Sudden Infant Death Syndrome in Relation to Weather and Optimetrically Measured Air Pollution in Taiwan

(1)

Sudden

Infant

Death

Syndrome

in Relation

to Weather

and

Optimetrically

Measured

Air

Pollution

in Taiwan

Harald

H. Kn#{246}bel,MD*;

Chien-Jen

Chen,

ScD*;

and

Kung-Yee

Liang,

PhD

ABSTRACT.

Objective.

To

examine

the

possible

role

of

weather

and

air

pollution

in

sudden

infant

death

syndrome

(SIDS)

and

suffocation.

Methods.

Poisson

regression

analysis

was

carried

out

to measure

the

association

between

daily

rates

of SIDS

per

1000

live

births

and

daily

average

values

of visibility

and

temperature

in Taiwan

between

1981

and 1991.

The

optimetrical

measure

of air

pollution

was

used

to

repre-sent

pollution

over

a whole

area

rather

than

at a point

source.

Results.

Mortality

from

SIDS

per

1000

live

births

was

3.3 times

greater

in the

lowest

incidence

of deaths

from

respira-tory

tract

infections,

temperature,

air pressure,

sunshine,

rainfall,

relative

humidity,

and

windspeed

increased

these

rate ratios

to 3.8

and

5.1, respectively.

This

suggests

that

the

relationship

between

air

pollution

and

511)5 is

not

biased

by

ecological

confounders.

For

temperature

the

rate

ratios

were

between

3.3

and

4.0.

Conclusions.

Our

findings

confirm

the

association

of

climatic

temperature

and

air

pollution

with

511)5.

Pediatrics

199596:1106-1110;

air pollution,

climatic

tern-perature, SIDS, suffocation.

ABBREVIATIONS. SIDS, sudden infant death syndrome; PMIcC,

weight of particles of diameter less than 10 pm/m3; PSI, pollutants standard index.

varied

in their

methods

and

findings,

as well

as

in

their

conclusions.8#{176}

Taiwan

has

experienced

a

consistent

reduction

of infant

mortality

in recent

de-cades,

whereas

511)5

and

suffocation

have

been

ris-ing durris-ing

the same

time

period.”

This

has

coincided

with

an equally

consistent

rise

in air pollution.’2

The

regional

distribution

of 511)5

has

shown

low

rates

in

From the *Imtitute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan, Republic of China; the Institute of Epidemiology, School of Public Health, National Taiwan University, Taipei, Taiwan, Republic of China, and the §Departinent of Biostatistics, School of Hygiene and Public Health, Johns Hopkins University, Baltimore, MD.

Received for publication May 16, 1994; accepted Nov 22, 1994.

Address correspondence to: Harald H. Kn#{246}bel,MD, ISSP Academia Sinica,

Taipei 11529, Taiwan, Republic of China.

the

undeveloped

unpolluted

regions,

while

general

infant

mortality

has

been

high

in those

regions

and

low

in

the

highly

urbanized

areas.11

We

therefore

decided

to investigate

the

relationship

between

SIDS

and

air pollution

in Taiwan.

Together

SIDS

and

suffocation

account

now

for

20%

of

the

total

infant

mortality

in

Taiwan,

repre-senting

a

yearly

rate

of close

to I

per

1000

live

births,

a figure

similar

to infant

mortality

from

5105

alone

in

western

countries.1’15

511)5

as a distinct

diagnosis

has

recently

been

introduced

because

suffocation

seemed

an

inappropriate

diagnosis

for

the

majority

of sudden

unexplained

infant

deaths.

Suffocation

as

cause

of

death

is

now

rarely

recorded

in

western

countries.16”7

The

situation

in Asia

is essentially

dif-ferent.

Taiwan

and

Japan

both

record

suffocation

rates

that

are

higher

than

SIDS

rates,

and

together

these

two

causes

of death

sum

up to values

similar

to

those

for

511)5

in

western

countries.’1’17

We

include

accidental

suffocation

diagnoses

in 511)5,

taking

the

view

that

from

a western

perspective

both

terms

would

be

synonymous.

In the

absence

of gravimetrically

measured

data

on

individual

pollutants

for

the

whole

island

of Taiwan

for

several

years

we

chose

visibility

as measured

at

all

the

meteorological

stations

throughout

the

island

as the

best

available

variable

to indicate

the

degree

of

air

pollution.

Visibifity

in our

study

is measured

as

visual

range,

which

makes

the

measurements

repre-sentative

for

areas,

whereas

aerosol

extinction

coef-ficients

from

photometers

are

point

measurements

like

the

gravimetrical

measures.

