Sage Journals: Discover world-class research

Abstract

Autism spectrum disorder is a developmental disorder that includes deficits in social communication and interaction, and restricted and repetitive behaviours, interests, or activities. This survey was done to assess autism spectrum disorder prevalence in 16- to 30-month-old children at an urban–rural distribution and determine the association of socioeconomic and demographic conditions. An observational cross-sectional study was conducted in 30 districts of Bangladesh. Three-stage cluster sampling was used with ‘Red Flag’, Modified Checklist for Toddlers and Diagnostic and Statistical Manual of Mental Disorders, 5th Edition, respectively. Data editing and analysis were done using CSPro 6 and SPSS 23. The survey included 37,982 households (71% rural, 29% urban) with 38,440 children. ‘Red Flag’ was positive in 209/10,000 children. Modified Checklist for Toddlers was positive in 149/10,000 children. Diagnostic and Statistical Manual of Mental Disorders, 5th Edition–positive autism spectrum disorder prevalence was 17/10,000 young children (boys 24/10,000, girls 9.8/10,000). Prevalence was higher in urban than rural (25/10,000 and 14/10,000 respectively). It was 77/10,000 in mothers aged 35–39 years and 23/10,000 in children of fathers aged 40 years. For families within the lowest wealth quintile, the autism spectrum disorder prevalence was 15/10,000. Autism spectrum disorder at very young ages in Bangladesh is still low. It was higher in urban areas, with the advanced age of parents, especially mothers, and in families with higher wealth quintiles.

Lay abstract

A nationwide survey was done in Bangladesh to assess autism spectrum disorder prevalence in 16- to 30-month-old children at urban–rural distribution and to determine the association with socioeconomic and demographic conditions. A three-stage cluster sampling method was used where districts from all divisions were selected in the first stage, census enumeration areas as blocks of households were selected in the second stage and households (within the blocks) were selected in the third stage. Thereby, it included 38,440 children from 37,982 households (71% rural, 29% urban) aged 16–30 months from 30 districts of eight divisions of Bangladesh. Screening was done with a ‘Red Flag’ tool and Modified Checklist for Toddlers and a final diagnosis using Diagnostic and Statistical Manual of Mental Disorders, 5th Edition for autism spectrum disorder. Autism spectrum disorder prevalence was 17 per 10,000 young children – in other words, one in 589 young children. Boys were found at higher risk of autism (one in 423 boys; one in 1026 girls). Prevalence of autism spectrum disorder was higher in urban environments than in rural ones – 25/10,000 and 14/10,000, respectively. More autism spectrum disorder children were found in advanced age groups of parents, especially mothers, and in households with a higher wealth quintile. This survey is significant as it covers both urban and rural areas and specifically targets very young children. The involvement of the Bangladesh Bureau of Statistics, as well as support from the entire healthcare system infrastructure, makes this survey more representative on a national level. Its results will form a database to support the development of an effective early intervention programme in Bangladesh. We hope it will prove useful for researchers, clinicians and frontline healthcare workers, and inform the decisions of policymakers and funders in Bangladesh.

Keywords

autism spectrum disorders pre-school children prevalence

Introduction

Autism spectrum disorder (ASD) is a developmental disorder that includes deficits in social communication and social interaction and the presence of restricted repetitive patterns of behaviour, interests, or activities that can persist throughout life (American Psychiatric Association, 2013). ASD and impairments associated with ASD often impose significant functional, social and financial impacts on affected individuals, their families and society (Ganz, 2007; Hoang et al., 2019; World Health Organization, 2018).

The prevalence of ASD has been increasing worldwide. It was reported as high as 168 per 10,000 or 1/59 in 8-year-old children in the United States in 2014 (Baio et al., 2018). The prevalence varies from 30 to 116/10,000 within the age range of birth to adulthood in Northern Europe (Elsabbagh et al., 2012). An overall prevalence of 36/10,000 with a higher prevalence in East Asia than in West Asia (51/10,000 and 35/10,000, respectively) was found in a meta-analysis of Asia (Qiu et al., 2019). A review study from South Asia, however, concluded that methodological differences in case definition, screening instruments and diagnostic criteria made it very difficult to compare the studies in this region (Hossain et al., 2017). In Bangladesh, in 2013, one survey reported the prevalence of autism in 0- to 9-year-old children was 15/10,000, with 6.8/10,000 in rural and 300/10,000 in urban areas (Non-Communicable Diseases Control (NCDC) Programme et al., 2013). A rural survey in 2016 reported autism in 7.5/10,000 cases among children aged 18–36 months (Akhter et al., 2018).

