Changes in blood pressure among students attending Glasgow University between 1948 and 1968: analyses of cross sectional surveysBMJ 2001; 322 doi: https://doi.org/10.1136/bmj.322.7291.885 (Published 14 April 2001) Cite this as: BMJ 2001;322:885
- Peter McCarron, lecturer in epidemiology and public health medicinea (, )
- Mona Okasha, PhD studenta,
- James McEwen, professor of public health medicineb,
- George Davey Smith, professor of clinical epidemiologya
- a Department of Social Medicine, University of Bristol, Bristol BS8 2PR
- b Department of Public Health, University of Glasgow, Glasgow G12 8RZ
- Correspondence to: Peter McCarron, Surveillance Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, 6130 Executive Boulevard, Executive Plaza North, Suite 4097, Bethesda, MD 20892-7350, USA
- Accepted 23 January 2001
Objectives: To examine the changes in blood pressure over time in a cohort of young adults attending university between 1948 and 1968.
Design: Cross sectional study.
Setting: Glasgow University.
Participants: 12 414 students aged 16-25 years—9248 men (mean age 19.9 years) and 3164 women (19.2 years)—who participated in health screening on entering university between 1948 and 1968.
Main outcome measures: Systolic and diastolic blood pressure.
Results: In male students mean systolic blood pressure adjusted for age decreased from 134.5 (95% confidence interval 133.8 to 135.2) mm Hg in those born before 1929 to 125.7 (125.0 to 126.3) mm Hg in those born after 1945, and diastolic blood pressure dropped from 80.3 (79.8 to 80.8) mm Hg to 74.7 (74.2 to 75.1) mm Hg. For female students the corresponding declines were from 129.0 (127.5 to 130.5) mm Hg to 120.6 (119.8 to 121.4) mm Hg and from 79.7 (78.7 to 80.6) mm Hg to 77.0 (76.5 to 77.5) mm Hg. Adjustment for potential confounding factors made little difference to these findings. The proportion of students with hypertension declined substantially in both sexes.
Conclusions: Substantial declines in systolic and diastolic blood pressure over time were occurring up to 50 years ago in young adults who were not taking antihypertensive medication. Since blood pressure tracks into adult life, the results of the cross sectional comparisons suggest that factors acting in early life may be important in determining population risk of cardiovascular disease. Changes in such factors may have made important contributions to the decline in rates of cardiovascular diseases, particularly stroke, seen in developed countries during the past century.
What is already known on this topic
What is already known on this topic Mortality from stroke in developed countries decreased throughout the 20th century; declines in mortality from coronary heart disease occurred more recently
Explanations may include favourable trends in risk factors, including blood pressure Since blood pressure tracks into adulthood, description of the trends occurring in blood pressure in young adults may help to elucidate the critical periods involved in influencing trends in blood pressure
What this study adds
What this study adds Declines in blood pressure over time were taking place up to 50 years ago in young adults
These changes did not result from use of antihypertensive medications and point to the importance of factors in early life in determining blood pressure and risk of subsequent cardiovascular disease
Mortality from stroke has declined dramatically in developed countries throughout most of the past century, 1 2 and more recently mortality from coronary heart disease has also decreased.1 Raised blood pressure is a major modifiable risk factor for both conditions.3 Since blood pressure in young adults is also positively associated with mortality from cardiovascular disease in later life,4 decreases in blood pressure in this age group could underlie declines in mortality from cardiovascular disease.
Blood pressure tracks from childhood through to later life.5 Examining trends in blood pressure in young adults, in whom blood pressure is largely uninfluenced by disease, medication, or behavioural changes consequent on morbidity, can help to elucidate the population determinants of blood pressure. To date, however, few studies have investigated such trends.6 We report on the changes in blood pressure over time in a cohort of Glasgow University students born between 1925 and 1950.
Full details of the study are available elsewhere.7 Briefly, students attending Glasgow University between 1948 and 1968 were invited to participate in an ongoing health survey consisting of a questionnaire and clinical examination. Data collected included sociodemographic details, health behaviours, and medical history. Height, weight, and blood pressure were also recorded.