Recently

the

close

relationship

between

visual

range,

aerosol

extinction

coefficients,

and

total

suspended

particulates

(TSP)

has

been

reconfirmed.4’18

Emissions

from

metropoli-tan

areas

are

known

to

cause

the

production

of

visibffity-reducing

aerosols

from

160

to

240

km

downwind,long

after

the

primary

precursors

have

been

diluted

to

low

concentrations.19

The

aerosols

that

have

the

most

effect

on reducing

visibility

are

in

the

diameter

range

from

0.2

to

2 tim,

those

inhaled

most

deeply

into

the

lungs.192#{176} Aerosols,

like

NH4HSO4

and

(NH4)2SO4

are

usually

not

measured

in

air

samplers,

but

are

important

pollutants

that

constitute

up

to

50%

of

submicrorneter

particles.2’

These

aerosols

induce

light

scattering,21’

and

thus

are

integratively

measured

with

visibifity.

As

visibifity

and

also

511)5

mortality

may

be influenced

in

part

by

ecological

confounders

such

as

temperature,

air

pres-sure,

sunshine

duration,

rainfall,

relative

humidity,

and

windspeed,

those

values

were

induded

in our

analysis.

at Viet Nam:AAP Sponsored on September 1, 2020

www.aappublications.org/news

(2)

Year

ARTICLES

1107

METHODS

Data from death certificates for all deaths of infants in Taiwan

above I week and below I year of age were made available to us

for the years 1981 to 1991 by the Taiwan Center for Vital Statistics. The underlying causes of death

are

assigned

according to the ninth revision of the International Classification of Diseases?’ We

studied deaths coded as SIDS (ICD9 798.0, in one year also coded

as 799.0), which accounted for 38% of the cases, and those coded as suffocation (E911, E912, and E913); 62% of the cases were subsumed under SIDS. As a potential confounder respiratory tract

infections (ICD9 460 to 466 and 480 to 487) were used in the

regression analysis, because they are believed to be a causal factor in SIDS.124

A forensic doctor examined 42% of SIDS cases, a coroner 15%, and the remaining death certificates were issued by a physician. A

forensic investigation by a forensic doctor or coroner indudes an examination of the body, inspection of the place of death, and an

interview with the family. These investigations are perfonned in addition to the primary examination by a physician. An autopsy is

performed only if an infanticide cannot be excluded, in about 3%

of the cases.n fr 0.5% of SIDS cases the status of the issuer of the

death

certificate was unknown, or the issuer was not a physician; therefore these cases were excluded from the analysis.

Data on weather were recorded by the Central Weather Bureau of the Ministry of Transport and Communications. Nine stations

located in the center of each of the nine main regions on the island were chosen to represent each region (in the dties of Taipei,

Hsinchu, Taichung, Chiayi, Tainan, Kaohsiung, Taitung, Hualien,

and ilan). On average an area of 2200 square km was covered by each weather station. The island groups of Penghu, Kinmen, Matsu, Lanyu, and Lutao as well as the central mountainous areas were exduded from the analysis because either the population was too small for the time series analysis to be performed or

because no weather station representative for the area was

avail-able. Daily mean values for all days in the years 1981 to 1991 of visibility, temperature, air pressure, rainfall, sunshine hours, rel-ative humidity, and windspeed were used. Visibility is defined as the furthest distance from which an object (or, at nighttime, a

diffuse light source) can be recognized with the unaided eye.

Sixteen measurements were made at intervals of 22.5#{176},in a circle

around the observing position. The visibility observed from

mea-surements over at least 202.5#{176}sectors was used. Therefore at

conditions of low visibility of 4.1 km (5% quantile of visibility

during the study period) the measurement is representative of an area of 30 j2; at mean visibifity of 11.6 km it represents 238 km2; at high visibility of 22.3 km (95% quantile) it represents 880 km2.

The measurements were repeated every 3 hours and followed a

precise standardized protocol.26

Daily gravimetrica! measures of air pollution [weight of parti-des of diameter <10 pm/m’ (PM1O), pollutants standard index

(PSI),

sulphur dioxide, and carbon monoxide] were made

avail-able from the Environmental Protection Agency for 1990 and 1991 for seven regions (Taipei, Hsinchu, Taichung, Chiayi, Tainan, Kaohsiung, and Hualien). Gravimetrical measures, optimetrica! measures, and weather variables were correlated using Pearson and Spearman correlation coefficients to assess inferences. In

Taipei and Kaohsiung areas the mean daily values of the five

available stations were used for the correlation analysis, the other

regions had only one station. To make weather and pollution correlations comparable, only the 2 years for which data were also

available for gravimetrically measured pollutants were used.