In recent years, a growing interest in the early detection of ASD has developed to implement early intervention (Dawson, 2008). Several longitudinal studies revealed that the majority of diagnoses made around the second birthday were stable during a re-evaluation of children at age 4 or older (Charman et al., 2005; Cox et al., 1999; Freeman & Cronin, 2002; Lord, 1995; Moore & Goodson, 2003; Stone et al., 1999). Intervention in the early years of life is associated with a more rapid and stronger response in learning language and social skills than the time of teaching these skills at a later age (Courchesne et al., 2011; Dawson, 2008; Lewis et al., 2014). Many researchers also believe feel that ASD is not necessarily a lifelong disabling condition (L. K. Koegel et al., 1999; R. L. Koegel & Koegel, 2012) and most research clinics reported that most of the children could be enrolled in regular education classrooms with intervention and around 25% of children would completely lose the diagnosis (Cohen et al., 2006; Helt et al., 2008; Lovaas, 1987; Sallows & Graupner, 2005).

This study was done to assess ASD prevalence in 16- to 30-month-old children in Bangladesh on a national level using the Bangladesh public health structure.

Methods

Survey design and population

This research was based on a cross-sectional observational survey, conducted in rural and urban areas of sampled districts nationwide to obtain the national prevalence of autism among children aged 16–30 months in Bangladesh. The survey was conducted in 85 primary sampling units (PSUs)/enumeration areas (EAs) distributed among 30 randomly selected districts of eight divisions of the country (Figure 1). Rural and urban areas were selected proportionately.

Figure 1.

Map of the study area.

The population covered in this survey was children aged 16–30 months. So, only households (HHs) with children aged 16–30 months were selected.

Sample size and sampling design

The required sample size (n) was determined by the usual sample size determination formula for estimating proportion, which is given below:

n = p (1 - p) {[\frac{z (α / 2)}{d}]}^{2} \times f

where p is an a priori proportion for autism in 8-year-old children in the United States in 2014 = 147/10,000 or 1.47%; z (α/2) is the critical value of the Normal distribution at α / 2 (for a confidence level of 95% and α of 0.05) = 1.96; d, allowing a maximum relative error of ±10%, = 0.0147 and f is the design effect = 1.5.

Thus,

n = 1.47 (1 − 0.0147) [1.96/0.0147]² × 1.5 = 38,624.

For operational feasibility, the sample size was rounded to 38,250.

For operational feasibility, the sample size was rounded to 38,440. The survey followed a three-stage cluster sampling method, where districts from all divisions were selected in the first stage. The census EAs as blocks of HHs were selected in the second stage and HHs (within the blocks) were selected in the third stage.

Training of survey staff

Three-day training of the trainers (ToT) was conducted with the participation of Medical Officers, Clinical and Education Psychologists, Instructors, Senior Teachers of Autism Schools and Therapists of the Institute where the training was conducted by the Project Director/Paediatric Neurologist, Deputy Project Director, Research Consultant and Research Medical Officer imparted the training.Then, a team of two Master Trainers subsequently was engaged for training 55–65 survey staff, including Medical Officers (MOs), Health Inspectors (HIs) and Health Assistants (HAs) for 2 days on the topics mentioned above at the respective district Civil Surgeon office. Each Master Trainer covered training in two sample districts one after another. After completing 2 days of training at the Civil Surgeons (CS) office, each survey team started enumeration in one of their assigned sample areas on the third day, where Master trainers, Quality control Officers (QCO) and Coordinators (MOs) closely monitored their field operation and made corrections where needed. Before conducting field-based training, all the assigned MOs (Coordinators & QCOs) were given 3 days of training on Autism & NDDs, including the survey questionnaire by the senior officers of the concerned Institute. The training of about 280 MOs in 30 districts was arranged in two phases before commencing field training and survey. A 1-day training of all the CS of sample districts (30 CS) was also done before field data collection, which helped them with effective coordination of the survey.

Data collection methods, instruments and screening of ASD

A survey team of three members (two Health Assistants as interviewers and one Health Inspector as team supervisor) was assigned to each PSU/EA to collect data from HH heads or mothers or caregivers of children aged 16–30 months. The HHs of each PSU/EA were updated by the survey team. To collect data, face-to-face interviews took place with a structured questionnaire.

The survey covered 85 PSUs/EAs in each district completing 2550 PSUs in 30 districts (30 × 85). There were 17 teams to cover 85 PSUs. A total 1530 survey staffs from 30 sample districts were assigned from the Directorate General Health Services (DGHS). Regarding the screening of ASD and identification of autism cases, the survey teams performed the first stage screening of ASD among the surveyed children using the ‘Red Flag’ tool (Autism Speaks, 2018). Trained MOs applied ‘Modified Check Lists for Toddlers’ ‘M-CHAT’ (Robins et al., 2004) on the referred ‘Red Flag’-positive children. An expert group of trained Neurologists, Medical Doctors or Psychologists performed a final diagnosis of ASD in the third stage using the Diagnostic and Statistical Manual for Mental Health, 5th Edition (DSM-5) for ASD (Kulage et al., 2014) (Figure 2).

Figure 2.

Flow chart of survey activities.

