Sodium in drinking water and adolescent blood pressure by Jean Charlotte Forseth A thesis submitted in partial fulfillment of the requirements for the degree of MASTER OF NURSING Montana State University © Copyright by Jean Charlotte Forseth (1981) Abstract: The blood pressures of thirty-three healthy fourteen through sixteen year old high school students from a community with drinking water sodium of 378 mg/l were compared with blood pressures of thirty-four students of similar age and health from a control community that had 2.1 mg/l of sodium in the drinking water supply. Data were collected by questionnaire and measurement of variables known to affect blood pressure. Mean systolic and mean diastolic pressures for each group and for males and females were analyzed by t-test. Intervening variables were analyzed for correlation with blood pressure, then analysis of co-variance produced adjusted mean systolic and diastolic pressures for the two groups after controlling for the effect of the intervening variables. The high sodium group had higher systolic and diastolic means than the control group but only mean female diastolic pressure was significantly higher than the control group mean.  STATEMENT OF PERMISSION TO COPY In presenting th is thesis in partia l fu lfillm en t of the require­ ments fo r an advanced degree a t Montana State University, I agree that the Library shall make i t free ly available fo r inspection. I further agree that permission fo r extensive copying of th is thesis fo r scholarly purposes may be granted by my major professor, o r, in her absence, by the Director of L ibraries. I t is understood that any copying or publication of th is thesis fo r financial gain shall not be allowed without my written permission. Signature C. _________ Date 'Tha jj' <£ f, _______________ SODIUM IN DRINKING WATER AND ADOLESCENT BLOOD PRESSURE by JEAN CHARLOTTE FORSETH A thesis submitted in partia l fu lfillm en t of the requirements fo r the degree of MASTER OF NURSING Approved: Chairperson, Graduate Committee Head, Major Department graduate Dean MONTANA STATE UNIVERSITY Bozeman, Montana June, 1981 iii TABLE OF CONTENTS Page VITA ............................................ i i TABLE OF CONTENTS ................................................................................. i i i LIST OF TABLES . . . . . . . . . . . .................................................. v ABSTRACT ................................ vi CHAPTER 1. INTRODUCTION ........................................................................... I Hypothesis ................................................................................. 6 2. LITERATURE REVIEW ..................................................................... 7 Hypertension Epidemiology ........................................................... 7 Factors in Hypertension Development ................................. 9 Animal Studies . ..................................................................... 11 Human S tud ies ............................................................................. 14 Epidemiological Studies ......................................................... 17 Summary......................................................................................... 20 3. METHODOLOGY................................................................................. 21 Introduction ............................................................................. 21 Defin ition of Terms................................................................. 22 Community Characteristics ..................................................... • 23 S am p le ......................................................................................... 26 Procedure..................................................................................... 28 Part One Q ues tionna ire ................................................. 28 Part Two Q ues tionna ire ................................................. 29 Blood P re s s u re ................................................................. 30 Water T e s tin g ..................................................................... 32 Methods of A na lys is ................................................................. 33 Page 4. DATA ANALYSIS..................................................................... 36 Sample Description ............................................................. 36 Stage I A na lys is ................................................................. 38 Stage 2 A na lys is .................... 41 S ignificant Variables ............................................. 45 5. CONCLUSIONS......................................................................... 53 Discussion . ............................ 56 L im ita t io n s ......................................................................... 58 Recommendations............................................ 60 REFERENCES CITED . ............................................................................. 65 APPENDIX . . . ......................................... .... 69 iv VLIST OF TABLES Page Table 1. Community Population C ha rac te ris tics ..................... . . . 24 2. County Economic Comparisons ................................................. 25 3. Cases Which Did Not Meet Sample Selection C rite r ia ..................................................................................... 27 4. Chemical Analysis of High School Water ............................. 32 5. D istribution of Subjects by Gender and Group ................. 37 6. Age D istribution by Town........................................................ 37 7. Differences in Blood Pressure Means ................................. 39 8. Family History of Hypertension ............................................. 42 9. Height/Weight Ratio Correlation With Blood P re ssu re s ......................................................................... 46 10. S ignificant Personality Variables With Blood Pressure, Total Sample ............................................. 47 11. Adjusted Female Mean Blood Pressures ................................. 51 12. Adjusted Male Mean Blood Pressures..................................... 