Each region throughout the study period was then divided

according to the daily values of visibility and temperature. To link several regions for a combined analysis we divided the number of deaths occurring on each day in each category by the daily num-ber of live births, which was computed by dividing the monthly

number of live births by the number of days in each month. To

control for seasonality ofbirths, the number of live births 3 months before death (3 months being the peak age of SIDS and

suffoca-tion) was taken as denominator. For the computation of crude rates within each category, the total number of deaths was divided by the total number of live births.

As there is no consensus about the time between disease onset

and death from SIDS, we analyzed our data under two different assumptions. First, a time lag of no more than several hours was assumed. Second, we allowed for a time lag of several days

between the hypothetical disease onset and death.2 Under this

second hypothesis the visibility and weather values on the 9 days before death are considered to be related to the incidence of SIDS. Poisson regression analysis was performed to control for po-tential confounders. For this analysis the natural

logarithm

of the daily rate of SIDS per 1000 live births was explained by visibifity

and temperature. The daily number of live births was included to control for the different population sizes of the regions and

sea-sonality of births; the month of death to control for seasonality of

deaths; the proportion of births in each level of urbanization (the

levels were cities, towns, and rural townships) to control for the different development state and the different social composition of the regions; the daily rate of respiratory tract infections to control for their possible influence on SIDS, temperature, air pressure,

sunshine

hours,

rainfall,

relative humidity, and windspeed to control for their potential influence on visibility. All statistical analysis was performed with the Statistical Analysis System (SAS

Institute, Cary, NC).

RESULTS

Rates

of SIDS

consistently

rose

throughout

the

11

years

of the

study

period

(Fig

1),

in contrast

to

gen-eral

infant

mortality

which

decreased

from

8.9

per

1000

live

births

in 1981

to 4.8

per

1000

live

births

in

1991.11

Among

the

weather

variables

investigated,

only

visibility

had

a clear

secular

trend

with

consis-tently

decreasing

values,

mirroring

the

trend

of 511)5

(Fig

1).

The

seasonal

pattern

of

511)5

showed

the

highest

rate

in

January,

the

coldest

month,

with

an

additional

peak

in March

(Fig

2). Weather

variables

varied

widely

with

the

season.

Similar

to

tempera-ture

variations

were

the

seasonal

variations

of

sun-shine

duration

and

most

notably,

air

pressure

(data

not

shown).

Visibility

was

lowest

and

inversely

cor-related

with

the

high

rate

of SIDS

in March

(Fig

2).

Rainfall

was

highest

during

the

summer.

Average

windspeed

showed

little

seasonal

variation.

In

a

combined

regression

analysis

only

visibility

and

temperature

were

consistently

significantly

rates.

The

influence

of weather

variables

on

visibifity

as well

as the

relationship

of gravimetrical

air pollution

measures

to visibifity

was

assessed

via

correlation

coefficients

for

the

years

1990

and

1991

(Table

1). Data

were

available

for seven

regions,

for a

total

of

5028

observations.

Temperature

and

sun-shine

hours

were

significantly

correlated

to visibility,

with

correlation

coefficients

(Pearson/Spearman)

of

Fig 1.

Yearly trends of visibility and of mortality from SIDS in

Taiwan 1981 to 1991.

(3)

(I)

.c

0 >

a

(I)

CI)

E

11 .

10

Cl)

>.

0

Jan

Feb Mar Apr May Jun Jul Aug Sep Oct Nov

Dec

Month

Fig

2. Monthly trends of visibility and mortality from SIDS in Taiwan 1981 to 1991.

TABLE 1. Correlation of Optimetrically Measured Air Pollu-tion to Gravimetrically Measured Air Pollution and Weather in Taiwan

PM10 Pearson!

Spearman*

Visibility

Pearson/Spearman

PSI 0.97 0.97 -0.60 -0.62

PM10

1.00 1.00 -0.59 -0.61

Sulfur

dioxide

0.51

0.53

-0.40

-0.41

Carbon monoxide 0.44 0.48 -0.46 -0.53

Visibthty -0.59 -0.61 1.00 1.00

Temperature -0.02 -0.03 O.38 0.39

Air/Pressure 0.06 0.05 -0.19 -0.22

Rainfall -0.16 -0.27 -0.08 -0.06

Relative humidity -0.13 -0.15. -0.17 -0.19

Sunshine hours 0.12 0.13 0.29 0.26

Windspeed -0.12 -0.07 -0.03 -0.07

Abbreviations: PM10, weight of particles of diameter <10 pm/m3;

PSI, pollutants

standard

index.