Survey period

The preparatory phase of the survey was between January and April 2017. The field data collection took place in two phases: first phase from May to June 2017 and the second phase from October to December 2017.

Supervision, coordination and quality assurance

Thirty (30) District Coordinators (MOs of Civil Surgeon Offices) and 250 Quality Control Officers (MOs of Upazila Health Complexes) acted as supervisors in the data collection operation, where 250 Upazila Health and Family Planning Officers (UHFPO) acted as Upazila Coordinators and 30 Civil Surgeons (CS) acted as District Chief Coordinators (DCCs).

To validate collected data, a ‘post enumeration check’ (PEC) was conducted both in rural and in urban areas at least in 2% of the surveyed HHs within a week after completion of the main survey.

Regarding validation of screening of ASD, an expert group of Medical Officers and Psychologists from Dhaka re-examined 10% of M-CHAT-negative cases (i.e. one in 10) using DSM-5.

Data editing and analysis

Manual editing and coding were performed as per the editing guidelines and code book. The data entry, cleaning and duplication detection were performed using CSPro 6. Data analysis and tabulation process were carried out using SPSS 23.

Ethical implications

Ethical clearance was obtained from the ethical committee of the Bangladesh Medical Research Council. Furthermore, prior to its commencement, approval for conducting the nationwide survey was obtained from the Bangladesh Bureau of Statistics (BBS).

Patient and public involvement

Children with ASD and without ASD and their families were involved in setting the research questions and the outcome measures. They were actively involved in field-level data collection and screening processes. Patients and their families were also central to the dissemination of the baseline information, which helped to motivate community involvement during and beyond the survey. Participants were all residents in a stratified last stage random sample of enumeration areas (the geographical area canvassed by one census representative) from the latest census. Before the baseline survey, permission was received from the authentic government authority (Ministry of Health and Family Welfare and Bangladesh Bureau of Statistics). Individual consent forms were read and signed by every respondent, all of whom were told that they were free to decline to answer any question. Parents or caregivers gave individual consent for each child who contributed to the screening process. All sorts of privacy for the data providers were maintained. Confidentiality of data was strictly ensured. Data were used only for the purpose of this survey and research.

Results

The survey was undertaken in 37,982 HHs in 30 districts distributed across eight divisions. House-to-house interviews, including ‘red flags’, were completed in 38,440 children aged 16–30 months. Subsequently, 804 were screened positive for ‘Red Flags’, 574 were M-CHAT-positive and 67 were DSM-5-positive (Table 1).

Table 1.

Distribution of districts, Upazila, HHs surveyed, and children screened positive by Red Flag, M-CHAT and diagnosed as ASD by DSM-5.

Sl. no.	District	No. of Upazila/Thana	Total HHs surveyed	No. of children (16–30 months)
Sl. no.	District	No. of Upazila/Thana	Total HHs surveyed	Total children surveyed	ASD screened by Red Flag	ASD screened by M-CHAT	ASD diagnosed by DSM-5
01	Dhaka	29	1080	1096	38	18	5
02	Narayanganj	5	1256	1271	60	27	2
03	Gazipur	5	1215	1230	21	17	1
04	Tangail	12	1167	1192	18	12	1
05	Kishorganj	13	1381	1400	52	22	8
06	Mymensingh	12	1176	1187	19	13	0
07	Chittagong	25	1431	1464	18	14	4
08	Feni	6	1425	1451	36	22	5
09	Comilla	16	1311	1330	24	14	1
10	Chandpur	8	1303	1315	18	13	1
11	Barisal	10	946	950	22	11	3
12	Jhalakati	4	1192	1203	36	31	2
13	Khulna	14	1192	1206	14	8	0
14	Satkhira	7	1283	1293	14	9	3
15	Bogra	12	1276	1288	31	26	1
16	Faridpur	9	1281	1297	22	17	4
17	Gaibandha	7	1277	1285	46	43	0
18	Gopalganj	5	1250	1267	17	13	2
19	Jessore	8	1248	1256	17	17	1
20	Kurigram	9	1283	1289	20	18	1
21	Kushtia	6	1285	1299	27	24	5
22	Naogaon	11	1303	1319	22	19	4
23	Pabna	9	1302	1316	46	30	4
24	Panchagarh	5	1302	1322	21	14	0
25	Rangamati	10	1194	1201	9	7	0
26	Shariatpur	6	1310	1326	17	15	1
27	Sirajganj	9	1365	1379	33	21	2
28	Sunamganj	11	1306	1330	25	22	2
29	Sylhet	12	1286	1306	26	25	4
30	Thakurgaon	5	1356	1372	35	32	0
Total		300	37,982	38,440	804	574	67

HHs: households; M-CHAT: Modified Checklist for Toddlers; ASD: autism spectrum disorder; DSM-5: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition.

Table 2 presents the general characteristics of the study sample. Notably, 71.1% of children lived in rural areas with 51.9% boys and 48.1% girls.