A total of 15 322 students (11 755 men and 3567 women) participated in the study. Students born before 1925 (319), with unknown date of birth (2), or aged over 25 years at examination (1324) were excluded from the analyses, as were two students with lower systolic than diastolic blood pressure and 104 students without a record of blood pressure. Students with missing data on potential confounding variables (1159) were also excluded.
Year of birth was divided into five bands: 1925-9, 1930-4, 1935-9, 1940-4, and 1945-50. Linear regression analyses were used to examine trends in blood pressure. Fully adjusted mean blood pressures were then calculated, controlled for the effects of smoking (yes/no), height (metres), body mass index (kg/m2), father's social class (I-V), and age at menarche (≤11, 12, 13, 14, ≥15 years). Likelihood ratio tests were used to test for a linear trend in blood pressure for each one year increase in birth year. To assess the change over time in the proportion of individuals with normal and high blood pressure, blood pressure adjusted for age was categorised as optimal (<120/80 mm Hg), normal (120-129/80-84 mm Hg), high normal (130-139/85-89 mm Hg, hypertension stage 1 (140-159/90-99 mm Hg), hypertension stage 2 (160-179/100-109 mm Hg), or hypertension stage 3 (≥ 180/110 mm Hg)8 and presented for different birth bands. Trend in blood pressure category was examined by using ordinal logistic regression. Analyses were carried out with Stata 6.0.9
After exclusions, 12 412 students (9248 men and 3164 women) remained eligible for analysis. Male students were slightly older than female ones, with a mean age of 19.9 (SD 1.8) years versus 19.2 (1.4) years. Systolic blood pressure was higher in male students (130.9 (13.0) mm Hg) than in female students (124.0 (12.1) mm Hg), whereas diastolic blood pressure was marginally lower in male students (77.0 (8.6) mm Hg v 77.7 (7.5) mm Hg).
Trends in blood pressure
Table 1 shows the changes in blood pressure with increasing birth band. As results adjusted for age were similar for the full cohort and for those people with data on confounding variables, only the latter are reported here. There was little difference between blood pressure adjusted for age and fully adjusted blood pressure.
When year of birth was entered as a continuous variable, systolic and diastolic blood pressure in men declined by 0.40 (95% confidence interval 0.36 to 0.45) mm Hg and 0.28 (0.25 to 0.30) mm Hg respectively per one year increase in year of birth. For women the corresponding declines were 0.55 (0.48 to 0.61) mm Hg and 0.19 (0.15 to 0.23) mm Hg. Controlling for confounding variables had little effect on these findings.
Trends in hypertension
Most students had optimal, normal, or high normal blood pressure (table 2). Most students with high blood pressure had stage 1 hypertension; 87.4% of all men with hypertension and 92.3% of all women with hypertension were in this category. Table 3 shows a linear increase in normotension and decrease in hypertension with increasing birth band in both sexes (P<0.0001).
Substantial downward trends in blood pressure occurred in male and female students attending the University of Glasgow between 1948 and 1968, and remained after confounding factors were controlled for.
Strengths and weaknesses of the study
The current study is large, and data are available for both sexes. As details of how blood pressure was recorded and of the identity of the observers were not available we cannot assess observer variation. Almost 74% of valid readings ended in 0 or 5; however, although this might attenuate the magnitude of any trend uncovered, it would have little effect on the direction of these associations.10 Also, as diastolic blood pressure in women was lower than that in men at the beginning of the study period but higher at the end, and as any systematic change in measurement protocol should have affected readings in both sexes similarly, it is unlikely that such a change could have resulted in the substantial trends observed.