Daily

average

values from seven regions for 1990 and 1991, in total

5028 observations.

* Correlation index.

§P

< .0001.

0.38/0.39

and

0.29/0.26,

respectively.

Visibility

was

highly

correlated

to PM10

(coefficients

0.59/0.61)

and

PSI

(0.60/0.62).

This

was

higher

than

the

correlation

of

PM10

to

sulphur

dioxide

and

carbon

monoxide.

The

correlation

of

visibility

to

sulphur

dioxide

was

only

slightly

less,

and

of visibility

to carbon

monox-ide

was

higher

than

the

correlation

of PM10

to those

two

single

pollutants

(Table

1).

Visibility

and

temperature

were

divided

into

cat-egories

and

tested

for

consistency

of

their

effect

throughout

the

whole

range

of

values.

The

ratio

of

the

crude

rate

of SIDS

per

1000

live

births

from

the

highest

to the lowest

before

death

(Table

2). There

was

a consistent

trend

of increasing

SIDS

from

the

highest

to

the

lowest

visibility

cate-gory

(P

<

.01,

based

on

test

for

trend).

In

the

regression

model

where

population

size,

season,

de-gree

of

urbanization,

the

incidence

of

respiratory

tract

infections,

and

the

other

weather

variables

were

controlled,

the

effect

of

visibility

was

even

stronger

TABLE 2. Postperinatal SIDS Rate by Daily Average Visibility (in han) on Day of Death and During the 9 Days Before Death

SIDS, Visibility on Day of Death

Visibility, No. of Crude Crude RR Adjusted RR

(lan) Births* Ratet (95% CI) (95% CI)

1-3 183 1.14 3.3 (2.0-5.2) 3.8(2.8-5.1)

4-9 2117 0.84 2.4 (1.5-3.8) 2.9(22-3.9)

10-15 1130 0.73 2.1 (1.3-3.3) 2.7(2.0-3.7) 16-21 326 0.53 1.5 (0.9-2.4) 2.0(1.5-2.7)

22-37

60

0.35 1.0 1.0

SIDS, Visibility I to 9 Days Before Death

Visibility, (han) No. of Crude Crude RR Adjusted RR Births* Ratet (95%

CI)

(95% CI) 1-3 33 1.20 3.4 (1.5-7.9) 5.1 (3.2-8.1)

4-9 2318 0.89 2.6 (1.2-5.5) 3.6 (2.4-5.5)

10-15 1215 0.65 1.9 (0.9-4.0) 3.2(2.1-4.9)

16-21

225 0.43 1.2(0.6-2.7) 1.7 (1.1-2.6)

22-37 24 0.35 1.0 1.0

Abbreviations: SIDS, sudden infant

death

syndrome;

RR, rate

ratio; CI, confidence interval.

“in l000s.

t Deaths from SItS per 1000 live births.

Controlling for daily number of live births, month, daily average

temperature, air pressure, sunshine, rainfall, relative humidity, windspeed, daily

incidence

of

respiratory tract infections, and

level of urbanization.

and

equally

consistent.

The

adjusted

rate

ratio

was

3.8

for

the

day

of

death

and

5.1

for

the

average

visibility

on

the

9 days

before

death

(Table

2).

The

rate

ratio

of 511)5

per

1000

live

births

from

the

high-est

to

the

lowest

temperature

for

the

day

of death

was

4.0

and

remained

at 3.3

after

adjustment

(Table

3).

For

the

temperature

on

the

9 days

before

death

the

crude

rate

ratio

was

4.0

and

the

adjusted

rate

ratio

was

3.5.

The

other

weather

variables,

air

pres-sure,

sunshine

duration,

rainfall,

relative

humidity,

and

windspeed

were

either

not

consistent

in

the

TABLE 3. Postperinatal SIDS Rate by Daily Average

Temper-ature in Degrees Celsius on Day of Death and During

the

9 Days Before Death

SIDS, Temperature on Day of Death

Temperature No. of Crude Crude RR Adjusted RR Births” Ratet (95% CI) (95% CI)

7-13 179 1.63 4.0 (3.3-4.8) 3.3(2.7-4.1)

14-18

827

1.21 3.0 (2.5-3.5) 2.4(2.1-2.9)

19-23 938 0.88 2.1 (1.8-2.5) 1.8 (1.6-2.1) 24-28 1433 0.49 1.2 (1.0-1.4) 1.1 (1.0-1.3)

29-33 438 0.41 1.0 1.0

SIDS, Temperature I to 9 Days Before Death

Temperature No. of Crude Crude RR Adjusted RR

Births” Ratet (95% CI) (95% CI)

7-13 77 1.77 4.0 (3.1-5.1) 3.5(2.8-4.5)

14-18 958 1.33 3.0 (2.5-3.6) 2.4(2.0-2.8) 19-23 935 0.84 1.9 (1.6-2.2) 1.6(1.4-1.9)

24-28 1515 0.45 1.0 (0.8-1.2) 0.9(0.8-1.1)

29-33 329 0.45 1.0 1.0

Abbreviations: SIDS, sudden infant death syndrome; RR, rate ratio; CI, confidence interval.