Table 2.

General characteristics of the study participants.

	Frequency	%
Area
Rural	27,315	71.1
Urban	11,125	28.9
Sex
Boys	19,953	51.9
Girls	18,487	48.1
Age (months)
16–20	13,762	35.8
21–25	13,300	34.6
26–30	11,378	29.6
Age of mothers
<20 (years)	7147	18.7
20–29	25,305	66.3
30–39	5411	14.2
40 and above	313	0.8
Age of the fathers
<20	216	0.5
20–29	16,861	44.2
30–39	17,024	44.6
40 and above	4075	10.7

a. Prevalence of ASD

Red Flag screen was positive in 209 per 10,000 young children (boys 242; girls 172). M-CHAT screen was positive in 149 per 10,000 young children (boys 174; girls 122).

ASD prevalence was 17.4 per 10,000 young children (boys 24/10,000, girls 9.8/10,000) – that means one in 589 young children (one in 423 boys; one in 1026 girls). Rural prevalence was 14/10,000 and urban was 25/10,000 (Table 3).

Table 3.

Distribution of screened positive and diagnosed ASD children by sex and area of residence using Red Flag, M-CHAT and DSM-5.

Area of residence and sex of children	Red Flag (positive)		M-CHAT (positive)		DSM-5 (positive)		Total no. of children
Area of residence and sex of children	n	Per 10,000	n	Per 10,000	n	Per 10,000	Total no. of children
Area
Rural	561	209.31	382	145.31	38	13.88	27,315
Urban	229	206.84	175	159.19	29	24.86	11,125
National	790	208.62	557	149.20	67	16.97	38,440
Sex of children
Boy	474	242.21	338	174.46	47	23.65	19,953
Girl	316	172.37	219	121.95	20	9.75	18,487
Total	790	208.62	557	149.20	67	16.97	38,440

ASD: autism spectrum disorder; M-CHAT: Modified Checklist for Toddlers; DSM-5: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition.

b. Associations of ASD

Age of children

ASD prevalence was 13/10,000 in the age group 16–20 months and this rose to 25/10,000 in the age group 26–30 months (Table 4).

Table 4.

Distribution of screened positive and diagnosed ASD children (by age range) using Red Flag, M-CHAT and DSM-5.

Age of children (month)	Per 10,000		No. of children
Age of children (month)	M-CHAT (positive)	DSM-5 (positive)	No. of children
16–20	137.57	12.65	13,713
21–25	150.60	14.71	13,263
26–30	161.61	24.81	11,464
Total	149.32	17.42	38,440

ASD: autism spectrum disorder; M-CHAT: Modified Checklist for Toddlers; DSM-5: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition.

Parental age

ASD prevalence was high in children of mothers aged 35–39 years (77/10,000) compared with that in children of mothers aged less than 20 years (14/10,000). Similarly, it was more prevalent in children of fathers aged 40 years and above (23/10,000), whereas it was 14/10,000 in children of fathers aged 20–24 (Table 5).

Table 5.

Distribution of ASD children by parent’s age (by DSM-5).

	Mother			Father
	n	Per 10,000	Total no. of mothers	n	Per 10,000	Total no. of fathers
Parental age (in years)
Less than 20	10	14.29	7147	0	0.00	216
20–24	15	9.92	15,694	6	14.25	5286
25–29	22	20.74	9611	16	12.86	11575
30–34	10	22.88	4191	21	17.73	10930
35–39	9	77.42	1220	14	22.16	6094
40 and above	0	0.00	313	9	23.18	4075
Total	66+	16.95	38176	66	16.95	38176

ASD: autism spectrum disorder; DSM-5: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition.

In the lowest wealth quintile, ASD prevalence was 15 per 10,000, while in the highest it was more than double this − 33 per 10, 000 (Table 6).

Table 6.

Distribution of ASD children by wealth quintile (by DSM-5).

Children by wealth quintile	n	Per 10,000	Total no. of children
Lowest	12	14.54	7658
Second	9	10.22	7689
Middle	8	11.53	7665
Fourth	12	15.90	7705
Highest	26	33.20	7723
Total	67	16.97	38,440

ASD: autism spectrum disorder; DSM-5: Diagnostic and Statistical Manual of Mental Disorders, 5th Edition.

Maternal healthcare in pregnancy

A higher ASD prevalence was found in children of mothers who did not receive any antenatal care (ANC) (22/10,000) compared with that in children of mothers who had ANC from trained personnel (received at least 6-month-long training) working in the field (13/10,000).

Prenatal maternal physical and/or mental health

ASD prevalence was 15/10,000 in children of mothers who experienced no physical or mental stresses in pregnancy but 44/10,000 in children whose mothers faced physical stresses, 42/10,000 in children of mothers who faced mental stresses and 52/10,000 in children of mothers who experienced both. The relationships were found to be statically significant at <0.5 levels determined using the chi-square test.

