Self-Perpetuating Effects of Birth Size on Blood Pressure Levels in Elderly People
It has been suggested that essential hypertension is determined by 2 separate mechanisms: a growth-promoting process in childhood and a self-perpetuating mechanism in adult life. We report a clinical study of 500 people taken from a cohort of 7086 men and women who were born in Helsinki from 1924 to 1933, and whose body size at birth was recorded. As expected, blood pressure levels were inversely related to birthweight and birth length. These associations, however, were confined to the 213 people who had previously been diagnosed as having hypertension. In them, a 1-kg increase in birthweight was associated with a 6.4-mm Hg (95% confidence interval, 1.0 to 11.9) decrease in systolic blood pressure recorded at the clinic, and with a 9.3-mm Hg (95% confidence interval, 2.1 to 16.5) decrease recorded by ambulatory measurement. We conclude that pathological processes initiated in utero become self-perpetuating in adult life and lead to hypertension. Among elderly people with established hypertension, these processes have a strong effect on blood pressure levels, because they are processes that do not respond well to treatment.
In children, blood pressure rises with increasing age.1 The rise is greater in those who are growing more rapidly. This led Lever and Harrap1 to suggest that essential hypertension is determined by a growth-promoting process that leads to elevated blood pressure levels in early adult life, but that might not, by itself, be sufficient to cause hypertension. Because it is known that in secondary hypertension, raised blood pressure may persist after removal of the primary cause, eg, a phaeochromocytoma, they postulated a second mechanism, a self-perpetuating process, that acts in adult life.
People with hypertension tend to have had low birthweight.2,3 We propose that slow fetal growth initiates a self-perpetuating process. One possible self-perpetuating process, proposed by Brenner and Chertow,4 is initiated by the reduced number of nephrons found in people who were small at birth. Reduced numbers of nephrons lead to increased pressure in the remaining glomerular capillaries. This, combined with the loss of nephrons that accompanies normal aging, is thought to lead to the development of glomerulosclerosis, which in turn leads to further loss of nephrons and a self-perpetuating cycle of rising blood pressure and further nephron loss.
We report a study of the blood pressures of elderly men and women taken from a cohort of 7086 people born in Helsinki from 1924 to 1933.5,6 The birth weights of this cohort were recorded. A previous analysis has shown that the 1958 people who were receiving antihypertensive medication had low birthweight and short body length at birth.3 After birth, they gained in weight and height rapidly so that their body size ‘caught up.‘ We have now examined the effects of birth size on blood pressure levels in a sample of 500 people from the cohort.
We studied 500 elderly men and women from a cohort of 7086 people who were born at the Helsinki University Central Hospital from 1924 to 1933, went to school in Helsinki and were still living in Finland in 1971. Each person had detailed measurements of body size at birth. Details of this study sample have been published.7 The subjects comprised people still living in the Helsinki area, and they were invited to attend a clinic in the morning after an overnight fast to have an oral glucose tolerance test. At the clinic, blood pressure was measured from the right arm in the sitting position, using a standard sphygmomanometer, and was recorded as the mean of 2 successive readings with the subject having sat for at least 10 minutes. Subjects taking medication for the treatment of hypertension were asked not to take medication on the morning of clinic attendance, because some medications (eg, thiazide diuretics) alter glucose-insulin levels. We defined subjects as hypertensive if they reported that they had been previously diagnosed by a physician as having hypertension. Each subject’s height was measured using a Kawe stadiometer; weight was measured on a Seca-α 770 scale. Waist circumference was measured using a soft tape according to standard practice. The percentage of body fat was determined by a handheld bioelectric impedance analyzer (Omron Body Logic Body Fat Analyzer from Omron Healthcare).