52 13. Comparison of Findings With Those of Tu th ill and Calabrese (1979) ................................................. • 55 v i ABSTRACT The blood pressures of th irty -th ree healthy fourteen through sixteen year old high school students from a community with drinking water sodium of 378 mg/1 were compared with blood pressures of th irty -fou r students of s im ilar age and health from a control commun­ ity that had 2.1 mg/1 of sodium in the drinking water supply. Data were collected by questionnaire and measurement of variables known to a ffec t blood pressure. Mean systolic and mean d iasto lic pressures for each group and for males and females were analyzed by t - te s t . Intervening variables were analyzed for correlation with blood pressure, then analysis of co-variance produced adjusted mean systolic and d iasto lic pressures fo r the two groups a fte r controlling for the effect of the intervening variables. The high sodium group had higher systolic and d ias to lic means than the control group but only mean female d iasto lic pressure was s ign ifican tly higher than the control group mean. CHAPTER I INTRODUCTION For many years, there has been controversy about the ro le of sodium in hypertension development. During c lin ic a l experience in a rural community, the investigator discovered that the water was very soft and tasted lik e sodium bicarbonate. In the course of ethno­ graphic interviews with an elderly population sample, histories of hypertension were common. Hypertension prevalence increases with age, so many elderly persons with hypertension would not be an uncommon finding. Upon further investigation, water analysis (State Water Quality Bureau, 1977) revealed the drinking water sodium level to be 440 mg per l i t e r which could provide a source of an additional gram of sodium per day. The water sodium level was not known by members of the health care professions or the public in that town. The American Heart Association in 1957 (Your 500 Milligram D iet) proposed a lim it of 20 mg/1 of sodium in drinking water to protect persons on low sodium diets from additional sources of sodium. Several studies have associated water sodium levels lower than 400 mg with elevated blood pressure in animals and humans. Findings of a possible connection between water sodium and hyper­ tension in th is community stimulated investigation into the likelihood 2of water sodium's contribution to elevated blood pressures in a healthy sample of the population. The sample selected fo r study was a group of teenagers, 14 through 16 years old, because that age group would be less lik e ly to have much chronic disease and medication use which could in terfe re with a study of blood pressure levels. Blood pressures of a healthy group of adolescents in a town with high drinking water sodium (experimental) were compared with a comparable group in a town with a low (2 .5 mg/1) drinking water sodium level (control group). A study of the effec t of water sodium levels and blood pressure could have significance for the population of a town with high sodium in drinking water. I f blood pressures were elevated in young healthy people, that e ffec t could be expected to continue since blood pressure rises with age. More hypertension or stroke prevalence could be associated with water sodium intake. The e ffec t of higher drinking water sodium levels than had been previously studied was the focus of th is study. Sodium's e ffec t on health was measured by blood pressure. Such a study could add to data about water sodium hazards and contribute to determination of the necessity fo r drinking water sodium standards. The findings could aid communities in providing public awareness of hidden sources of sodium and its e ffec t on health. The high sodium community's water supply comes from three deep 3wells. Water from shallow wells is usually mineralized with calcium and magnesium (hard water). In deep wells , higher sodium levels are found because calcium and magnesium are exchanged fo r sodium in rock formations as water seeps into the ground. I f ion exchange materials are not availab le , or the rock is limestone in composition, levels of sodium in ground water are low (Calabrese and T u th i l l , 1978). Other sources of sodium in drinking water are a r t i f ic ia l water softening, addition of large amounts of sodium hypochlorite fo r water disinfection, or sodium hydroxide to precip itate iron from the water. Road salting is another source of sodium contamination of drinking water and is in greatest use in large metropolitan areas of the northern United States (National Academy of Sciences, 1977). I t is most lik e ly that deep well ion exchange is mainly responsible for the high sodium level in the study community. There has been no standard set for the level of sodium in drinking water nationally or on a state leve l. The Safe Drinking Water Act which became effective with Interim Primary Regulations in June 1977, did not l i s t a maximum allowable level fo r sodium in drinking water. The Environmental Protection Agency did not include a standard for sodium because there were insu ffic ien t and conflicting studies about the e ffec t o f sodium on health on which to base a standard. I t was also believed that the minority affected by s a lt restricted diets could be managed by the ir physicians without forcing 4the majority to pay fo r further water treatment (U. S. Environmental Protection Agency, 1971). A standard of 100 mg/1 was suggested (National Academy of Sciences, 1977) to assure that less than 10% of d ietary intake of sodium would be provided by drinking water. Montana has 40 communities with water sodium levels over 20 mg/1. Of these, 28 communities have water sodium levels over 270 mg/1, the level used by the Montana Water Quality Bureau (unknown year) to advise town mayors of potential harm to persons on sodium restric ted diets. One town has water sodium of 798 mg/1. The problem is most prevalent in the eastern portion of the state . A portion of the reluctance of Federal agencies to establish a standard fo r sodium in drinking water may be related to findings which have illu s tra ted that most people consuming high sodium do not become hypertensive. Because i t is d i f f ic u lt fo r persons cu ltu ra lly accustomed to high sodium intake to modify th e ir d iets , and desalini zation of water is very expensive, some believe that i t is not necessary to reduce sodium for an entire population. I t may suffice to provide early detection and treatment for the minority of sa lt sensitive individuals (Abramowicz, 1980). I f such individuals could be id en tified and receive follow up care prior to hypertension development, such an approach might be adequate. Without that a b il i ty , i t seems necessary that investigation continue into hazards of high sodium intake and ways to reduce unnecessary sodium intake. 