“in l000s.

1Deaths from SIDS per 1000 live births.

§

Controlling for daily number of live births, month, daily average

visibility, air pressure, sunshine, rainfall, relative humidity, daily incidence of respiratory tract infections. and level of urbanization.

at Viet Nam:AAP Sponsored on September 1, 2020

(4)

ARTICLES 1109

direction

of their

effect

from

single

to multivariate

analysis

or became

nonsignificant

in the

multivariate

analysis

(data

not

shown).

Therefore

they

were

not

believed

to

have

an

independent

important

effect

and

were

used

only

for

adjustment

of visibility

and

temperature.

511)5 mortality

was

low

in conditions

of

clear

air

and

temperatures

above

24#{176}C,

and

high

in

conditions

of low

visibifity

and

low

temperature.

Air

pollution,

measured

optimetrically,

had

a

con-sistent

relationship

to SIDS

that

was

as strong

and

significant

as the relationship

between

SIDS

and

tem-perature.

Other

weather

variables,

including

air

pressure,

sunshine

duration,

rainfall,

relative

humid-ity,

and

windspeed

were

inconsistent

in

their

rela-tion

to 511)5

and

seemed

not

to be independent

from

each

other.

The

values

of

visibility

on

the

9 days

before

death

had

a slightly

better

predictive

power

for

5105

than

the

values

on

the

day

of death.

DISCUSSION

Temperature

has

been

independently

and

consis-tently

identified

as a risk

factor

for

SIDS

in different

parts

of the

world.3

Paradoxical

overheating

is the

most

frequently

suggested

mechanism,2”#{176}

but

low

temperature

may

also

increase

the

effect

of

sub-stances

like

pollutants

on

the

respiratory

system.

Also

there

is a higher

prevalence

of respiratory

in-fections

in

colder

seasons,

and

a multiplication

of

risk

when

overbedding

and

infection

are

combined.24

Concern

has

also

been

expressed

about

how

closely

climatic

temperature

and

room

temperature

corre-late.3’

As

heating

is

still

uncommon

in

Taiwanese

households,

inside

and

outside

temperatures

are

closer

than

in countries

where

heating

is the

norm.

Recently,

air

pollution

has

been

studied

primarily

on

the

basis

of

gravimetrically

measured

data

that

offer

detailed

information

on

single

pollutants.8’9

Such

data

may

vary

with

the

distance

of the

station

from

strong

individual

polluters,

and

may

under-

or

overestimate

the

real

exposure

of

people

in

a given

area.

Optimetrical

data

give

an

average

reading

that

is more

generalizable,

but

may

be

more

influenced

by

weather

and

climate

than

gravimetrical

readings.

To

adjust

for

that

influence

we

have

included

a

va-riety

of weather

data

into

our

analysis.

Taiwan

has

a

wide

range

from

heavy

to

very

low

pollution,’2

a

mild

climate

ranging

from

subtropical

to

tropical,

and

a low

prevalence

of certain

maternal

risk

factors

(eg,

smoking),

and

is thus

a suitable

place

to study

the

effects

of

air

pollution

on

511)5.

Studies

on

air

pollution

and

511)5

have

used

a

wide

variety

of

methodologies

with

conflicting

results.

Some

studies

that

failed

to show

a correlation

have

been

carried

out

in areas

with

a very

narrow

band

of air pollution

values.9”0

The

results

of

our

study

are

presented

as

population-based

rates

that

allow

easy

comparison

among

areas.

Our

findings

suggest

that

air pollution

is

linked

to

SIDS,

and

thus

reopen

the

discussion.

The

likelihood

that

SIDS

is not

a homogeneous

cause

of death,

but

may

consist

of

etiologically

different

subgroups

should

lead

to an

underestimation

of

ef-fect.

Variation

in the

different

social

composition

of

the

regions

was

controlled

by

including

level

of

ur-banization

in

the

regression

analysis.

Urbanization

level

reflects

the

differences

in

the

three

variables

that

are

usually

used

to measure

social

class-edu-cation,

income,

and

occupation.

