of Early Learning), these tools may not generalize to non-Western cultures where air quality is poorest. An alternative is to measure infant cognition using specially-designed looking-based tasks. Multiple aspects of visual cognition that are predictive of later cognitive abilities (Rose et al., 2012) can be measured reliably in the first year across cultures (Rose, 1994; Wijeakumar et al., 2019). For instance, visual processing speed measured using looking-based tasks in both infancy and toddlerhood is significantly predictive of working memory and executive function scores at 11 years (Rose et al., 2012). A second challenge is that infants spend much of the first year indoors. Consequently, data from outdoor monitoring stations may not accurately reflect air quality exposure critical for early brain development, particularly in contexts where use of, for instance, solid fuels in cooking may lead to differences between indoor and outdoor air quality. Recent advances in air quality monitoring allow for the measurement of PM_2.5 directly in homes; this may be critical as PM_2.5 is possibly the most neurotoxic component of residential air quality (Block and Calderón-Garcidueñas, 2009).

Here, we examined the relationship between poor air quality and cognition in infancy using looking-based measures of cognition and in-home measures of PM_2.5. To assess visual cognition, we used a specially-designed, transculturally-relevant visual cognition task which assessed infants’ visual working memory and processing speed (Figure 1A and B; see Ross-Sheehy et al., 2003), capitalizing on infants’ tendency to look away from visual familiarity and toward visual novelty. This task has been examined in several prior studies (Oakes et al., 2006; Oakes et al., 2009), but this is the first study to use this task longitudinally with a sample of non-western infants. Infants visually explored two displays that blinked ‘on’ and ‘off’. On one side – the ‘no change’ side – the same colored squares were always presented; on the other side, one randomly-selected square changed color after each blink. If infants begin looking at the ‘no change’ side and they can remember the colors in working memory, they should lose interest in this display, releasing fixation to visually explore the ‘changing’ display. Here, they should detect the novel color and sustain looking to this display, leading to a strong change preference – a high proportion of looking to the changing side. This change preference is modulated by visual working memory capacity. When the number of items on the screen is small (i.e. the memory load is low), infants can detect the difference between the ‘no change’ and ‘change’ displays, and they show a higher change preference. With increasing memory load (and increasing difficulty), infants may falsely detect changes on the ‘no change side’ and have difficulty releasing fixation, leading to a lower change preference (Perone et al., 2011). Thus, the change preference score yields a quantifiable measure of infants’ visual working memory abilities. Note that if infants begin looking at the ‘change’ side, they should remain looking at this display with a consistently high change preference score (see Methods for further discussion).

Our study was conducted in Shivgarh, India, a rural community in Uttar Pradesh, one of the states in India that has been most strongly impacted by poor air quality (Pandey et al., 2021). We report data from 215 families from a range of socio-economic backgrounds (see Table 1). Infants were enrolled at 6 months (N=108) or 9 months of age (N=107) at which time they completed an in-lab assessment of visual cognition (see Figure 1A) as well as standardized psychomotor and cognitive assessments. A similar assessment was repeated a year later when the same infants were 18 or 21 months of age (see Methods). We examined these two cohorts based on evidence that there is an improvement in visual working memory capacity between 6 and 8 months. In particular, Ross-Sheehy and colleagues (Ross-Sheehy et al., 2003) showed that 6.5-month-old infants demonstrate greater-than-chance change preference scores with a memory load of one item, while 10- and 13-month-old infants showed greater-than-chance change preference scores for memory loads of two and three items. Similarly, 6.5-month-old infants could remember one spatial location, while 8- and 12.5-month-old infants showed evidence of remembering multiple locations (Oakes et al., 2011). By assessing infants in rural India on either side of this transition, we hoped to either replicate a similar transition or, alternatively, to reveal a delay in this transition across cultures.

We focused on three primary measures from the preferential looking task: (1) the shift rate, that is, the number of looks back and forth (per second), (2) the change preference score for trials where infants’ ‘first look’ was to the ‘no change’ side, and (3) the change preference score for trials where infants’ ‘first look’ was to the ‘change’ side. Use of the shift rate measure was motivated by prior work showing that visual processing speed in infancy is predictive of longer-term cognitive outcomes (Rose et al., 2012). The latter two measures are an adaptation of the standard ‘change preference’ looking measure from this task. Note that because the ‘first-look no-change’ measure starts at 0 (i.e. looking to no change) and the ‘first-look change’ measure starts at 1 (i.e. looking to change), chance performance is no longer anchored at 0.5. Although this differs from prior work, the change was motivated by evidence that the standard measure is not predictive longitudinally, while the adapted measures are (see Methods for discussion). Note further that we focus on the ‘first look no-change’ measure below (referred to as the ‘change preference’ score for simplicity), as the only significant finding for the ‘first look change’ measure was a decrease in the change preference score from year 1 to year 2 (see Table 10 and Table 11).