Discussion

Principal findings

Among 38,440 children, 804 (209/10,000) had early signs of ASD. Using M-CHAT, 574 were screened positive for ASD, and ASD was diagnosed in 67 (17 per 10,000) cases (Table 1). ASD was highest in the 26–30 months age group, followed by 21–25 months and 16–20 months (Table 3). Prevalence was higher in urban environments (Table 4). It was 5.5 times higher for children with 35–39 years old mothers than for those of <20 years. A higher prevalence was also found in fathers aged above 40 years compared with fathers in the 20–24 age group (Table 5). Families within the lowest wealth quintile had more than double the number of children with ASD than families within the highest quintile (Table 6), whereas ASD prevalence was 15/10,000 of children of mothers who experienced no physical or mental stress in pregnancy but 44/10,000 in children whose mothers faced physical stress, 42/10,000 in children of mothers who faced mental stresses and 52/10,000 in children of mothers who experienced both. The relationships were found to be statistically significant.

Strengths and limitations of this study

With the help of the Bangladesh Bureau of Statistics, this study covered a large and representative sample of the country which was a strength of the study. Support from the whole health system, including field-level staff who are familiar with house-to-house visits, was taken. The population sampled was a very young age group in which interventions would be more beneficial.

A limitation of the study was the age group, which was children aged 16–30 months, which could not represent ASD prevalence in all age groups of the country. Full clinical features of ASD might not appear at this low age range and the diagnosis might be difficult at this age. As there were 1700 data collectors, as well as 250 quality control medical officers, who were all part-time – working in regular government jobs, and sometimes transferred, standardization of the data collection faced limitations. In addition, in busy government health facilities, the location of M-CHAT assessments was not ideal. Furthermore, some areas like Kishoreganj were hard to reach because of natural calamities and the urban sampling was challenging because people were reluctant to allow access to their homes.

Comparison with other studies

The prevalence rate in this study was in agreement with the prevalence rate of the DSH 2013 survey which was 15.5/10,000 (NCDC Programme et al., 2013) and it is difficult to compare this rate because of different and limited sampling frames. The 2013 survey looked for autism in 0- to 9-year-old children and was conducted in seven Upazilas and one urban ward in eight districts, whereas this survey covered 300 Upazilas and 11,000 urban areas. The prevalence found in this study is comparable with a community study done in Northwest India (15/10,000) with an age group of 1–10 (Raina et al., 2017). Another study in South India found 15/10,000 in their 1- to 5-year age group (Poovathinal et al., 2016). Prevalence in this study is in disagreement with the very high findings of 168/10,000 in the US survey (Baio et al., 2018) and 264/10,000 in a South Korean survey (Kim et al., 2011). This might be due to the lower age range when full clinical features might not appear and cause diagnosis difficulties. Other important causes of variation were the methodological differences in the case definition and case-finding procedures and diagnostic criteria. Hoang et al. (2019) and Perera et al. (2009) found high levels of ASD at this young age, which was contrary to our findings (75/1000 and 105/10,000 respectively). This might be due to the use of Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV) instead of DSM-5 (Kulage et al., 2014; Wilson et al., 2013; Worley & Matson, 2012). Rural prevalence in this study (Table 4) was roughly double that of the 2016 IPNA rural survey (6.8/10,000) (Akhter et al., 2018) and 2013 DSH survey (7.5/10,000) (NCDC Programme et al., 2013). Comparisons here were inappropriate as the sampling frames were different and limited in both studies. The higher urban population in this survey was in agreement with a few other studies (Becker, 2010; Hoang et al., 2019; Lauritsen et al., 2014). Environment-related factors might contribute to this difference. Hazardous air pollutants usually concentrate in urban environments causing exposures during pregnancy and early childhood, which might have a potential association (Kalkbrenner et al., 2010; Windhamet al., 2006). The association of ASD with aged mothers (Table 5) was in agreement with the results of a meta-analysis by Sandin et al. (2016) where there was an association between advancing maternal age and the risk of autism. Many discussions with explanations were given regarding maternal age and chromosome anomalies (Sandin et al., 2016), variation and other genomic alterations, including epigenetic dysfuncions (Christian et al., 2008; Marshall et al., 2008), and the accumulated exposure to various toxins over time which could also result in genomic and epigenetic alterations in germ cells of older parents (Williams & Ross, 2007; Yauk et al., 2008). The association between ASD and paternal age (Table 5) was consistent with the finding of Reichenberg et al. (2006) where de novo mutations with advancing age or alterations in genetic imprinting were mentioned as the possible cause. Results regarding the wealth quintile on ASD were conflicting. Our findings (Table 6) agreed with some studies (Durkin et al., 2010; Thomas et al., 2012)and disagreed with others (Rai et al., 2012; Sun et al., 2014).