5One such study demonstrated higher blood pressures in adolescents consuming 107 mg/1 of sodium in community drinking water (Tu th ill and Calabrese, 1979). Nursing is a profession concerned with promotion of health and health-seeking behaviors. High sodium intake is of importance to nurses in teaching patients and fam ilies affected by hypertension, renal insuffic iency, a rte riosc lero tic disease, congestive heart fa ilu re , liv e r fa ilu re and toxemia of pregnancy. Therefore, nurses, as well as other health professionals, should be aware of high water sodium levels and include such information in health promotion ac tiv itie s . Groups which could be affected by high sodium intake have been id en tified . Newborns fed formula diluted with high sodium concentration water could be at increased risk for the development of hypernatremia, especially when i l l . The elderly population could be at higher risk fo r severe dehydration development when consuming high sodium water in the presence of the reduced kidney function which accompanies the aging process. In some schools in Montana, i t is s t i l l common practice to d is tribu te sa lt tablets to athletes with plenty of water intake during hot weather. In a high water sodium community, th is practice could be quite dangerous. Those hazards of high sodium intake should certa in ly be addressed by those respon­ sible for the provision and promotion of health. In addition, i f high water sodium content is shown to adversely 6affect the blood pressure of normal, healthy individuals, health teaching about the water sodium level and the necessity fo r decreasing dietary sa lt intake may be beneficial fo r the entire population from an early age. This study was done to add to knowledge about the risk from high sodium water to the healthy population. Studies of sodiun) risks may stimulate appropriate health interventions at the community level and within health care re fe rra l areas. Hypotheses The study was undertaken to answer th is question. Are blood pressure levels higher in 14 through 16 year old healthy subjects exposed to high sodium in drinking water than a sim ilar group which has low sodium in the drinking water? The following hypotheses were tested to address the question: 1. There w ill be an increase in the mean systolic and mean d iasto lic blood pressure values in the high sodium group over the low sodium group. 2. The upward s h ift in mean systolic and mean d iasto lic blood pressure values found w ill be proportionately higher than that found in Tu th ilT s (1979) study of high school sophomores. There w ill be a dose-response relationship between findings of the two studies. 73. There w ill be a higher prevalence rate of hypertension in the high sodium group than the low sodium group. CHAPTER 2 LITERATURE REVIEW Hypertension Epidemiology Controversy has surrounded lite ra tu re and research relating to hypertension in man for years. There has been no ongoing study of the incidence of hypertension over time. The prevalence rate of hypertension has not changed in recent years. The rate among adults was 18.1 per 100 population in the 1971-1974 National Health Survey and 18.2 in 1960-1962 (Borhani, 1979). M orta lity from hypertension and hypertensive heart disease, however, has decreased dramatically from 1950 to 1976. Hypertension is a very powerful risk factor fo r coronary heart disease. In recent years there has been a decrease in stroke mortality and the rate of mortality from coronary heart disease as well. In the Proceedings of the Conference on the Decline in Coronary Heart Disease Morta lity (National Institu tes of Health, 1979), changes in risk factors such as hypertension, physical a c tiv ity , dietary patterns, smoking and changes in treatment were cited as having an effec t on coronary heart disease m orta lity . Thus, 9multi fac to ria l causes are possible explanations of the decline in vascular disease m orta lity . Public awareness, education, screening, and treatment of hypertension have increased in recent years. While disease prevalence has not changed, hypertension mortality has declined. I t is suggested that is because of increased case finding and treatment. Factors in Hypertension Development Some basic mechanisms of hypertension in man have been described by Guyton (1976). One type of hypertension, volume loading hypertension, is caused by excess water and sa lt intake in patients with reduced kidney mass. The actual mechanism for this hypertension is the in i t ia l increased cardiac output caused by excessive water load which accompanies sa lt retention in the compro­ mised kidney. Following continued sa lt loading (controlled administra­ tion of high saline dose), the to ta l peripheral resistance increases. Since the basic mechanism of hypertension is either increased cardiac output or increased peripheral resistance, hypertension continues during the period of increase in volume. Guyton (1976) did experi­ ments in animals in which he demonstrated that increased to ta l sa lt in the body does not cause hypertension unless there is a concomitant rise in the flu id volume. As long as the kidneys are even p a r t ia lly 10 functional and can excrete water, sa lt is the factor that determines the volume of ex trace llu la r f lu id in the body. Renal arte ry constriction with chronic renin secretion is another mechanism by which hypertension is caused. Renin is secreted which causes angiotensin (a powerful vasoconstrictor) to increase peripheral resistance. Over a period of time, excretion of water and sa lt is depressed, thus contributing further to increased cardiac output and long-term hypertension (Guyton, 1976). Neurogenic factors have been proposed as a cause of hyper­ tension through prolonged stimulation of the sympathetic nervous system. I t is believed that abnormal sympathetic stimulation of the kidneys eventually causes a secondary renal abnormality which main­ tains the hypertensive state (Guyton, 1976). D iasto lic pressure is a reflection of to ta l peripheral resistance which is the mechanism in neurogenic hypertension (Groer and Shekleton, 1979). Essential hypertension is a term used to describe hypertension of unknown o rig in . In persons with essential hypertension, the urinary output-pressure curve is shifted to a higher leve l; the a rte ria l pressure