Six-month-old infants showed lower change preference scores than 9-month-old infants (Figure 1C), Χ² (1)=3.66, p=0.056 η_p ² = 0.02 (for full results, see Table 2). In addition, change preference scores decreased as the memory load increased, Χ² (2)=12.53, p=0.002 η_p ² = 0.01 (see Figure 1D). Similar effects were observed in an analysis of infants’ rate of looking back and forth between the displays (i.e. shift rate; see Table 3). Infants’ shift rate decreased as the memory load increased, Χ² (2)=16.52, p<0.001 η_p ² = 0.02 (see Figure 1E). In addition, infants’ standardized cognitive scores in year 1, F(1) = 10.94, p=0.001 η_p ² = 0.050, and year 2, F(1) = 15.99, p<0.001 η_p ² = 0.080, were consistently lower for low SES infants (see Figure 1F and G and Table 4).

Figure 1

Download asset Open asset

We recorded in-home air quality using a laser particle sensor (Air Visual Node, Atlanta Healthcare, Inc) placed in the home (see Figure 2A) for 3 continuous days during each assessment period. Field workers were instructed to place the sensor in the room where infants slept or spent most of their time. We re-assessed air quality up to six times for each family (every three months in-between lab visits; M visits = 4, SD = 1.16). These data were modelled and a participant-level score which adjusted for season and time of day was extracted for use in the analyses (see Methods). We focused on PM_2.5 concentrations expressed as US Air Quality Index (AQI) values. This index maps PM_2.5 concentrations measured in μg/m³ to a more intuitive categorical scale where AQI values under 50 are good, values from 51 to 100 indicate moderate air quality, values 101–150 indicate air that is unhealthy for sensitive groups, and values higher than 151 are considered unhealthy extending up to hazardous (>301). Note that AQI values can be readily converted to μg/m³ using on-line calculators (e.g. https://www.airnow.gov/aqi/aqi-calculator/).

Figure 2

Download asset Open asset

As can be seen in Figure 2B, air quality in Shivgarh was quite poor, with AQI values often higher than 151, with an overall mean of 207. This is comparable to recent AQI data for the northern states in India that ranged from 186 to 267 (Pandey et al., 2021). We also compared in-home air quality to daily outdoor air quality from the nearest monitoring station in Lucknow (red line in Figure 2B; see Central Pollution Control Board in India: https://cpcb.nic.in). Annual fluctuations in in-home air quality generally mirrored fluctuations in outdoor air quality in Lucknow, although with a lower peak in the winter of 2018–2019 (for similar seasonal variations in indoor and outdoor air quality, see Rohra and Taneja, 2016). We note that this lower peak occurred during a pause in our indoor air quality data collection; thus, we cannot evaluate if this is a mismatch or simply a result of limited data during this period in the indoor air quality model.

In-home air quality (PM_2.5) was strongly influenced by daily meal preparation, with peaks in poor air quality occurring during meal preparation times (see Figure 2C; see Mukhopadhyay et al., 2012 for similar meal preparation findings). Air quality was poorest in households using cow dung for fuel and best in households using LPG, F(2) = 3.23, p=0.042 η_p ² = 0.030 (see Figure 2D; Table 5). Note that we were not able to look at the impact of cooking fuel independently from SES as most families that used LPG fuel fell into the high SES tertile (see Table 1).

Next, we explored the association between air quality (PM_2.5) and cognitive measures while controlling for both age cohort and SES (see Methods). Critically, infants living in homes with poor air quality had poorer visual cognitive performance. Infants from households with poorer air quality had lower change preference scores in year 1, an effect that was attenuated in year 2, Χ² (1)=5.91, p=0.015 η_p ² = 0.006 (see Figure 3A and Year x Air Quality interaction in Table 6). We also found a strong negative association of air quality with visual processing speed (shift rate), Χ² (1)=4.69, p=0.03 η_p ² = 0.02 (see Air Quality main effect in Table 7). As can be seen in Figure 3B, infants from households with poor air quality showed slower rates of visual processing. Note that both effects shown in Figure 3 were robust in models that controlled for SES using the modified Kuppuswamy Scale which aggregates effects of family occupation, education, and income (see Mohd Saleem, 2020). Moreover, in both cases, models that included air quality captured a significant proportion of variance above and beyond the baseline cognitive models: models comparing the baseline change preference model (Table 2) to a model that included air quality (Table 6) showed that the air quality model captured a greater proportion of variance in the change preference scores, Χ² (2)=6.35, p=0.04; similarly, comparing the baseline visual processing speed model (Table 3) to a model that included air quality (Table 7) showed that the air quality model captured a greater proportion of variance in processing speed, Χ² (1)=4.74, p=0.03.