Trends in blood pressure
In a study from Queen's University, Belfast, blood pressure declined during the period 1948-70 from 131/78 mm Hg to 123/72 mm Hg in male students and from 122/78 mm Hg to 110/70 mm Hg in female students. 11 12 US national health and nutrition examination surveys from 1960 to 1991 showed small increments in both mean systolic and mean diastolic blood pressure in men and women aged 18-29 years between the first two surveys (1960-2 and 1971-4), but since then blood pressure has declined in both sexes.6 A limitation of the above studies is that the role of confounding was not adequately assessed.
Trends in hypertension
The prevalence of hypertension in the current study is similar to that in the Johns Hopkins precursors study.13 In that study, which was smaller and restricted to male medical students, 3.5% of participants (mean age 22 years) who graduated between 1948 and 1964 had grade 2 hypertension or worse compared with 2.5% of male students in Glasgow. In the US national health and nutrition examination surveys the prevalence of stage 2 hypertension or worse in white men aged 18-29 increased between the first and third surveys and then declined, whereas in similarly aged women the prevalence showed a continual decline over the survey period, 1960-91.6
Explanation of trends
The substantial decline in blood pressure is not accounted for by confounding factors. Neither can the use of antihypertensive drugs explain the trends reported here as most of the measurements were carried out before antihypertensive drugs were available and to our knowledge none of the students was taking such medication.
A recent systematic review of trials of salt restriction reported that blood pressure was lower when salt intake was reduced, although the effects were small.14 However, in adolescents randomised in infancy to a low or a normal sodium diet, blood pressure was lower after 15 years in individuals in the intervention arm compared with those in the control group, suggesting that sodium restriction in infancy may have greater effects on later blood pressure than reduction in salt intake in adulthood.15 Although caution is required in interpreting these findings as the proportion of participants followed up was low, alterations in salt intake in early life, through dietary change and possibly through increasing use of refrigeration to store foodstuffs,1 may explain some of the decline in blood pressure found in the current study.
Other dietary changes during the period when most of participants in this study were born and grew up may be important. Between 1903 and 1934 there were large increases in fruit and vegetable consumption, especially among affluent people—the group from which most of the participants in this study came.16 Changes in consumption of other foods were proportionately much less. Accurate assessment of levels of dietary constituents in the first half of the last century is problematic, but there was certainly a trend towards increasing consumption of vegetables, fruit, and cereals. Importantly, fruit and vegetables, unlike other foods, were not rationed during the second world war, resulting in a general improvement in quality of diet across the social spectrum.17–19 That such changes may have contributed to a progressively more favourable blood pressure profile is supported by an American dietary trial in which there were substantial declines in blood pressure in people assigned diets rich in fruit and vegetables compared with controls on a more typical American diet.20
Most studies have reported that birth weight is inversely associated with subsequent blood pressure.21 However, even if the increase in birth weight in the United Kingdom since at least the 1970s had been occurring throughout the last century the increment is likely to have been modest, and certainly no more than 550 g in Scotland over the century (Information and Statistics Division, Common Services Agency, Scotland: unpublished SMR2 data). 22 23 Since it has been estimated that a 1000 g rise in birth weight is required for a 2 mm Hg drop in systolic blood pressure in people at age 50,21 any upward trend in birth weight in the current cohort could not explain the large declines in blood pressure seen. Finally, while it seems that postnatal growth has a role in determining later blood pressure this is poorly understood and findings are conflicting.24–26
The substantial declines in blood pressure occurring in young adults between 30 and 50 years ago in the United Kingdom may have had their origin in early life and may explain some of the decline in rates of cardiovascular disease over the past century.
We thank Alan Kerr, Christine Hamilton, and Heather Learmonth for entering the data and Jonathon Sterne for advice on statistical analyses.
Contributors: PMcC, GDS, and JMcE discovered the original data and designed the study. PMcC carried out the analyses and drafted the paper. GDS, MO, and JMcE commented critically on subsequent drafts. PMcC will act as guarantor.
Funding Chest, Heart and Stroke, Scotland; Stroke Association; NHS Management Executive; Cardiovascular Disease and Stroke Research and Development Initiative.
Competing interests None declared.