Higher

level

of

ur-banization

is associated

with

a greater

proportion

of

the

population

that

has

achieved

higher

or secondary

education,

higher

income,

more

prestigious

occupa-tion,

lower

total

infant

mortality,

and

fewer

births

to

teenage

mothers.”2

This

is reflected

in the

regions

according

to

their

composition

of

cities,

towns,

and

rural

townships.

As

the

different

distribution

of

so-cial

risk

factors

for

511)5

in different

areas

is well

characterized

through

the

level

of

urbanization,

for

which

we

control

in our

analysis,

we

maintain

that

social

risk

factors

do

not

constitute

a potential

bias

in

this

study.

Infant

mortality

from

respiratory

infec-tions

and

total

infant

mortality

were

low

in the

de-veloped

regions

where

pollution

is high,

while

511)5

was

low

in

the

less

developed

regions

where

pollu-tion

is

low.”

Maternal

smoking

is

extremely

rare,

regardless

of

the

region.

Teenage

mothers

are

more

prevalent

in

underdeveloped

unpolluted

areas,’1’2

which

should

lead

also

to an underestimation

of the

effect.

Infants

are

usually

not

autopsied

in

Taiwan,

though

studies

comparing

clinical

and

pathological

diagnosis

have

shown

that

autopsy

findings

nor-mally

are

not

able

to differentiate

511)5

and

suffoca-tion,

and

usually

do

not

change

the

clinical

diag-nosis

of

SIDS.37

Furthermore,

any

misclassification

of

explained

causes

under

SIDS

should

be

either

non-differential

with

regard

to

the

exposure

or

occur

more

likely

as

overdiagnosis

of

511)5

in

the

under-developed

unpolluted

regions,

thus

introducing

a

bias

towards

the

null.

The

use

of visual

range

as a measure

of air

pollu-tion

gives

more

weight

to the

small

particles

that

are

inhaled

most

deeply

into

the

lungs.’9’#{176} Furthermore,

as visual

range

gives

an

average

reading

for

a whole

area,

there

is less

chance

that

exposure

will

be

mis-classified

than

there

is with

local

air

samples.

Visual

range

is a good

measure

of air pollution

in the

mild

Taiwan

climate,

although

in

cooler

climate

zones

modifications

might

be needed

to allow

for

the

nat-ural

occurrence

of fog.

The

association

of pollution

with

511)5

as

found

in this

study

is consistent

with

previously

proposed

pathways

involving

the

respi-ratory

system.

Both

assumptions

about

the

disease

onset,

ie,

several

hours

or

up

to

9 days,

show

a

consistently

strong

association

with

air

pollution.

The

influence

of the

psycho-emotional

dimension

on

health

has

been

taken

up

recently

in

medical

research.

The

newly

created

term

seasonal

affective

disorder

gives

a

first

glimpse

of

the

influence

of

weather

and

climate

on

human

life.

“Wet-cold

weather

patterns,”

which

have

been

mentioned

in a

study

on

SIDS,39

could

influence

infants

at the

level

of

the

autonomic

nervous

system

and

thus

could

possibly

subtly

interfere

with

vital

functions

like

res-piration.

Such

influence

could

increase

the

suscepti-biity

of

infants

to

a malfunctioning

of

respiration

and

be

one

of the

stones

in the

risk-factor

mosaic

of

SIDS.

Research

into

the

physiology

of

infant

devel-opment

and

adaptation

to climatic

and

environmen-tal changes

needs

to be expanded

to clarify

the

(5)

sible

role

of climate

and

weather

in the

etiology

of

511)5.

REFERENCES

1. Bonser RSA, Knight BH, West RR. Sudden infant death syndrome in Cardiff, association with epidemic influenza and with

temperature-1955-1974. mt j Epidemio!. 1978;7:335-340

2. Campbell MJ. Sudden infant death syndrome and environmental

temperature: further evidence for a time-lagged relationship. Med I Aust. 1989;151:365-367

3. Beal 5, Porter C. Sudden infant death syndrome related to climate. Acta Paediatr Scand. 1991;80:278-287

4. Schwartz J. Particulate air pollution and daily mortality in Detroit.

Environ Res. 199156:204-213

5. Fairley D. The relationship of daily mortality to suspended particulates in Santa Clara county, 1980-1986. Environ Health Perspect. 199089: 159-168

6. Mazumdar 5, Schimmel Higgins ITF. Relation of daily mortality to air

pollution. An analysis of 14 London winters, 1958/59-1971/72. Arch Environ Health. 198237:213-220