must remain elevated in order to maintain the excretion of water and sa lt through the kidneys at an amount equal to intake. The cause of th is s h ift in the pressure curve is unknown (Guyton, 1976). Several studies have attempted to explain this phenomenon on the basis of sa lt intake. This study also addressed 11 the problem of essential hypertension development in re la tion to sodium intake. Several theories have been proposed to p a rt ia lly explain a hypothesis of sa lt induced hypertension. Guyton (1976) believed that essential hypertension could be explained (as are many known causes of hypertension) by a kidney abnormality in regulation of sa lt and water. This theory was supported by findings of a s h ift in the urinary output-pressure curve in patients with essential hyper­ tension. Urinary output levels can only be maintained i f the blood pressure is. high. When blood pressure f a l ls , the urinary out­ put drops to zero. He postulated the pathology was either thickened glomerular membranes which would reduce the glomerular f i l t r a t io n rate and increase tubular reabsorption of sa lt and water or in ­ creased afferen t a rte r io la r resistance caused by vascular sclerosis. Neither theory has been d irec tly ve rified in the lite ra tu re . This pathology explanation does not c learly postulate sa lt as a cause of the hypertension; rather, sa lt intake would be a contributor to the development of hypertension following subtle kidney changes. Animal Studies Several studies have proposed and tested the theory that a mechanism for sa lt induced hypertension is an increase in in tracellu lar 12 sodium in the peripheral a rte r ia l wall (Tobian and Binion, 1952; Haddy, 1974; Edmondson, H ilton , Thomas, Jones, and Patrick, 1975; Madden, Smith, and Llaurado, 1979). Madden et a l. (1979) demon­ strated a net movement of sodium to a compartment in the mesenteric a rte ria l wall in rats. The rats (not Dahl ra ts , the sa lt-sensitive genetic s tra in ) had hypertension induced by substituting 2% sodium chloride for drinking water ad l ib . The experimental rats in 8 weeks showed a 14% increase in blood pressure when compared with control animals. The researchers developed a model in which a sodium compartment within the ce ll was postulated to contain the excessive sodium. There was slowed sodium exchange in the mesenteric artery wall of salt-loaded ra ts , and decreased ex trace llu lar flu id to in trace llu la r flu id ra tio was indicative of in trace llu la r sodium accumulation. The authors believed th is sh ift could have a stiffen ing effec t on the a rte ria l walls leading to hypertension development. I . K. Dahl has been a persistent advocate of the role of sa lt in hypertension development since the 1950's. In the intervening years, he developed a genetic stra in of sa lt-sens itive rats which have served as subjects in much hypertension animal research. Genetic factors were demonstrated to play a part in the development of essential hypertension in rats . Dahl (1972) found that the sa lt- sensitive strain developed hypertension with high sa lt intake while the control strain did not display elevated blood pressures. 13 Evidence that the genetic fa u lt is in the kidneys as postulated by Guyton was provided in a recent study (Tobian, 1977). Bianchi (1974) and Dahl (1972) independently demonstrated hypertension may be caused by a renal defect in sodium excretion. Defin itive physiologic and histologic studies of the pathology of sa lt induced hypertension are not possible in humans. Corbet, Kuller, Blaine and Damico (1979) studied the effects of a high sa lt d ie t in swine to demonstrate a model of its occurrence in man. Swine were chosen because of the ir sedentary and obese features; they were believed to be s im ilar to man in that regard. Serum cholesterol, potassium, sodium, gross pathology and histology of the aortic vessels, heart walls and vessels, kidneys, and adrenal glands were u tiliz ed to study the high sa lt d ie t effect in experimental and control animals. Unfortunately, urinary excretion of sodium and water were not measured, nor were ex trace llu la r-in trace llu la r f lu id ra tios . The 3% sa lt d iet used was quite high in comparison with American d iets . Both systolic and d ias to lic blood pressures were elevated in the experimental animals, 7-12 mm over the control animals' values. There was no difference between groups in the variables weight, serum sodium, potassium or cholesterol. The report did not discuss results of gross pathology and histology studies which may have added to knowledge about mechanisms of sa lt induced hypertension. 14 Human Studies In studies of humans, genetic factors have been implicated which consistently demonstrate decrease in sodium excretion in black males given a sodium load. In the discussion of one study, (Lu ft, Grim, Fineberg, and Weinberger, 1979a) the genetic stimulus was speculated to be an adaptive response to o rig ina lly liv ing in areas with high temperature where dietary sodium was not read ily available. The black race exhibits more e ff ic ie n t means of temperature regulation in secretion of less sweat per hour than whites performing the same work load, so i t is conceivable that they and other peoples originating in semi tropical areas may have developed more e ff ic ie n t mechanisms of sodium conservation. Such mechanisms of sodium conservation could be hazardous to blacks liv ing in America or other countries with excessive sa lt intake. Pickering (1980, p. 13) wrote a c r it ic a l review of some research which has supported the role of sa lt in hypertension. He restated the theory of autoregulation f i r s t proposed by Guyton: According to the argument, excessive intake of sa lt increases extracellu lar f lu id volume, plasma volume and cardiac output. This leads to an increase in a rte ria l pressure, which in turn leads 15 to adaptive changes in peripheral blood vessels, some effected through the baroreceptor reflexes. This vasoconstriction results in an increased peripheral resistance that eventually maintains the a rte r ia l pressure as the cardiac output returns to normal. The portion of the autoregulation theory which he omitted in his restatement was that the increased pressure then becomes essential fo r the maintenance of normal urinary/sodium excretion balance with intake (Guyton, 1976; Coleman, Samar, Murphy, 1979). Pickering then concluded that excessive sa lt intake is not the cause of