Conclusions

The prevalence of ASD in the very early years of life was still low in Bangladesh. ASD prevalence was higher in urban areas, with mothers of higher age groups and families in the lowest wealth quintiles. It was also higher in mothers who had physical and mental stress during the pregnancy period.

Recommendations

Further studies would be needed to elucidate the reasons for these findings. This survey might help in the development of an effective programme through which the government and NGOs could be able to work together. Early intervention could be started if these children with ASD could be identified at an early age. They could substantially contribute to society and the nation with proper training and education.

Footnotes

Acknowledgements

We would like to express our very great appreciation to the Ministry of Health and Family Welfare of Bangladesh, in particular, the Directorate General of Health Services to conduct this survey as part of the project ‘Establishment of Institute of Paediatric Neurodisorder and Autism’. We are thankful to the Bangladesh Bureau of Statistics, A House of Survey Research (SURCH) for their statistical support. We are also grateful to Psychiatrists from the National Institute of Mental Health, the Community Health Care Initiatives in Bangladesh, Civil surgeons of the selected 30 districts and their colleagues and the Bangladesh Medical Research Council. We would also like to extend our thanks to the Medical officers, Psychologists, Therapists, Teachers and all other staff of IPNA, BSMMU for conducting this door to door survey. Finally, we are grateful to all the parents of children who cooperated with their valuable time.

Data editing and analysis

Manual editing and coding were performed as per the editing guidelines and codebook. The data entry, cleaning, duplication detection was performed using CSPro 6. Data analysis and tabulation process were carried out using SPSS 23.

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.

Disclaimers

The views expressed in this article are our own and do not repreent the official positions of any institutions or funders.

Ethical implications

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This survey was funded by the Ministry of Health and Family Welfare, the Government of the People’s Republic of Bangladesh.

Patient and public involvement

ORCID iDs

Shaheen Akhter

Jannatara Shefa

References

Akhter

Hussain

A. H. M. E.

Shefa

Kundu

G. K.

Rahman

Biswas

(2018). Prevalence of autism spectrum disorder (ASD) among the children aged 18-36 months in a rural community of Bangladesh: A cross sectional study. F1000Research 7, 424.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders: DSM-5 (5th ed.). American Psychiatric Publishing.

Autism Speaks. (2018). Learn the signs of autism. https://www.autismspeaks.org/what-autism/learn-signs

Baio

Wiggins

Christensen

D. L.

Maenner

M. J.

Daniels

Warren

. . .Dowling

N. F.

(2018). Prevalence of autism spectrum disorder among children aged 8 years – Autism and Developmental Disabilities Monitoring Network, 11 sites, United States, 2014. MMWR Surveillance Summaries, 67(6), 1–23. https://doi.org/10.15585/mmwr.ss6706a1

Becker

K. G.

(2010). Autism and urbanization. American Journal of Public Health, 100(7), 1156–1157.

Charman

Taylor

Drew

Cockeri

L. L. H.

Brown

J. A.

Baird

(2005). Outcome at 7 years of children diagnosed with autism at age 2: Predictive validity of assessments conducted at 2 and 3 years of age and pattern of symptom change over time. Journal of Child Psychology and Psychiatry, 46(5), 500–513.

Christian

S. L.

Brune

C. W.

Sudi

Kumar

R. A.

Liu

Karamohamed

. . .Cook

E. H.

Jr. (2008). Novel submicroscopic chromosomal abnormalities detected in autism spectrum disorder. Biological Psychiatry, 63, 1111–1117.

Cohen

Amerine-Dickens

Smith

(2006). Early intensive behavioral treatment: Replication of the UCLA model in a community setting. Journal of Developmental and Behavioral Pediatrics, 27, 145–155.

Courchesne

Campbell

Solso

(2011). Brain growth across the life span in autism: Age-specific changes in anatomical pathology. Brain Research, 1380, 138–145.

10.

Cox

Klein

Charman

Baird

Baron-Cohen

Swettenham

. . .Wheelwright

(1999). Autism spectrum disorders at 20 and 42 months of age: Stability of clinical and ADI-R diagnosis. Journal of Child Psychology and Psychiatry and Allied Disciplines, 40(5), 719–732.

11.

Dawson

(2008). Early behavioral intervention, brain plasticity, and the prevention of autism spectrum disorder. Development and Psychopathology, 20, 775–803.

12.

Durkin

M. S.

Maenner

M. J.

Meaney

F. J.

Levy

S. E.

DiGuiseppi

Nicholas

J. S.

. . .Schieve

L. A.

(2010). Socioeconomic inequality in the prevalence of autism spectrum disorder: Evidence from a U.S. cross-sectional study. PLOS ONE, 5(7), Article e11551. https://doi.org/10.1371/journal.pone.0011551

13.

Elsabbagh

Divan

Koh

Y. J.

Kim

Y. S.

Kauchali

Marcin

. . .Fombonne

(2012). Global prevalence of autism and other pervasive developmental disorders. Autism Research, 5(3), 160–179.

