Introduction: Sedentary lifestyle and energy rich food have been associated with the risk of developing type 2 diabetes; limited data are available on environmental conditions in childhood on this risk later in life. The objective was to study if residency in the first 20 years of life affected the risk of developing type 2 diabetes.

Methods: In a cohort of 17811 men (48%) and women, mean age 53 years (range 33-81) participating in the population-based Reykjavík Study from 1967-91, 29% grew up in rural and 35% in coastal areas for an average of 20 years before moving to urban Reykjavík, but 36% lived in Reykjavík from birth. The prevalence of type 2 diabetes according to residency in early life was examined.

Results: The relative risk of developing type 2 diabetes was 43% lower in men (RR 0.57; 95% CI 0.43-0.77) and 26% lower (RR 0.74; 95% CI 0.56-0.99) in women living in rural areas for the first 20 years of their life compared with those living in urban Reykjavík from birth. The low prevalence among those that grew up in rural areas was maintained through the age categories of 55-64 years and 65 years and older.

Conclusions: Our findings indicate that persons growing up in rural areas in early 20^th century Iceland had lower risk of  developing type 2 diabetes later in life when compared with peers living in Reykjavík from birth. We postulate a prolonged effect of early development on glucose metabolism and risk of developing type 2 diabetes.

Olafsdottir E^1,2, Aspelund T^1,4, Torfadottir JE², Steingrimsdottir L^3,5, Sigurdsson G^4,5, Thorsson B¹, Benediktsson R^4,5, Eiriksdottir G¹, Valdimarsdottir UA², Gudnason V^1,4  

¹Icelandic Heart Association, ²Center of Public Health Sciences, ³Faculty of Food Science and Nutrition, ⁴Faculty of Medicine, University of Iceland, ⁵Landspítali University Hospital

Figure 1. Prevalence of type 2 diabetes in men (a) and women (b) living in urban Reykjavik in 1967, according to age categories and residency for the first 20 years of their life, in a rural area (29%), a coastal area (35%) or in urban Reykjavik (36%) from birth.

Figure 2.  Prevalence of type 2 diabetes, adjusted to sex and the age of 60 years, in men(48%) and women living in urban Reykjavik in 1967 according to residency in different regions of the country for the first 20 years of their life. The different regions of the country are subdivided into rural or coastal areas and urban Reykjavik. The two horizontal lines show the adjusted mean prevalence in rural (2.7%) and coastal (5.0%) areas. Each column is shown with error bars (with 95% CI) and numbers of cases behind each column are as follows.

Rural (total 5195): West and Northwest peninsula (1640), North (1002), East (290), South (2263).  Coastal (total 6140): West and Northwest peninsula (1918), North (1233), East (935), South (743), Reykjanes peninsula (1311).  Urban Reykjavik (6476).

 

Table I. Prevalence of type 2 diabetes in men and women living in urban Reykjavik in 1967, according to age categories and residency for the first 20 years of their life, in rural areas, coastal villages or in urban Reykjavik. 17 811 men and women born in 1907-1935 examined between 1967 and 1991 with a mean prevalence of 3.6% for type 2 diabetes.

Residency	Rural	Coastal	Urban
Men
Number	2415	2840	2990
n with T2D	72	132	150
Age in years, mean (± SD)	53.7 (±8.7)	52.2 (±8.5)	51.2 (±8.3)
Age range	34-79	33-78	33-78
% T2D diagnosed at study entry	61.5	66.7	58.5
T2D prevalence % (n)	3.0 (72)	4.6 (132)	5.0 (150)
"           at age <55	2.3	2.8	3.6
"           at age 55-64	2.8	6.6	6.5
"           at age 65+	7.3	11.5	11.6
Women
Number	2635	3078	3227
n with T2D	73	90	109
Age in years, mean (± SD)	54.2 (±9.1)	53.5 (±8.9)	52.3 (±8.9)
Age range	33-81	33-80	33-80
% T2D diagnosed at study entry	52.1	42.7	40.2
T2D prevalence % (n)	2.8 (73)	2.9 (90)	3.4 (109)
"           at age <55	1.2	1.5	1.5
"           at age 55-64	3.4	3.9	5.5
"           at age 65+	7.9	7.8	9.8

 

Table II.  The relative risk (RR) of developing type 2 diabetes according to residency for the first 20 years of life in (a) rural versus urban area and (b) coastal versus urban area with 95% confidence interval (CI).  % risk reduction and p-values are also shown.