essential hypertension by c iting data from a study (Framingham Heart Study Data displayed by Dauber et a l . , 1967) which showed no s ignificant elevation of blood pressure in persons whose urinary sodium excretion varied between 8 and 13 Cm. The percentage of persons with high blood pressure increased from 20% at less than 8 Gm sodium excretion to 36% at greater than 13 Gm representing twice as many hypertensive subjects in 44-45 year olds. Since sodium excretion is equal to sodium intake on a 24 hour basis, the study Pickering u tilized lends some support to the sa lt hypothesis rather than d e fin ite ly refuting i t as was his in ten t. S ir George Pickering also cited three studies (Thomas, Ledingham and B e ilin , 1978; Simpson, Waal-Manning and B o ll i , 1978; and Bing, 16 Thurston and Swales, 1979) which demonstrated that persons with highest blood pressures in a population excrete the same amount of sodium in 24 hours as those with low pressures. That statement was intended to prove that sodium intake had no e ffec t on hypertension but the populations sampled were not defined as having high sodium intake. Following fu rther lite ra tu re review, he stated that i t seemed reasonable to conclude that humans, lik e Dahl ra ts , may vary in th e ir sa lt sens itiv ity . Another c r it ic a l review by Coleman, Samar, and Murphy, (1979) discussed studies which have conflicting results in findings related to blood flow; an increase in blood flow precedes a peripheral resistance increase in the autoregulation theory. They concluded that the lack of d irec t experimental evidence is prolonging controversy surrounding the theory of sa lt induced hypertension. Some experimental studies in humans have been completed recently. The giving of an 800 mEq sodium load (in great excess of that consumed by Western societies) resulted in a mean blood pressure (calculated from summation of 1/3 of the pulse pressure and the d ias to lic pressure) increase from 83.3 to 100.3 in fourteen men. Failure to use a control group which was not s a lt loaded opened the study to c ritic ism . The effect of testing blood and urine and I.V . f lu id administration was unknown. Controls were included for intake of potassium, phosphorus and drinking water sodium (Lu ft, et a l . , 1979b) 17 A study u t iliz in g sim ilar methodology, smaller sample size and no control group fa iled to demonstrate any increase in blood pressure (Romoff, Keusch, Campese, et a l . , 1979). Epidemiological Studies Papers presented at the Joint U.S.-U.S.S.R. Symposium on Hypertension reported on an investigation of a population liv ing in a region in Russia where two to five times more s a lt than normal (values not given) was present in the drinking water. This sample was divided into those who consumed excessive dietary sa lt (more than 350 mEq/day) and those who did not (150-180 mEq/day). The two groups were further s p lit into healthy and hypertensive sub-groups. Hyper­ tension was found four times more often in individuals who consumed excessive dietary sa lt. These groups were then subjected to experi­ mental treatment; placed in the hospital where a ll received diets of , 9 Gm sa lt per day. Intravenous sa lt load and subsequent diuresis with Furosemide caused decreased sodium and water excretion in the hyper­ tensive group who had used excessive sa lt. This did not occur in persons with hypertension who consumed normal amounts of sa lt (Nekrasova, Gazaryan, Fatula, Suvorov, Zhemerikina and Chernova, 1978). Changes in prostaglandins and renin were seen in healthy high sa lt consumers when compared with the ir low sa lt control group. The report was a 18 translation from Russian and d if f ic u lt to understand. Thus, i t would appear that experimental physiologic studies are s t i l l lim ited in number, scope and consistency of findings in demonstrating any conclusive evidence that sodium intake is a factor in the development of essential hypertension. Few studies considered control fo r many possible intervening variables. Much more evidence supporting sodium's role in hypertension than refuting i t has been published in the las t 20 years, however. Interested public health professionals are continuing to develop studies to demonstrate the effec t of sodium on the mean blood pressure of normal community samples. In an epidemiological study (combined with physiological para­ meters) seven areas with soft water were compared with six areas having hard water supplies. Mean systolic blood pressure, plasma cholesterol and serum sodium were elevated in areas with higher sodium-calcium ra tios . Validation of higher sodium intake by measuring urinary sodium excretion was not done, however. The researcher demonstrated sign ifican t negative correlation between hard water (containing low sodium) and death rates from cardiovascular disease. Cebus monkeys were studied and found to develop high serum sodium levels when given soft drinking water fo r 8 weeks (Toutounchian, 1977). Tu th ill and Calabrese (1979) published a report of th e ir contin­ uing studies aimed at documenting the need for a national standard for drinking water sodium. High school sophomores were compared in a 19 community with 107 mg/1 drinking water sodium content with a matched community having a sodium level of 8 mg/1. I t is generally accepted that patients on a sodium restricted diet would be harmed by the additional unknown intake of over 200 mg per day of sodium in drinking water which contains 100 mg/1 sodium. The purpose of Tuth ill and Calabrese's studies was to determine effects of additional sodium intake on normal healthy human populations. Communities in the study were sim ilar in size, median family income, education, and ethnic composition which was believed to control d ie tary , genetic, and stress factors. Blood pressures were compared in 300 white sopho­ mores. Many variables were studied by questionnaire. An upswing in blood pressure was demonstrated in the high sodium community with the increase higher in females than males. Female systolic and d iasto lic means were 5.11 mm higher in the high sodium community. Male values were 3.58 mm higher for systolic pressure and 2.69 mm higher for d iasto lic pressures. Using one way analysis of covariance, values were adjusted in relation to the 18 variables studied by questionnaire. In males, the difference in d iasto lic pressure means decreased to nonsignificance with the main effec t coming from the variable "time since las