Ambulatory blood pressure was monitored on a subsample of 169 subjects; 72 of these subjects (43%) were hypertensive, similar to the percentage in the whole subsample. Those currently on therapy resumed it before monitoring commenced. We did not have sufficient equipment to offer ambulatory monitoring to all subjects, but invited subjects to participate as equipment became available. We used a portable noninvasive SpaceLabs Medical 90207 monitor. The recorder was attached before midday to the nondominant arm. Blood pressure was measured at 20-minute intervals from 6:00 am to 9:00 pm and at 60-minute intervals from 9:00 pm to 6:00 am The data were read by 90219 ABP PC Interface software. We averaged the ambulatory blood pressure over 24 hours. The ethical committee of the National Public Health Institute, Helsinki, approved the study.
We analyzed the data using multiple linear regression and tabulation of means. Probability values refer to analyses with continuous variables.
The mean age of the 500 subjects was 70 years (range, 65 to 75 years); 186 (37%) were men. Two hundred thirteen (43%) had been diagnosed as having hypertension, of whom 75 (35%) were men and 175 (82%) were currently on antihypertensive medication. Table 1 compares the body sizes of the normotensive and hypertensive subjects at birth and at present. The hypertensive subjects had lower birthweight, although this was not statistically significant. They had shorter body length at birth (P=0.02 in both sexes combined) and lower placental weight (P=0.04). These associations with birth size were little changed by adjustment for the duration of gestation (length, P=0.01; placental weight, P=0.04) or by exclusion of people born before term. People with hypertension currently had a higher body mass index, waist circumference, and percentage body fat. Table 1 shows the mean blood pressures in the 2 groups. As expected, subjects with hypertension had higher mean blood pressures recorded either at the clinic or by ambulatory recording. Mean blood pressures at the clinic were 158/91 mm Hg in men and 160/88 in women. This difference was not statistically significant, and we have therefore not adjusted blood pressure values for sex in our presentation, although we have done so in calculating probability values.
Birth Size and Blood Pressure
Systolic blood pressure decreased with increasing birthweight and birth length. It was unrelated to the duration of gestation. Table 2 shows that these associations with small body size at birth were confined to people with hypertension (interaction between birthweight and hypertension, P=0.06; interaction between birth length and hypertension, P=0.05). Among people with hypertension a 1-kg increase in birthweight was associated with a 6.4-mm Hg (95% confidence interval [CI], 1.0 to 11.9) decrease in systolic blood pressure. The corresponding figure for normotensives was −1.2 (95% CI, −6.7 to 4.3), whereas for the total study sample it was 3.5 (95% CI, −0.6 to 7.6). There were similar trends with diastolic blood pressure, although they were not statistically significant. If hypertension was defined as systolic blood pressure >160 mm Hg, recorded at the clinic, rather than self-reported diagnosis, the findings were similar. This definition categorizes 228 (46%) people as hypertensive. When we used lower values of systolic pressure, 150 or 140 mm Hg, the majority of the people were categorized as hypertensive. Nevertheless, the interaction between birthweight and hypertension remained. Trends in men and women were similar.
We examined the same associations in the 169 people who had ambulatory blood pressure measurements. The mean blood pressures were 133/79 mm Hg in men and 132/75 mm Hg in women, and the correlation between ambulatory measurements and those made at the clinic were 0.60 for systolic blood pressure and 0.63 for diastolic pressure. Table 3 shows that the trends with systolic blood pressure over 24 hours were similar to those for pressure recorded at the clinic (interaction between birthweight and hypertension, P=0.02; interaction between birth length and hypertension, P=0.07). Among people with hypertension a 1-kg increase in birthweight was associated with a 9.3-mm Hg (95% CI, 2.1 to 16.5) decrease in systolic blood pressure. The corresponding figure for normotensives was −1.9 (95% CI, −7.8 to 3.9), whereas for hypertensive and normotensive subjects combined, it was 3.9 (95% CI, −0.7 to 8.5). There were no significant trends with diastolic blood pressure.
Current Body Size and Blood Pressure
Table 4 shows the simultaneous effects of body length at birth and current body mass index. Both short length and high body mass index were associated with increased systolic blood pressure. When we examined the trends in systolic pressure among normotensive and hypertensive subjects, we found that the trend with current body mass index was present only among normotensive subjects. Table 5 shows that among people with hypertension systolic pressure was unrelated to body mass index, waist circumference, or percentage body fat (interaction between body mass index and hypertension, P<0.001). There were similar trends with diastolic pressure (interaction, P=0.01).