		RR	95% CI	95% CI	%
	Model		Lower	Upper	reduction	p-value
Men (a)	1	0.52	0.39	0.69	48	<0.0001
	2	0.57	0.43	0.77	43	0.0002
	3	0.56	0.42	0.76	44	0.0001
	4	0.57	0.43	0.77	43	0.0002
Men (b)	1	0.87	0.68	1.09	13	0.2138
	2	0.91	0.72	1.16	9	0.4497
	3	0.91	0.72	1.16	9	0.4415
	4	0.91	0.71	1.15	9	0.4289
Women (a)	1	0.70	0.52	0.93	30	0.0154
	2	0.74	0.56	0.99	26	0.0435
	3	0.75	0.56	1.00	25	0.0555
	4	0.76	0.57	1.02	24	0.0710
Women (b)	1	0.78	0.59	1.02	22	0.0765
	2	0.77	0.59	1.01	23	0.0594
	3	0.78	0.59	1.03	22	0.0760
	4	0.78	0.60	1.03	22	0.0782

Model 1: Adjusted for age.

Model 2: Adjusted for age, BMI, systolic blood pressure and serum triglycerides.

Model 3: Adjusted additionally for smoking status, education and leisure time physical activity.

Model4: Adjusted additionally for time of entry into the study.

 

 

Table III.  Baseline characteristics when entering the Reykjavik Study (1967-1991) according to residency in a rural coastal or urban area for the first 20 years of life.

	Men			Women
Variables	Rural	Coastal	Urban	Rural	Coastal	Urban
Number	2487	2972	3140	2708	3168	3336
Age in years (±SD)	53.6 (±8.7)	52.3 (±8.6) ***	51.3 (±8.3) ***	54.2 (±9.2)	53.5 (±8.9) *	52.4 (±8.9) ***
Cholesterol, mmol/L (±SD)	6.36 (±1.07)	6.43 (±1.07) *	6.35 (±1.07)	6.65 (±1.22)	6.69 (±1.26) *	6.51 (±1.21) **
Triglycerides, mmol/L, median (IQR)	1.03 (0.65)	1.09 (0.67) ***	1.14 (0.71) ***	0.90 (0.51)	0.92 (0.54) ***	0.93 (0.56) ***
Glucose, mmol/L (±SD)	5.36 (±0.85)	5.40 (±0.87) *	5.46 (±0.94) ***	5.12 (±0.78)	5.18 (±0.83) **	5.18 (±0.89) ***
BMI, kg/m2 (±SD)	25.6 (±3.3)	25.7 (±3.4)	26.0 (±3.6) ***	25.0 (±4.3)	25.2 (±4.3)	25.1 (±4.3) *
Systolic BP, mm Hg (±SD)	140.2 (±19.2)	139.6 (±18.7)	140.9 (±19.5) ***	137.3 (±20.8)	138.2 (±20.9) *	136.6 (±20.2)
Diastolic BP, mm Hg (±SD)	88.1 (±10.3)	87.9 (±10.1)	88.6 (±10.8) *	84.0 (±10.2)	84.3 (±10.3)	84.0 (±10.1)
CHD prevalence†  %	2.9	3.1	2.6	0.5	0.9*	0.8*
Current smoking %	49.6	56.7***	59.3***	38.5	38.0	44.1***
Education more than secondary %	23.7	20.2***	21.8*	8.1	6.3**	10.8***
Sports activity at 20-29y %	17.0	15.3	21.3***	7.7	7.6	10.7***
Leisure time physical activity at entry %	11.9	10.0*	13.8*	10.0	11.4	14.4***

Significance estimates: *p<0.05; **p<0.01; *** p<0.001 for age-adjusted comparison, rural residency used as reference.

^† Prevalence from history of MI, PCI, and CABG in hospital records.