t meal". Females' systolic and d iasto lic and male systolic readings remained highly significant following the adjust ment. Two way analysis of covariance did not change those results except that the decrease in mean difference in males' d iasto lic values 20 disappeared. Further studies are planned with th ird graders in which sa lt intake w ill be ve rified by sodium excretion. Summary There are a large number of variables known to a ffec t blood pressure. The development of hypertension is probably of m ulti­ fac to ria l causation. The theoretical framework of th is study is based upon findings that high sodium intake is one of the factors contributing to the development of hypertension and investigates the relationships between sodium levels in drinking water and blood pressure levels in a young, normal population. CHAPTER 3 METHODOLOGY Introduction The methodology of th is study was patterned a fte r that of Tu th ill and Calabrese (1979) to allow comparisons between findings as much as possible. Methods were not rigorously replicated, however. To answer the question whether 14 to 16 year olds drinking high sodium water in the community supply would have higher blood pressures than a control group, blood pressures were measured and information on other factors was collected by a two-part questionnaire. The study design was pre-experimental in nature and consisted of measurement of the depend­ ent variable , blood pressure and selected intervening variables. The independent variab le , drinking water sodium lev e l, had been present for years in the high sodium town (High Na). The town selected fo r sample comparison of blood pressure means had low drinking water sodium. In other words, the study used post-test only, nonequivalent control group design (Waltz and Bausell, 1981). Three measures of blood pressure for each consenting 14 through 16 year old high school student were obtained in December 1980. Measurement of blood pressures fo r each of the sample groups was done 22 on subsequent days and testing was begun and ended at the same time of day for both groups to provide fo r testing day internal v a lid ity . The groups were tested under sim ilar conditions so fa r as possible. The data collectors were not aware what independent variable was being studied. Subjects were not made aware of the experimental-control group differences between water sodium levels although such knowledge could not be ruled out. Additional measures for control of possible intervening variables were obtained by questionnaire. Some items were sent home fo r parental assistance; others were completed at the time of blood pressure measurement for independent student responses. Height and weight were measured a t the time of blood pressure recording. Signi­ fican t intervening variables were controlled by post hoc s ta tis tic a l analysis. Defin ition of Terms For the purposes of th is study, the following defin itions were used: Subjects: High school students who had attained th e ir fourteenth birthday but not the ir seventeenth birthday when the f i r s t questionnaire was distributed . Adolescent hypertension: Subjects 14-16 years old with sustained blood pressure values exceeding or equal to 23 140 mm systo lic or 90 mm d iasto lic in three repetitions as measured in th is study (Weidman, 1979). Systolic blood pressure: That pressure recorded upon auscultation of the f i r s t Korotkoff sound a t the point where the f i r s t of two consecutive sounds was heard. D iastolic blood pressure: That pressure recorded a t the disappearance of sound; the f i f t h Korotkoff sound. Prevalence of hypertension: The rate of students found in screening to have adolescent hypertension per 100 students. Community Characteristics The communities selected for the study differed in drinking water sodium levels . Characteristics of the communities were compared prior to community selection to assure s im ila rity on as many variables as possible. The high sodium community was a t 4245 fee t elevation; the control community was 5280 fee t in a ltitu de . Acclimatization to high, a ltitude may be related to lower blood pressures (Slonin, 1974) but no evidence was found to associate differences in blood pressure with the a ltitude levels between the study communities. Table I displays population d istribu tion for the two communities from 1980 24 Census of Population and Housing Advance Reports. Table I Community Population Characteristics Low Naa High Nab County Population 2154 2359 Town Population 1302 1182 White 1297 1171 Blacks 0 0 American Indian 2 5 Asian 0 2 Other 3 3 Housing Units 574 592 a Sodium level in drinking water 2.1 mg/1 taken during data collection (State Water Quality Bureau) ^ Sodium level in drinking water 378 mg/1 taken during data collection (State Water Quality Bureau) Since 1970, population in the low sodium community increased by 8.5% (102 persons) and the high sodium community decreased by 14.1% (194 persons). A major portion of the population loss for the experimental community occurred when the Milwaukee Railroad ceased operations in the town in March,1980. The economic loss to the community could a ffec t the blood pressures of that community's c h il­ dren since stress and lower socioeconomic status have been associated 25 with hypertension. Data about county economic condition is presented in Table 2 from the Montana Employment and Labor Force Monthly Report (August 1980). The study towns are the major population concen­ tra tion for the ir respective counties. Table 2 County Economic Comparisons Low Na High Na Labor Force 1195 1245 Unemployment Rate 4.0 3.1 6 Month Unemployment Insurance Paid $41,567 $15,815 As can be seen from Table 2, although population and a major source of employment dropped in the high sodium community, unemploy­ ment was higher in the low sodium community in the quarter prio r to th is study. I f unemployment had an e ffec t on adolescent blood pressure, i t would be in a d irection opposite the hypothesis of th is study. Other s ta tis tics about the counties are of in terest to the hypothesis that sodium in drinking water contributes to elevated blood pressure in normal individuals. The death rate from heart disease in 1976 in the high sodium county was 240 per 100,000 population. The low sodium county had a heart disease death rate of 227 per 100,000 26 population comparisons of death rates fo r cerebrovascular disease (stroke) were even more s ign ifican t to a study of hypertension. In