7. Firket M. Fog along the Meuse valley. Trans Farad Soc. 193632:

1192-1197

8. Hoppenbrouwers T, Calub M, Arakawa K, Hodgman JE. Seasonal relationship of sudden infant death syndrome and environmental pd-lutants. Am

J

Epidemiol. 1981;113:623-635

9. Greenberg MA, Nelson KE, Carnow BW. A study of the relationship

between sudden infant death syndrome and environmental factors. Am

I

Epidemic!. 1972;98:412-422

10. Audliems A, Barnes A. Sudden infant deaths and dear weather in a

subtropical environment. Soc Sci Med. 1987;24:51-56

Ii. Kn#{246}belHH, Yang WS, Ho MS. Urban and regional differences in infant

mortality in Taiwan. Soc Sci Med. 199439:815-fl

12. Environmental Protection Agency Taiwan ROC. Report on Air Quality in

Taiwan. Taipei, Taiwan: EPA; 1989

13. Grether JK, Schulman J, Croen LA. Sudden infant death syndrome

among Asians in California. IPediatr. 1990;116:525-528

14. Ponsonby AL, Jones ME, Lumley J, Dwyer D, Gilbert N. Climatic

temperature and variation in the incidence of sudden infant death syndrome between the Australian states. Med IAust. 1992;156:246-251 15. Sivan Y, Shen G, Schonfeld T, Nitzan M, NutmanJ. Sudden infant death

syndrome in the Tel Aviv and Petah Tivka districts. Isr

J

Med Sci.

1992;28:430-435

16. Osterreichisches Statistisches Zentralamt. Demographisches Jahrbuch Osterreichs. Wien: Osz; 1990

17. Sato Y, Kagehara B, Mukai T, Endo T. Diagnosis of asphyxia on the sudden infant death-prone sleeping position and vomit aspiration. Jpn

I

Legal Med. 1992;46:407-412

18. Ozkaynak H, Schatz AD, Thurston GD. Relationships between aerosol extinction coefficients derived from airport visual range observations and alternative measures of airborn particulate mass.

J

Air Pollut Control Assoc. 198635:937-942

19. White WH, Anderson JA, Blumenthal DL, et al. Formation and tram-port of secondary air pollutants: ozone and aerosols in the St. Louis

urban plume. Science. 1976;194:187-189

20. Charison RJ, Ahlquist NC, Selvidge H, Mac Ready PB. Monitoring of

atmospheric aerosol parameters with the integrating nephelometer. I Air Po!!ut Control Assoc. 1969;19:937-942

21. Charlson RJ, Vanderpol MI, Covert DS, Waggoner DS, Ahlquist NC.

Sulfuric add-aminonium sulfate aerosol: optical detection in the St.

Louis region. Science. 1974;184:156-158

22. Horvath H, Charlson RJ. The direct optical measurement of atmo-spheric air pollution. Am md Hyg Assoc

J.

196930:500-509

23. World Health Organization. International Statistical Classification of Dis-eases, Injuries and Causes of Death. Geneva, Switzerland: World Health

Organization; 1975 Revision

24. Gilbert R, Rudd P, Berry PJ, et al. Combined effect of infection and

heavy wrapping on the risk of sudden unexpected infant death. Arch Dis Child. 199267:171-177

25. Yang WS, Kn#{246}bel HH, Kuo PY. Study of Social Factors ofSudden

Unex-pected Deaths in Taiwan. Taipei, Taiwan: ROC National Science Council; July 1993 (Report No. NSC8I-0301-H-001)

26. Central Weather Bureau. Manual of Ground and Climate Measurements.

Taipei, Taiwan: Central Weather Bureau; 1982

27. Thom CC, Oft WR. A proposed uniform air pollution index. Atmos Environ. 1976;10:261

28. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13-22

29. Azaz Y, fleming PJ, Levine M, McCabe R, Stewart A, Johnson P. The relationship between environmental temperature, metabolic rate, sleep state, and evaporative water loss in infants from birth to three months.

Pediatr Res. 199232:417-423

30. fleming P1, Azaz Y, Wigfield R. Development of thermoregulation in infancy: possible implications for SIDS. Clin Pathol. 1992; 45(suppl): 17-19

31. Wailoo NW, Petersen SA, Whittaker H, Goodenough P. The thermal

environment in which 3-4 months old infants sleep at home. Arch Dis Child. 1989;64:600-604

32. Ministry of the Interior ROC. Taiwan-Pu/den Demographic Fact Book.

Taipei, Taiwan: Ministry of the Interior; 1981-1991 (inclusive)

33. Department of Budget Accounting & Statistics, Taiwan PrOVincial Gov-ernment ROC. Reporton theSurveyofFamilylncome & Expenditure Taiwan

Province, ROC. Taipei, Taiwan; Taiwan PrOVincial Government ROC; 1993

34. Bureau of Budget Accounting and Statistics, Taipei Municipal

Govern-ment. The Statistical Abstract of Taipei Municipality. Taipei, Taiwan: Taiwan Municipal Government; 1981-1991 (inclusive)

35. Leadbeatter 5, Knight B. Sudden infant death and suffocation. Br Med

J.