Figure 3

Download asset Open asset

One concern with these findings is that change preference and shift rate are measured in the same task and are correlated. Thus, it is possible the findings above reflect some shared variance rather than separate effects. To examine this, we re-ran the change preference / air quality analysis, adding shift rate as a predictor of the change preference score. This analysis revealed a main effect of shift rate, Χ² (1)=17.36, p<0.001 η_p ² = 0.02, confirming the relationship with the change preference score; however, the Load and Year x Air Quality interaction remained significant (see Supplementary file 2). Thus, the effects shown in Figure 3A are statistically robust, even when shift rate is included in the change preference model. We also re-ran the shift rate / air quality model, including the change preference score as a predictor. This analysis revealed a main effect of change preference score, Χ² (1)=23.02, p<0.001 η_p ² = 0.02. Critically, however, the Load and Air Quality main effects remained significant even when the change preference score was included as a predictor (see Supplementary file 3).

Another key question is why the change preference measure only showed a robust association with air quality in the first year. As discussed in the Methods, we modulated the working memory load in year 2 to make the task more challenging and age-appropriate; thus, it is possible we made the task too hard for some infants, dampening our ability to detect individual differences in working memory. To explore this possibility, we re-ran the change preference / air quality analysis, only using data from load 2 trials (i.e. the ‘medium’ load in year 1 and the ‘low’ load in year 2; see Methods). In this analysis with identical task stimuli, we still found a significant Year x Air Quality interaction, F(1,340) = 6.83, p=0.009 η_p ² = 0.02 (see Supplementary file 4).

Finally, we note that air quality did not significantly impact standardized cognitive scores in year 1 (measured using the Mullen Scales) or year 2 (measured using the ASQ; see Tables 8 and 9). This may indicate that the effects of air quality are specific to the visual cognitive system rather than impacting the more general aspects of cognition and psychomotor function assessed by these measures.

The present findings indicate that poor air quality (PM_2.5) is associated with slower visual processing speed in the first two years of life and poorer visual working memory scores in year 1. These negative impacts were evident only for looking-based measures of cognition. Our results contrast with findings from prior studies that have failed to show an association between outdoor air quality and cognition in early development. It is possible that this difference reflects the broader range of PM_2.5 exposure in our sample. For instance, Guxens and colleagues (Guxens et al., 2014) reported no systematic relationships between air quality and cognitive measures across six European birth cohorts (although they did find effects on psychomotor function); however, air quality ranged from AQI values of 53–72. Our mean AQI value (207) was three to four times higher. Thus, infants in the present report were exposed to much poorer air quality which might explain the strong relationship with visual cognition.

It is also possible that indoor levels of PM_2.5 are critical to cognition in infancy. Most prior studies have looked at outdoor air quality, although several studies have reported a negative impact of poor indoor air quality on cognition as assessed via questionnaires (Midouhas et al., 2019; Gonzalez-Casanova et al., 2018; Vrijheid et al., 2012; Midouhas et al., 2018). Interestingly, Vrijheid et al., 2012 found that use of a gas cooker in the home showed a stronger negative association with standardized cognitive scores after 14 months. Our findings extend this work by also showing an association between poor indoor air quality and visual cognition as early as 6 months of age.

One question raised by our findings is why visual working memory effects were isolated to year 1. We investigated one possibility – that the working memory task was too difficult in year 2. Results suggested that this was not the case: when we compared identical conditions across years 1 and 2 we still found an inverse relationship between change preference scores and air quality in year 1 but not in year 2. Another possibility is that the impact of indoor air quality on visual cognition wanes in year 2. Our findings for shift rate argue against this possibility as these findings were robust across both years. The robustness of the shift rate effect may suggest that visual processing speed is a particularly sensitive measure in infancy, consistent with other work showing that visual processing speed in infancy is predictive of schooling outcomes 11 years later (Rose et al., 2012). A third possibility for why the impact of air quality on change preference scores wanes in year 2 is that other factors which are not correlated with air quality have a stronger impact on visual working memory in year 2. For instance, we are currently examining how interactions with caregivers impact infants’ visual working memory abilities. We suspect these interactions have a strong influence on visual working memory development, particularly in year 2 after extended interactions accumulate over time. Interestingly, infant-led interactions – which have been shown to support working memory development (Landry et al., 2006; Perone and Spencer, 2013) – appear to be quite frequent in low SES families in our sample. It is possible such positive influences counter the impact of poor air quality in some families.

Another interesting result from the present study was the specificity of our findings to the visual cognition task. In particular, although results from the Mullen and Ages and Stages Questionnaire showed robust relationships with SES suggesting good sensitivity, these measures showed no associations with air quality. Thus, air quality may have specific effects in infancy, targeting early emerging cognitive systems. Future work will be needed to determine if this is the case. For instance, it would be interesting to examine if air quality has an impact on auditory and/or statistical learning processes in the first two years of life (Saffran et al., 1996).