14.

Freeman

B. J.

Cronin

(2002). Diagnosing autism spectrum disorder in young children: An update. Infants and Young Children, 14(3), 1–10.

15.

Ganz

M. L.

(2007). The lifetime distribution of the incremental societal costs of autism. Archives of Pediatrics & Adolescent Medicine, 161(4), 343–349.

16.

Helt

Kelley

Kinsbourne

Pandey

Boorstein

Herbert

Fein

(2008). Can children with autism recover? If so, how? Neuropsychology Review, 18, 339–366.

17.

Hoang

V. M.

T. V.

Chu

T. T. Q.

B. N.

Duong

M. D.

Thanh

N. M.

. . .Bui

T. T. H.

(2019). Prevalence of autism spectrum disorders and their relation to selected socio-demographic factors among children aged 18–30 months in northern Vietnam. International Journal of Mental Health Systems, 13, 29.

18.

Hossain

M. D.

Ahmed

H. U.

Jalal Uddin

M. M.

Chowdhury

W. A.

Iqbal

M. S.

Kabir

R. I.

. . .Sarker

(2017). Autism spectrum disorders (ASD) in South Asia: A systematic review. BMC Psychiatry, 17(1), Article 281.

19.

Kalkbrenner

A. E.

Daniels

J. L.

Chen

J. C.

Poole

Emch

Morrissey

(2010). Perinatal exposure to hazardous air pollutants and autism spectrum disorders at age 8. Epidemiology, 21(5), 631–641.

20.

Kim

Y. S.

Leventhal

B. L.

Koh

Fombonne

Laska

Lim

E. C.

. . .Grinker

R. R.

(2011). Prevalence of autism spectrum disorders in a total population sample. American Journal of Psychiatry, 168, 904–912.

21.

Koegel

L. K.

Koegel

R. L.

Shoshan

McNerney

(1999). Pivotal response intervention II: Preliminary long-term outcomes data. Journal of the Association for Persons With Severe Handicaps, 24, 186–198.

22.

Koegel

R. L.

Koegel

L. K.

(2012). The PRT pocket guide: Pivotal response treatment for autism spectrum disorder. Paul H. Brooks Publishing.

23.

Kulage

K. M.

Smaldone

A. M.

Cohn

E. G.

(2014). How will DSM-5 affect autism diagnosis? A systematic literature review and meta-analysis. Journal of Autism and Developmental Disorders, 44(8), 1918–1932.

24.

Lauritsen

M. B.

Astrup

Pedersen

C. B.

Obel

Schendel

D. E.

Schieve

. . .Parner

E. T.

(2014). Urbanicity and autism spectrum disorders. Journal of Autism and Developmental Disorders, 44(2), 394–404.

25.

Lewis

J. D.

Evans

A. C.

Pruett

J. R.

Botteron

Zwaigenbaum

Estes

. . .Piven

(2014). Network inefficiencies in autism spectrum disorder at 24 months. Translational Psychiatry, 4, Article e388.

26.

Lord

(1995). Follow-up of two-year-olds referred for possible autism. Journal of Child Psychology and Psychiatry and Allied Disciplines, 36(8), 1365–1382.

27.

Lovaas

O. I.

(1987). Behavioral treatment and normal educational and intellectual functioning in young autistic children. Journal of Consulting and Clinical Psychology, 55, 3–9.

28.

Marshall

C. R.

Noor

Vincent

J. B.

Lionel

A. C.

Feuk

Skaug

. . .Scherer

S. W.

(2008). Structural variation of chromosomes in autism spectrum disorder. The American Journal of Human Genetics, 82, 477–488.

29.

Moore

Goodson

(2003). How well does early diagnosis of autism stand the test of time? Follow-up study of children assessed for autism at age 2 and development of an early diagnostic service. Autism, 7(1), 47–63.

30.

Non-Communicable Diseases Control Programme, Revitalization of Community Health Care Initiatives in Bangladesh, Bangladesh Medical Research Council, & Department of Pediatric Neuroscience, Dhaka Shishu Hospital. (2013). Survey of autism and neurodevelopmental disorders in Bangladesh. http://www.dghs.gov.bd/images/docs/Autism/Survey%20of%20Autism%20-%20SKB-0403.pdf

31.

Perera

Wijewardena

Aluthwelage

(2009). Screening of 18–24-month-old children for autism in a semi-urban community in Sri Lanka. Journal of Tropical Pediatrics, 55(6), 402–405. https://doi.org/10.1093/tropej/fmp031

32.

Poovathinal

S. A.

Anitha

Thomas

Kaniamattam

Melempatt

Anilkumar

Meena

(2016). Prevalence of autism spectrum disorders in a semiurban community in south India. Annals of Epidemiology, 26(9), 663–665.

33.