the high sodium county, the death rate for stroke was 160 per 100,000 population; whereas in the low sodium county, that rate was 45.5 per 100,000 (Montana V ita l S ta tis tic s , 1976). Sample The sample was drawn from two rural Montana communities which are 58 road miles apart. Potential subjects were fourteen through sixteen year old high school students whose parents consented to allow partic ipation . Three years of age were used to produce an adequate sample size in these rural towns. Following data collection, subjects were further screened for home drinking water not provided by the community supply, for the presence of diabetes, heart or kidney disease, h istory, medication use, or use of bottled water. The e l ­ imination of those cases resulted in a sample size of 33 in the high water sodium community and 34 in the low water sodium community. To form the fin a l sample of healthy subjects whose home was supplied by c ity water, cases were eliminated as shown in Table 3. 27 Table 3 Cases Which Did Not Meet Sample Selection C rite ria Low Sodium Town High Sodium Town Non-city water supply 17 . 21 Bottled water 0 2 History heart disease 4 6 Diabetes 0 I History kidney disease 4 0 Current medication use 4* 4* Some subjects eliminated for more than one reason. Of 97 questionnaires distributed in class in the control group, 61 were returned and 58 students completed the blood pressure testing and completed part two of the questionnaire. In the high sodium town, 78 questionnaires were d istributed , 74 were returned. There were 67 subjects in the experimental town who completed part two and had th e ir blood pressures taken (two of these were eliminated following discovery of lack of parental consent). 28 Procedure Portions of a questionnaire developed by Tu th ill and Calabrese (1979) were used in development of the two-part questionnaire used for data co llec tion . The f i r s t part of the questionnaire was sent home with students, two to three weeks prior to data co llec tion . The second part was completed at the time of blood pressure screening. A copy of both parts of the questionnaire comprise Appendix A. Part One Questionnaire Demographic data were obtained in the f i r s t portion of the questionnaire as were those questions in which parental assistance would improve accuracy. In the interest o f b rev ity , items on Tu th ill and Calabrese's instrument relating to prior experience with blood pressure taking, number of cups of hot chocolate consumed per week (may be milk or water based) and glasses of water consumed with lunch were deleted. The la s t page of Tu th ill and Calabrese's instrument was deleted because students given a p ilo t test (n.-5) had d if f ic u lty completing i t without great assistance from the researcher. Some questions students had d if f ic u lty with were, fo r example: "Have you ever been troubled by 'cold sweats'?", "Do you feel that you are bothered by a ll sorts (d iffe ren t kinds) of ailments in d iffe ren t parts 29 of your body?" Association with blood pressure was unclear fo r some of those questions. Items added to the questionnaire sought further information about family history of hypertension, black lico rice use and the source of the home drinking water supply. The items consisted of mixed multip le- choice and f i l l- in - th e -b la n k questions. Questions sought information about gender, grade in school, liv ing s ituation , head of household education and occupation, age, and length of time lived in the town. General health, physical a c t iv ity , recent gain or loss of weight, specific diseases and medication use were reported. The student was asked to estimate the amount of sa lt added to meals, salty snacks, soft drinks, coffee or tea and water consumption. Part one of the questionnaire was returned in a sealed envelope to the school, then the entire group of returned questionnaires was sealed and sent to the researcher. Confidentiali ty of response was assured and maintained throughout the study by iden tifica tion of each section of data collection by id en tifica tion number. Part Two Questionnaire Other variables have been mentioned in the lite ra tu re as having an effec t upon blood pressure. Personality factors of anger and anxiety were shown to be associated with hypertension in adults (Baer, Collins, Bourianoff, and Ketchel, 1979; McClelland, 1979). A se lf-report 30 instrument developed and validated with adults by Baer et a l . (1979) was adapted for use with adolescents and included in the second part of the questionnaire. Terminology of some items was changed based on a p ilo t study with a small sample of teenagers in a larger town than the study communities, but fu ll-s ca le validation was not carried out. To avoid parental influence in responses to the personality questions and those about smoking and beer consumption, they were administered at school prior to blood pressure screening. Questions to e l i c i t informa­ tion about the prior night's sleep and the length of time since the las t meal were answered in the second questionnaire as well. Blood Pressure Blood pressure screening was done by four volunteer nurses." Their r e l ia b i l i t y in judging blood pressure readings in response to a train ing film was tested. A "Y" stethoscope was used to ve rify the ir findings on ten readings with those of the researcher prio r to the beginning of the study. Every hour o f screening included a recheck of each screener with the dual stethoscope on a random basis. Each screener was given a 10 minute break and returned to a d iffe ren t station every hour following the break. Students remained s ittin g during completion of the questionnaire and sat for one fu ll minute p rio r to in it ia t io n of each of three blood pressure readings taken at each station. The righ t arm was used 31 for a ll readings. Subjects carried a number iden tified booklet which corresponded to the number of th e ir questionnaire. Each blood pressure was recorded on a d iffe ren t page of the booklet to prevent screener bias from previous recordings. (See Appendix B). Readings o f blood pressure were obtained using mercury-gravity manometers which were calibrated to zero prior to both day's screening. Cuff width and length were measured for each subject a t the f i r s t station. The bladder of the cu ff encircled the arm without overlapping and the cuff width was at least two-thirds the length of the arm but not extending into the antecubital space