Strengths of the current study include our use of multiple measures to assess infant cognition including looking-based measures, the longitudinal design, inclusion of a large sample, and a high density of air quality measurements for each household. We also included a continuous range of SES families from the same micro-cultural context, allowing us to tease apart influences of SES while holding culture relatively constant.

Regarding limitations, we note that effect sizes were relatively low in the present study. In addition, there was some instability in the air quality devices over time (see Methods). Note that although laser particle sensors have shown robust correlations with beta attenuation monitors in laboratory and field tests (Zhang and Srinivasan, 2020), it would be ideal in future work to use personal air quality monitors such as gravimetric devices that correct light scatter and use weighted PM_2.5. This would enable a direct, localized measure of the air each child is exposed to.

Another limitation of the present study was the absence of more detailed information about cooking practices in the home. Although we identified the primary cooking fuel used in each household, it is likely some households used multiple cooking methods, and other fuels for warming, smoking, and so on (Rohra and Taneja, 2016; Mukhopadhyay et al., 2012). Future work will also be needed to carefully examine whether poor air quality has a causal influence on cognition in early development. Given recent data on the impact of poor air quality on the developing brain in animals, future work using neuroimaging tools might be particularly useful to clarify mechanistic pathways in infancy.

Our data suggest that global efforts to improve air quality could have benefits to infants’ emerging cognitive abilities. This, in turn, could have a cascade of positive impacts, including positive impacts on families as well as economic consequences as improved cognition can lead to improved economic productivity longer-term and reduce the burden on healthcare and mental health systems (Moffitt et al., 2011). Our results showing links between indoor air quality and cooking materials also suggest that efforts to reduce cooking emissions in homes should be a key target for intervention. This requires both increased availability of clean technologies and uptake of such technologies in rural households where traditional methods of meal preparation might be a barrier (see Mukhopadhyay et al., 2012). Our findings can motivate both policymakers and families to improve air quality as this should positively boost the neurocognitive health of young infants.

All data and code are available at https://osf.io/fspzb/.

1. 1. Spencer JP
   2. Forbes S

   (2023) Open Science Framework

   ID fspzb. Air quality and visual working memory.

   https://osf.io/fspzb/

1. 1. Ahmed I
   2. Ali SM
   3. Amenga-Etego S
   4. Ariff S
   5. Bahl R
   6. Baqui AH
   7. Begum N
   8. Bhandari N
   9. Bhatia K
   10. Bhutta ZA
   11. Biemba G
   12. Deb S
   13. Dhingra U
   14. Dube B
   15. Dutta A
   16. Edmond K
   17. Esamai F
   18. Fawzi W
   19. Ghosh AK
   20. Gisore P
   21. Grogan C
   22. Hamer DH
   23. Herlihy J
   24. Hurt L
   25. Ilyas M
   26. Jehan F
   27. Kalonji M
   28. Kaur J
   29. Khanam R
   30. Kirkwood B
   31. Kumar A
   32. Kumar A
   33. Kumar V
   34. Manu A
   35. Marete I
   36. Masanja H
   37. Mazumder S
   38. Mehmood U
   39. Mishra S
   40. Mitra DK
   41. Mlay E
   42. Mohan SB
   43. Moin MI
   44. Muhammad K
   45. Muhihi A
   46. Newton S
   47. Ngaima S
   48. Nguwo A
   49. Nisar I
   50. O’Leary M
   51. Otomba J
   52. Patil P
   53. Quaiyum MA
   54. Rahman MH
   55. Sazawal S
   56. Semrau KE
   57. Shannon C
   58. Smith ER
   59. Soofi S
   60. Soremekun S
   61. Sunday V
   62. Taneja S
   63. Tshefu A
   64. Wasan Y
   65. Yeboah-Antwi K
   66. Yoshida S
   67. Zaidi A

   (2018) Population-Based rates, timing, and causes of maternal deaths, stillbirths, and neonatal deaths in South Asia and sub-Saharan Africa: a multi-country prospective cohort study

   The Lancet. Global Health 6:e1297–e1308.

   https://doi.org/10.1016/S2214-109X(18)30385-1

   • PubMed
   • Google Scholar
2. 1. Ben-Shachar M
   2. Lüdecke D
   3. Makowski D

   (2020) Effectsize: estimation of effect size indices and standardized parameters

   Journal of Open Source Software 5:2815.

   https://doi.org/10.21105/joss.02815

   • PubMed
   • Google Scholar
3. 1. Block ML
   2. Calderón-Garcidueñas L

   (2009) Air pollution: mechanisms of neuroinflammation and CNS disease

   Trends in Neurosciences 32:506–516.

   https://doi.org/10.1016/j.tins.2009.05.009

   • PubMed
   • Google Scholar
4. 1. Bos I
   2. De Boever P
   3. Int Panis L
   4. Sarre S
   5. Meeusen R

   (2012) Negative effects of ultrafine particle exposure during forced exercise on the expression of brain-derived neurotrophic factor in the hippocampus of rats