Qiu

Shi

Cui

. . .Qiao

(2019). Prevalence of autism spectrum disorder in Asia: A systematic review and meta-analysis. Psychiatry Research, 284, 112679. https://doi.org/10.1016/j.psychres.112679

34.

Rai

Lewis

Lundberg

Araya

Svensson

Dalman

(2012). Parental socioeconomic status and risk of offspring autism spectrum disorders in a Swedish population-based study. Journal of the American Academy of Child & Adolescent Psychiatry, 51(5), 467–476.

35.

Raina

S. K.

Chander

Bhardwaj

A. K.

Kumar

Sharma

Kashyap

. . .Bhardwaj

(2017). Prevalence of autism spectrum disorder among rural, urban, and children (1-10 years of age). Journal of Neurosciences in Rural Practice, 8(3), 368–374.

36.

Reichenberg

Gross

Weiser

Bresnahan

Silverman

Harlap

. . .Susser

(2006). Advancing paternal age and autism. Archives of General Psychiatry, 63, 1026–1032.

37.

Robins

D. L.

Casagrande

Barton

Chen

C. M.

Dumont-Mathieu

Fein

(2004). Validation of the modified checklist for Autism in toddlers, revised with follow-up (M-CHAT-R/F). Pediatrics, 133(1), 37–45.

38.

Sallows

G. O.

Graupner

T. D.

(2005). Intensive behavioral treatment for children with autism: Four-year outcome and predictors. American Journal on Mental Retardation, 110, 417–443.

39.

Sandin

Schendel

Magnusson

Hultman

Surén

Susser

. . .Reichenberg

(2016). Autism risk associated with parental age and with increasing difference in age between the parents. Molecular Psychiatry, 21(5), 693–700. https://doi.org/10.1038/mp.2015.70

40.

Stone

W. L.

Lee

E. B.

Ashfor

Brissie

Hepburn

S. L.

Coonrod

E. E.

Weiss

B. H.

(1999). Can autism be diagnosed accurately in children under 3 years? Journal of Child Psychology and Psychiatry and Allied Disciplines, 40(2), 219–226.

41.

Sun

Allison

Auyeung

Cohen

S.-B.

Brayne

(2014). Parental concerns, socioeconomic status, and the risk of autism spectrum conditions in a population-based study. Research in Developmental Disabilities, 35, 3678–3688.

42.

Thomas

Zahorodny

Peng

Kim

Jani

Halperin

Brimacombe

(2012). The association of autism diagnosis with socioeconomic status. Autism, 16(2), 201–213. https://doi.org/10.1177/1362361311413397

43.

Williams

J. H. G.

Ross

(2007). Consequences of prenatal toxin exposure for mental health in children and adolescents: A systematic review. European Child & Adolescent Psychiatry, 16, 243–253.

44.

Wilson

C. E.

Gillan

Spain

Robertson

Roberts

Murphy

C. M.

. . .Murphy

D. G.

(2013). Comparison of ICD-10R, DSM-IV-TR and DSM-5 in an adult autism spectrum disorder diagnostic clinic. Journal of Autism and Developmental Disorders, 43(11), 2515–2525.

45.

Windham

G. C.

Zhang

Gunier

Croen

L. A.

Grether

J. K.

(2006). Autism spectrum disorders in relation to distribution of hazardous air pollutants in the San Francisco Bay area. Environmental Health Perspectives, 114(9), 1438–1444.

46.

World Health Organization. (2018). Factsheet: Autism spectrum disorders. https://www.who.int/news-room/fact-sheets/detail/autism-spectrum-disorders

47.

Worley

J. A.

Matson

J. L.

(2012). Comparing symptoms of autism spectrum disorders using the current DSM-IV-TR diagnostic criteria and the proposed DSM-V diagnostic criteria. Research in Autism Spectrum Disorders, 6(2), 965–970.

48.

Yauk

Polyzos

Rowan-Carroll

Somers

C. M.

Godschalk

R. W.

Van Schooten

F. J.

. . .Kovalchuk

(2008). Germ-line mutations, DNA damage, and global hypermethylation in mice exposed to particulate air pollution in an urban/industrial location. Proceedings of the National Academy of Sciences of the United States of America, 105, 605–610.

Autism spectrum disorder among 16- to 30-month-old children in Bangladesh: Observational cross-sectional study

Abstract

Lay abstract

Keywords

Introduction

Methods

Survey design and population

Sample size and sampling design

Training of survey staff

Data collection methods, instruments and screening of ASD

Survey period

Supervision, coordination and quality assurance

Data editing and analysis

Ethical implications

Patient and public involvement

Results

Age of children

Parental age

Maternal healthcare in pregnancy

Prenatal maternal physical and/or mental health

Discussion

Principal findings

Strengths and limitations of this study

Comparison with other studies

Conclusions

Recommendations

Footnotes

Acknowledgements

Data editing and analysis

Declaration of conflicting interests

Disclaimers

Ethical implications

Funding

Patient and public involvement

ORCID iDs

References