We have examined a sample of elderly men and women taken from a large cohort of people born in Helsinki, Finland, from 1924 to 1933. Although rates of coronary heart disease are high in Finland, access to medical care is similar to that in other European countries in which, unlike the United States, health services are mostly funded by central government. The use of a national identification number makes it possible to trace people rapidly and link them to national record systems, which include records of people receiving medication for chronic disease.
We have previously shown that the 1958 people in this Helsinki cohort who had hypertension, defined by receiving antihypertensive medication, had lower birthweight and shorter body length at birth than did other people.3 Consistent with this association with established hypertension, and with many previous studies of blood pressure levels in populations of different ages, we found that small body size at birth was associated with raised systolic pressure8 (Table 2). In the total sample, the increase in systolic pressure associated with a 1-kg decrease in birthweight was small, 3.5 mm Hg, which is similar to what has been shown previously.8 We found, however, that the association was only present among people who had hypertension (Tables 2 and 3⇑). We confirmed the findings from blood pressures measured in the clinic with ambulatory measurements made on a subsample. Ambulatory measurements give a picture of blood pressure over a 24-hour period and are now widely used in clinical practice. Reassuringly, the clinic and ambulatory measurements were highly correlated. Although the subsample was selected by the availability of equipment on the day of clinic attendance, similar proportions of people had hypertension as in the sample as a whole.
Among people with hypertension a 1-kg increase in birthweight was associated with a 6.4-mm Hg decrease in systolic pressure recorded at the clinic and a 9.4-mm Hg decrease recorded in a subsample by ambulatory measurements over 24 hours. These large differences occurred, despite the subjects being on treatment for hypertension. There were no similar trends with diastolic pressure. With increasing age, pulse pressure widens through rising systolic pressure. Diastolic pressure does not tend to rise, and one would not therefore expect it to be linked to self-perpetuating processes that act on blood pressure and are associated with aging.
Our findings are consistent with the hypothesis that slow growth in utero initiates a self-perpetuating mechanism that leads to hypertension. We suggest that during childhood and early adult life, the various regulatory mechanisms that control blood pressure ensure that the pathological processes associated with slow fetal growth lead only to a small increase in blood pressure. In older people, however, age-related damage begins a cycle of rising blood pressure, further damage, and the development of hypertension. Increasingly, people with lesions acquired in utero become hypertensive, and the association between birthweight and blood pressure becomes focused in this group. It also becomes amplified by the self-perpetuating cycle. This framework of ideas may illuminate the paradox that low birthweight has only small effects on blood pressure levels in the general population,9 but has large effects on morbidity and mortality from cardiovascular disease.10
Two possible self-perpetuating mechanisms involve nephron number and arterial elastin. The number of nephrons each person has at birth varies widely, from 300 000 to 1 100 000.11 Small babies have fewer nephrons. As described in the introduction, this, combined with age-related nephron loss, could lead to a self-perpetuating cycle of rising blood pressure and nephron loss.4 Alternatively, Martyn and Greenwald12 have shown that people who were small at birth have less elastic arteries. The elasticity of larger arteries depends on the scleroprotein elastin, which is laid down in utero and during infancy and thereafter turns over slowly. Its half-life in humans is ≈40 years. Reduced elastin deposition in small babies leads to stiffness in the major arteries, which leads to raised pulse pressure. The loss of elastin with aging will amplify this increase in pulse pressure. It can readily be shown that stiffer, less compliant arteries alter the pressure wave generated by the contraction of the heart in such a way that systolic pressure increases while diastolic pressure tends to fall.13
We found that the strong association between high current body mass index and raised blood pressure, which has been described many times, was confined to people who were not hypertensive (Table 5). There were also strong associations between raised blood pressure and high waist circumference, and high percentage body fat, which again were confined to normotensive people. We suggest that the effects of high fat mass on raised blood pressure are masked in hypertensive subjects by the effects of the self-perpetuating processes linked to slow growth in utero.