   Neuroscience 223:131–139.

   https://doi.org/10.1016/j.neuroscience.2012.07.057

   • PubMed
   • Google Scholar
5. Software

   1. Datavyu

   (2014) Datavyu: A video coding tool

   Datavyu.

   http://datavyu.org
6. 1. Delgado Reyes L
   2. Wijeakumar S
   3. Magnotta VA
   4. Forbes SH
   5. Spencer JP

   (2020) The functional brain networks that underlie visual working memory in the first two years of life

   NeuroImage 219:116971.

   https://doi.org/10.1016/j.neuroimage.2020.116971

   • PubMed
   • Google Scholar
7. Software

   1. Dink J
   2. Ferguson B

   (2021) EyetrackingR, version version 0.2.0

   R Package.

   http://www.eyetracking-r.com/
8. 1. Fernald A
   2. Pinto JP
   3. Swingley D
   4. Weinberg A
   5. McRoberts GW

   (1998) Rapid gains in speed of verbal processing by infants in the 2nd year

   Psychological Science 9:228–231.

   https://doi.org/10.1111/1467-9280.00044

   • Google Scholar
9. 1. Freire C
   2. Ramos R
   3. Puertas R
   4. Lopez-Espinosa MJ
   5. Julvez J
   6. Aguilera I
   7. Cruz F
   8. Fernandez MF
   9. Sunyer J
   10. Olea N

   (2010) Association of traffic-related air pollution with cognitive development in children

   Journal of Epidemiology and Community Health 64:223–228.

   https://doi.org/10.1136/jech.2008.084574

   • PubMed
   • Google Scholar
10. 1. Gerlofs-Nijland ME
    2. van Berlo D
    3. Cassee FR
    4. Schins RPF
    5. Wang K
    6. Campbell A

    (2010) Effect of prolonged exposure to diesel engine exhaust on proinflammatory markers in different regions of the rat brain

    Particle and Fibre Toxicology 7:1–10.

    https://doi.org/10.1186/1743-8977-7-12

    • PubMed
    • Google Scholar
11. 1. Gonzalez-Casanova I
    2. Stein AD
    3. Barraza-Villarreal A
    4. Feregrino RG
    5. DiGirolamo A
    6. Hernandez-Cadena L
    7. Rivera JA
    8. Romieu I
    9. Ramakrishnan U

    (2018) Prenatal exposure to environmental pollutants and child development trajectories through 7 years

    International Journal of Hygiene and Environmental Health 221:616–622.

    https://doi.org/10.1016/j.ijheh.2018.04.004

    • PubMed
    • Google Scholar
12. 1. Grandjean P
    2. Landrigan PJ

    (2014) Neurobehavioural effects of developmental toxicity

    The Lancet. Neurology 13:330–338.

    https://doi.org/10.1016/S1474-4422(13)70278-3

    • PubMed
    • Google Scholar
13. 1. Guxens M
    2. Aguilera I
    3. Ballester F
    4. Estarlich M
    5. Fernández-Somoano A
    6. Lertxundi A
    7. Lertxundi N
    8. Mendez MA
    9. Tardón A
    10. Vrijheid M
    11. Sunyer J
    12. INMA Project

    (2012) Prenatal exposure to residential air pollution and infant mental development: modulation by antioxidants and detoxification factors

    Environmental Health Perspectives 120:144–149.

    https://doi.org/10.1289/ehp.1103469

    • PubMed
    • Google Scholar
14. 1. Guxens M
    2. Garcia-Esteban R
    3. Giorgis-Allemand L
    4. Forns J
    5. Badaloni C
    6. Ballester F
    7. Beelen R
    8. Cesaroni G
    9. Chatzi L
    10. de Agostini M
    11. de Nazelle A
    12. Eeftens M
    13. Fernandez MF
    14. Fernández-Somoano A
    15. Forastiere F
    16. Gehring U
    17. Ghassabian A
    18. Heude B
    19. Jaddoe VWV
    20. Klümper C
    21. Kogevinas M
    22. Krämer U
    23. Larroque B
    24. Lertxundi A
    25. Lertxuni N
    26. Murcia M
    27. Navel V
    28. Nieuwenhuijsen M
    29. Porta D
    30. Ramos R
    31. Roumeliotaki T
    32. Slama R
    33. Sørensen M
    34. Stephanou EG
    35. Sugiri D
    36. Tardón A
    37. Tiemeier H
    38. Tiesler CMT
    39. Verhulst FC
    40. Vrijkotte T
    41. Wilhelm M
    42. Brunekreef B
    43. Pershagen G
    44. Sunyer J

    (2014) Air pollution during pregnancy and childhood cognitive and psychomotor development: six European birth cohorts

    Epidemiology 25:636–647.