1989;299:455

36. Bunai Y, Ohya I, Brinkmann B. Pathological approaches to SIDS L Special emphasis on the histopathologicaldifferences betweenSlDS and

asphyxia. Jpn ILegal Med. 1992;46:405-406

37. Haas JE, Taylor JE, Bergman AB, et al. Relationship between epidemi-ological risk factors and dinicopathological findings in the sudden

infant death syndrome. Pediatrics. 199391:106-112

38. Heleniak EP. Seasonal affective disorder: lower or raised resting mets-bolic rate? Med Hypotheses. 1993;40:367

39. Deacon EL, Williams AL The incidence of the sudden infant death syndrome in relation to climate.

mt

I Biometeorol. 198226207-218

at Viet Nam:AAP Sponsored on September 1, 2020

(6)

1995;96;1106

Pediatrics

Harald H. Knöbel, Chien-Jen Chen and Kung-Yee Liang

Air Pollution in Taiwan

Sudden Infant Death Syndrome in Relation to Weather and Optimetrically Measured

Services

Updated Information &

http://pediatrics.aappublications.org/content/96/6/1106

including high resolution figures, can be found at:

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtml

entirety can be found online at:

Information about reproducing this article in parts (figures, tables) or in its

Reprints

http://www.aappublications.org/site/misc/reprints.xhtml

Information about ordering reprints can be found online:

(7)

1995;96;1106

Pediatrics

Harald H. Knöbel, Chien-Jen Chen and Kung-Yee Liang

Air Pollution in Taiwan

Sudden Infant Death Syndrome in Relation to Weather and Optimetrically Measured

http://pediatrics.aappublications.org/content/96/6/1106

the World Wide Web at:

The online version of this article, along with updated information and services, is located on

been published continuously since 1948. Pediatrics is owned, published, and trademarked by the

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has

http://www.aappublications.org/site/misc/Permissions.xhtml

http://www.aappublications.org/site/misc/reprints.xhtml

Sudden Infant Death Syndrome in Relation to Weather and Optimetrically Measured Air Pollution in Taiwan

Sudden

ABSTRACT.

Objective.

average

greater

incidence

confounders.

Previous

varied

an equally

undeveloped

suffocation

suffocation

higher

In the

pollution.

relationship

primary

NH4HSO4

usually

temperature,

Data from death certificates for all deaths of infants in Taiwan

assigned

A forensic doctor examined 42% of SIDS cases, a coroner 15%, and the remaining death certificates were issued by a physician. A

death

Data on weather were recorded by the Central Weather Bureau of the Ministry of Transport and Communications. Nine stations

avail-able. Daily mean values for all days in the years 1981 to 1991 of visibility, temperature, air pressure, rainfall, sunshine hours, rel-ative humidity, and windspeed were used. Visibility is defined as the furthest distance from which an object (or, at nighttime, a

The measurements were repeated every 3 hours and followed a

available stations were used for the correlation analysis, the other

control for seasonality ofbirths, the number of live births 3 months before death (3 months being the peak age of SIDS and

logarithm

and temperature. The daily number of live births was included to control for the different population sizes of the regions and

hours,

weather

2). Weather

of SIDS

visibifity

hours

TABLE 1. Correlation of Optimetrically Measured Air Pollu-tion to Gravimetrically Measured Air Pollution and Weather in Taiwan

dioxide

Relative humidity -0.13 -0.15. -0.17 -0.19

PSI, pollutants

slightly

whole

before

respiratory

TABLE 2. Postperinatal SIDS Rate by Daily Average Visibility (in han) on Day of Death and During the 9 Days Before Death

death

ratio; CI, confidence interval.

incidence

level of urbanization.

equally

lowest

weather

TABLE 3. Postperinatal SIDS Rate by Daily Average

1Deaths from SIDS per 1000 live births.

at Viet Nam:AAP Sponsored on September 1, 2020

ARTICLES 1109

direction

temperature.

as the relationship

values

parts

seasons,

outside

distance

gravimetrical

of certain

carried

reopen

ur-banization

population

rural

Infant

regardless

shown

regard

inhaled

measure