Our results could be influenced by preferential survival of subjects with particular profiles of birthweight, body mass index, and blood pressure. Although 74% of people invited to take part did so, our sample is not representative of the cohort. Sixty-three percent of the participants were women, which must in part reflect their longer life expectancy. Our findings, however, are based on internal comparisons within the sample, and we do not think it is likely that the differences in the associations among normotensive and hypertensive subjects can be attributed to selection bias. In our analyses of birth size and hypertension, we defined hypertension by a history of treatment for the disorder. Our findings would presumably be less impressive if the patients with hypertension had had fully effective therapy that restored their blood pressures to normal levels. The findings based on blood pressure recorded at the clinic (Table 2) were confirmed by ambulatory measurements on a subsample (Table 3). Although this subsample may not have been representative of the study sample, a similar percentage was hypertensive. The subsample was selected through the availability of the monitoring equipment on the day of attendance.
Implications for Clinical Practice
A strong effect of low birthweight on blood pressure levels in hypertensive patients has previously been suggested by findings in South Carolina, in which patients with low birthweight were found more often to require second-line therapy with calcium channel blockers and ACE inhibitors, as opposed to first-line therapy with diuretics or β-blocking agents.14 Our findings similarly suggest that the blood pressures of hypertensive patients who had low birthweight are more difficult to control by therapy. The absence of an association between body fat and blood pressure in hypertensive patients suggests that reduction in body fat is of less importance in relation to blood pressure control in hypertensive patients than it is in the general population.
Pathological processes initiated in utero become self-perpetuating in adult life and lead to hypertension. Among elderly people with established hypertension, they have a strong effect on blood pressure levels, because the processes do not respond well to treatment. Pathological processes associated with high current fat mass also contribute to the development of hypertension. They have a weaker influence on blood pressure levels once hypertension is established, possibly because their effect is obscured by the stronger effect of the processes initiated in utero.
- Received July 26, 2002.
- Revision received August 20, 2002.
- Accepted December 31, 2002.
Curhan GC, Chertow GM, Willett WC, Spiegelman D, Colditz GA, Manson JE, Speizer FE, Stampfer MJ. Birth weight and adult hypertension and obesity in women. Circulation. 1996; 94: 1310–1315.
Eriksson JG, Forsen T, Tuomilehto J, Osmond C, Barker DJP. Fetal and childhood growth and hypertension in adult life. Hypertension. 2000; 36: 790–794.
Eriksson JG, Forsen T, Tuomilehto J, Winter PD, Osmond C, Barker DJP. Catch-up growth in childhood and death from coronary heart disease: longitudinal study. BMJ. 1999; 318: 427–431.
Forsen T, Eriksson JG, Tuomilehto J, Osmond C, Barker DJP. Growth in utero and during childhood among women who develop coronary heart disease: longitudinal study. BMJ. 1999; 319: 1403–1407.
Robinson R. The fetal origins of adult disease. BMJ. 2001; 322: 375–376.
Eriksson JG, Forsen T, Tuomilehto J, Osmond C, Barker DJP. Early growth and coronary heart disease in later life: longitudinal study. BMJ. 2001; 322: 949–953.
Merlet-Benichou C, Leroy B, Gilbert T, Lelievre-Pegorier M. Retard de croissance intra-uterin et deficit en nephrons [Intrauterine growth retardation and inborn nephron deficit]. Medecine/Sciences. 1993; 9: 777–780.
Martyn CN, Greenwald SE. Mechanisms for in-utero programming of blood pressure. In: Barker DJP, ed. Fetal origins of cardiovascular and lung disease. New York, NY: Marcel Dekker; 2001.
Lackland DT, Egan BM, Syddall HE, Barker DJP. Associations between birthweight and antihypertensive medication in black and white Americans. Hypertension. 2002; 39: 179–183.