    https://doi.org/10.1097/EDE.0000000000000133

    • PubMed
    • Google Scholar
15. 1. Harris MH
    2. Gold DR
    3. Rifas-Shiman SL
    4. Melly SJ
    5. Zanobetti A
    6. Coull BA
    7. Schwartz JD
    8. Gryparis A
    9. Kloog I
    10. Koutrakis P
    11. Bellinger DC
    12. White RF
    13. Sagiv SK
    14. Oken E

    (2015) Prenatal and childhood traffic-related pollution exposure and childhood cognition in the project viva cohort (Massachusetts, USA)

    Environmental Health Perspectives 123:1072–1078.

    https://doi.org/10.1289/ehp.1408803

    • PubMed
    • Google Scholar
16. Software

    1. Hartig F

    (2020) DHARMa: residual diagnostics for hierarchical (multi-level / mixed) regression models

    R Packages.

    https://cran.r-project.org/web/packages/DHARMa/vignettes/DHARMa.html
17. 1. Heckman JJ
    2. Mosso S

    (2014) The economics of human development and social mobility

    Annual Review of Economics 6:689–733.

    https://doi.org/10.1146/annurev-economics-080213-040753

    • PubMed
    • Google Scholar
18. 1. Kuznetsova A
    2. Brockhoff PB
    3. Christensen RHB

    (2017) lmertest package: tests in linear mixed effects models

    Journal of Statistical Software 82:1–26.

    https://doi.org/10.18637/jss.v082.i13

    • Google Scholar
19. 1. Landry SH
    2. Smith KE
    3. Swank PR

    (2006) Responsive parenting: establishing early foundations for social, communication, and independent problem-solving skills

    Developmental Psychology 42:627–642.

    https://doi.org/10.1037/0012-1649.42.4.627

    • PubMed
    • Google Scholar
20. 1. Midouhas E
    2. Kokosi T
    3. Flouri E

    (2018) Outdoor and indoor air quality and cognitive ability in young children

    Environmental Research 161:321–328.

    https://doi.org/10.1016/j.envres.2017.11.026

    • PubMed
    • Google Scholar
21. 1. Midouhas E
    2. Kokosi T
    3. Flouri E

    (2019) The quality of air outside and inside the home: associations with emotional and behavioural problem scores in early childhood

    BMC Public Health 19:406.

    https://doi.org/10.1186/s12889-019-6733-1

    • PubMed
    • Google Scholar
22. 1. Moffitt TE
    2. Arseneault L
    3. Belsky D
    4. Dickson N
    5. Hancox RJ
    6. Harrington H
    7. Houts R
    8. Poulton R
    9. Roberts BW
    10. Ross S
    11. Sears MR
    12. Thomson WM
    13. Caspi A

    (2011) A gradient of childhood self-control predicts health, wealth, and public safety

    PNAS 108:2693–2698.

    https://doi.org/10.1073/pnas.1010076108

    • PubMed
    • Google Scholar
23. 1. Mohd Saleem S

    (2020) Modified kuppuswamy socioeconomic scale updated for the year 2020

    Indian Journal of Forensic and Community Medicine 7:1–3.

    https://doi.org/10.18231/j.ijfcm.2020.001

    • Google Scholar
24. Website

    1. Molenar JV

    (2016) Comparison of PM2.5 Measurements Using the AirVisual Monitor and Beta Attenuation Monitor

    Accessed May 10, 2017.

    https://www.airvisual.com/air-pollution-information/research/comparison-of-pm2-5-measurements-using-the-airvisual-monitor-and-beta-attenuation-monitor)
25. 1. Mukhopadhyay R
    2. Sambandam S
    3. Pillarisetti A
    4. Jack D
    5. Mukhopadhyay K
    6. Balakrishnan K
    7. Vaswani M
    8. Bates MN
    9. Kinney PL
    10. Arora N
    11. Smith KR

    (2012) Cooking practices, air quality, and the acceptability of advanced cookstoves in Haryana, India: an exploratory study to inform large-scale interventions

    Global Health Action 5:1–13.

    https://doi.org/10.3402/gha.v5i0.19016

    • PubMed
    • Google Scholar
26. 1. Oakes LM
    2. Ross-Sheehy S
    3. Luck SJ

    (2006) Rapid development of feature binding in visual short-term memory

    Psychological Science 17:781–787.

    https://doi.org/10.1111/j.1467-9280.2006.01782.x

    • PubMed
    • Google Scholar
27. 1. Oakes LM
    2. Messenger IM
    3. Ross-Sheehy S
    4. Luck SJ

    (2009) New evidence for rapid development of color-location binding in infants’ visual short-term memory

    Visual Cognition 17:67–82.

    https://doi.org/10.1080/13506280802151480

    • PubMed
    • Google Scholar
28. 1. Oakes LM
    2. Hurley KB
    3. Ross-Sheehy S
    4. Luck SJ

    (2011) Developmental changes in infants’ visual short-term memory for location

    Cognition 118:293–305.

    https://doi.org/10.1016/j.cognition.2010.11.007

    • PubMed
    • Google Scholar
29. 1. Pandey A
    2. Brauer M
    3. Cropper ML
    4. Balakrishnan K
    5. Mathur P
    6. Dey S
    7. Turkgulu B
    8. Kumar GA
    9. Khare M
    10. Beig G
    11. Gupta T
    12. Krishnankutty RP
    13. Causey K
    14. Cohen AJ
    15. Bhargava S
    16. Aggarwal AN
    17. Agrawal A
    18. Awasthi S
    19. Bennitt F
    20. Bhagwat S
    21. Bhanumati P
    22. Burkart K
    23. Chakma JK
    24. Chiles TC
    25. Chowdhury S
    26. Christopher DJ
    27. Dey S
    28. Fisher S
    29. Fraumeni B
    30. Fuller R
    31. Ghoshal AG
    32. Golechha MJ
    33. Gupta PC
    34. Gupta R
    35. Gupta R
    36. Gupta S
    37. Guttikunda S
    38. Hanrahan D
    39. Harikrishnan S
    40. Jeemon P
    41. Joshi TK
    42. Kant R
    43. Kant S
    44. Kaur T
    45. Koul PA
    46. Kumar P
    47. Kumar R
    48. Larson SL
    49. Lodha R
    50. Madhipatla KK
    51. Mahesh PA
    52. Malhotra R
    53. Managi S
    54. Martin K
    55. Mathai M
    56. Mathew JL
    57. Mehrotra R
    58. Mohan BVM
    59. Mohan V
    60. Mukhopadhyay S
    61. Mutreja P
    62. Naik N
    63. Nair S
    64. Pandian JD
    65. Pant P
    66. Perianayagam A
    67. Prabhakaran D
    68. Prabhakaran P
    69. Rath GK
    70. Ravi S
    71. Roy A
    72. Sabde YD
    73. Salvi S
    74. Sambandam S
    75. Sharma B
    76. Sharma M
    77. Sharma S
    78. Sharma RS
    79. Shrivastava A
    80. Singh S
    81. Singh V
    82. Smith R

    (2021) Health and economic impact of air pollution in the states of India: the global burden of disease study 2019

    The Lancet. Planetary Health 5:e25–e38.

    https://doi.org/10.1016/S2542-5196(20)30298-9

    • PubMed
    • Google Scholar
30. 1. Patten KT
    2. González EA
    3. Valenzuela A
    4. Berg E
    5. Wallis C
    6. Garbow JR
    7. Silverman JL
    8. Bein KJ
    9. Wexler AS
    10. Lein PJ

    (2020) Effects of early life exposure to traffic-related air pollution on brain development in juvenile Sprague-Dawley rats

    Translational Psychiatry 10:166.

    https://doi.org/10.1038/s41398-020-0845-3

    • PubMed
    • Google Scholar
31. 1. Perone S
    2. Simmering VR
    3. Spencer JP

    (2011) Stronger neural dynamics capture changes in infants’ visual working memory capacity over development

    Developmental Science 14:1379–1392.

    https://doi.org/10.1111/j.1467-7687.2011.01083.x

    • PubMed
    • Google Scholar
32. 1. Perone S
    2. Spencer JP

    (2013) Autonomous visual exploration creates developmental change in familiarity and novelty seeking behaviors

    Frontiers in Psychology 4:648.

    https://doi.org/10.3389/fpsyg.2013.00648

    • PubMed
    • Google Scholar
33. 1. Rohra H
    2. Taneja A

    (2016) Indoor air quality scenario in india—an outline of household fuel combustion

    Atmospheric Environment 129:243–255.

    https://doi.org/10.1016/j.atmosenv.2016.01.038

    • Google Scholar
34. 1. Rose SA

    (1994)

    Relation between physical growth and information processing in infants born in India

    Child Development 65:889–902.

    • PubMed
    • Google Scholar
35. 1. Rose SA
    2. Feldman JF
    3. Jankowski JJ

    (2012) Implications of infant cognition for executive functions at age 11

    Psychological Science 23:1345–1355.

    https://doi.org/10.1177/0956797612444902

    • PubMed
    • Google Scholar
36. 1. Ross-Sheehy S
    2. Oakes LM
    3. Luck SJ

    (2003) The development of visual short-term memory capacity in infants

    Child Development 74:1807–1822.

    https://doi.org/10.1046/j.1467-8624.2003.00639.x

    • PubMed
    • Google Scholar
37. 1. Saffran JR
    2. Aslin RN
    3. Newport EL

    (1996) Statistical learning by 8-month-old infants

    Science 274:1926–1928.

    https://doi.org/10.1126/science.274.5294.1926

    • PubMed
    • Google Scholar
38. 1. Suglia SF
    2. Gryparis A
    3. Wright RO
    4. Schwartz J
    5. Wright RJ

    (2008) Association of black carbon with cognition among children in a prospective birth cohort study

    American Journal of Epidemiology 167:280–286.

    https://doi.org/10.1093/aje/kwm308

    • PubMed
    • Google Scholar
39. 1. Sunyer J
    2. Esnaola M
    3. Alvarez-Pedrerol M
    4. Forns J
    5. Rivas I
    6. López-Vicente M
    7. Suades-González E
    8. Foraster M
    9. Garcia-Esteban R
    10. Basagaña X
    11. Viana M
    12. Cirach M
    13. Moreno T
    14. Alastuey A
    15. Sebastian-Galles N
    16. Nieuwenhuijsen M
    17. Querol X

    (2015) Association between traffic-related air pollution in schools and cognitive development in primary school children: a prospective cohort study

    PLOS Medicine 12:e1001792.

    https://doi.org/10.1371/journal.pmed.1001792

    • PubMed
    • Google Scholar
40. Software

    1. van Rij J
    2. Wieling M
    3. Baayen R
    4. van Rijn H

    (2020) Itsadug: interpreting time series and autocorrelated data using gamms, version version 2.4.1

    R Package.

    https://cran.r-project.org/web/packages/itsadug/citation.html
41. 1. Vrijheid M
    2. Martinez D
    3. Aguilera I
    4. Bustamante M
    5. Ballester F
    6. Estarlich M
    7. Fernandez-Somoano A
    8. Guxens M
    9. Lertxundi N
    10. Martinez MD
    11. Tardon A
    12. Sunyer J
    13. INMA Project

    (2012) Indoor air pollution from gas cooking and infant neurodevelopment

    Epidemiology 23:23–32.

    https://doi.org/10.1097/EDE.0b013e31823a4023

    • PubMed
    • Google Scholar
42. 1. Wang S
    2. Zhang J
    3. Zeng X
    4. Zeng Y
    5. Wang S
    6. Chen S

    (2009) Association of traffic-related air pollution with children’s neurobehavioral functions in quanzhou, China

    Environmental Health Perspectives 117:1612–1618.

    https://doi.org/10.1289/ehp.0800023

    • PubMed
    • Google Scholar
43. 1. Wijeakumar S
    2. Kumar A
    3. Delgado Reyes LM
    4. Tiwari M
    5. Spencer JP

    (2019) Early adversity in rural India impacts the brain networks underlying visual working memory

    Developmental Science 22:e12822.

    https://doi.org/10.1111/desc.12822

    • PubMed
    • Google Scholar
44. 1. Wood SN

    (2011) Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models

    Journal of the Royal Statistical Society Series B 73:3–36.

    https://doi.org/10.1111/j.1467-9868.2010.00749.x

    • Google Scholar
45. 1. Zhang H
    2. Srinivasan R

    (2020) A systematic review of air quality sensors, guidelines, and measurement studies for indoor air quality management

    Sustainability 12:9045.

    https://doi.org/10.3390/su12219045

    • Google Scholar

Author details

1. John P Spencer

   School of Psychology, University of East Anglia, Norwich, United Kingdom

   Contribution

   Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Investigation, Visualization, Methodology, Writing – original draft, Project administration, Writing – review and editing

   For correspondence

   j.spencer@uea.ac.uk

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7320-144X

2. Samuel H Forbes

   Department of Psychology, Durham University, Durham, United Kingdom

   Contribution

   Resources, Data curation, Software, Formal analysis, Visualization, Writing – review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1022-4676

3. Sophie Naylor

   School of Psychology, University of East Anglia, Norwich, United Kingdom

   Contribution

   Data curation, Formal analysis

   Competing interests

   No competing interests declared

4. Vinay P Singh

   Community Empowerment Lab, Lucknow, India

   Contribution

   Resources, Data curation, Supervision, Methodology, Project administration

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-1421-8775

5. Kiara Jackson

   School of Psychology, University of East Anglia, Norwich, United Kingdom

   Contribution

   Data curation, Formal analysis

   Competing interests

   No competing interests declared

6. Sean Deoni

   Department of Pediatrics, Brown University, Providence, United States

   Contribution

   Conceptualization, Data curation, Supervision, Funding acquisition, Methodology, Project administration, Writing – review and editing

   Competing interests

   Sean Deoni has grants or contracts, received consulting fees from and has patents planned, issued or pending with Nestle Nutrition and has received Payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Wyeth Nutrition and Mead Johnson Nutrition. The author has no other competing interests to declare

7. Madhuri Tiwari

   Community Empowerment Lab, Lucknow, India

   Contribution

   Data curation, Supervision, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

8. Aarti Kumar

   Community Empowerment Lab, Lucknow, India

   Contribution

   Conceptualization, Data curation, Formal analysis, Supervision, Funding acquisition, Methodology, Project administration, Writing – review and editing

   Competing interests

   No competing interests declared

• 773

  views

• 105

  downloads

• 2

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.