or a x il la . The radial pulse was palpated, the cuff rapid ly in fla ted to a point 30 mm above the disappearance of the radial pulse, then slowly deflated during auscultation throughout notation of systolic and d ias to lic values (Weidman, 1979). Radial pulse rate and dressed height and weight were recorded. Quiet was maintained during pressure recordings by having waiting students contained in an adjoining room in which questionnaires were completed. The students were very quiet and orderly throughout the screening. Precautions were taken to assure s im ila rity of data collection between communities. In scheduling testing days, basketball tourna­ ments and tests were avoided. Since Tuth ill and Calabrese (1979) had found a s ign ifican t e ffec t from the length of time since the las t meal, the same number of students was tested before lunch as a fte r 32 lunch in each town. Data collection was performed December 8, 1980 in the low sodium town and December 9, 1980 in the high sodium town. Water Testing Water samples were taken from the tap at each school on the days blood pressure screening was done in each community. Chemical analysis of the water performed by the State Chemistry Laboratory Bureau showed results lis ted in Table 4. On the day tested, the sodium value fo r both schools was lower than previous community water analysis reports indicated. Table 4 Chemical Analysis of High School Water Parameter: Low Na High Na Sodium 2.1 mg/1 378 mg/1 Potassium .5 mg/1 1.9 mg/1 Calcium 4.8 mg/1 5.2 mg/1 Magnesium 1.6 mg/1 3.6 mg/1 Some authors studying lower hypertension rates in communities with hard water have proposed that high potassium (Walker, Whelton, Saito , Russell, and Hermann, 1979) or calcium, magnesium, and cadmium 33 (Neri and Johansen, 1978) are protective elements for the development of hypertension. I f that hypothesis is true , the high sodium commun­ ity which has higher levels of those elements could have lower blood pressures because of that protective e ffec t. The differences between communities in levels of those ions was not large , however. Cadmium levels in both towns were less than .001 when tested in 1977 (high sodium town) and 1978 (low sodium town). Thus, sodium level in drinking water was confirmed as the independent variable fo r th is study at the time of data co llection . Methods of Analysis The subjects for th is study could not be randomly assigned to an experimental and a control group. Parametric s ta tis tics have random group assignment as a major assumption for the use of those s ta tis ­ tics in generalization to a larger population from a sample. Because of the need to determine whether the observed differences between the groups was a chance happening, and the power provided by parametric s ta tis tic a l procedures, much of the data analysis used those s ta tis ­ t ic s . The target population was composed of 14 through 16 year olds in the two towns and a ll students in that age group had an oppor­ tun ity to partic ipate in the study. Sample selection c r ite r ia were imposed following data co llection . 34 In Stage I analysis, mean blood pressures were subjected to Jt-test to analyze differences between towns fo r systolic and diastolic values for the town samples and by gender. Average systolic was computed using a ll systolic readings as was average d iasto lic to arrive at means fo r each parameter. The samples were divided for gender because Tu th ill and Calabrese's study (1979) had demonstrated differences in between town means when the ir samples were separated fo r males and females. In Stage 2 analysis, intervening variables were analyzed to assess th e ir e ffec t on the main question of difference of mean blood pressures between groups. Intervening variables were studied by correlational methods, using Spearman's rho fo r ordinal data and Pearson £ fo r in terval data to determine those variables which showed sign ifican t correlation with blood pressure. From those studies, multi fac to ria l analysis of variance studied the combined effects of several variables to determine which interactions or main effects were strongest in relationship to the blood pressure measures. A fin a l grouping of strongly in fluen tia l variables for each gender was tested by analysis of covariance which combines analysis of variance and regression techniques (ANCOVA). The dependent variable (systolic or d ias to lic blood pressure) was adjusted to elim inate the effects of factors and covariates other than water sodium le v e l. The s ta tis tic used with ANCOVA was multiple classification analysis which produced 35 a grand mean systolic and grand mean d ias to lic adjusted for factors and covariates fo r each gender. The amount of variation in that grand mean caused by the water sodium variable was then added to or sub­ tracted from the grand mean to give an adjusted mean systolic and d ias to lic pressure, controlled for the effects of other variables fo r each sample group. The level of significance to be met for determining chance occurrence was less than .05 probability . CHAPTER 4 DATA ANALYSIS Sample Description The sample for this study included subjects from the high schools of two rural communities in Montana. One group was from a town with low sodium in the drinking water (2.1 mg/1), the other from a town with high sodium in drinking water (378 mg/1). The subjects' ages were lim ited to those who had passed the ir fourteenth birthday but had not reached the ir seventeenth birthday by the time of questionnaire d is t r i ­ bution. Students who reported a history of heart or kidney disease, medication use, use of bottled water or water from other than c ity sources were not included. There were 34 subjects in the low sodium town who completed both parts of the questionnaire and blood pressure screening. Th irty - three subjects formed the high sodium town group. Breakdown of the two groups by gender is displayed in Table 5. 37 Table 5 D istribution of Subjects by Gender and Group Males Females Total Low Sodium Group 19 14 34* High Sodium Group 16 16 33* * One missing case each town. Although one subject in each town did not respond to the question about subject gender, those two cases were included in those analyses in which variables were correlated with blood pressure without sex breakdown. Tests of blood pressure differences by sex did not include the two missing cases. The age of subjects was s im ilarly distributed between the two towns as shown in Table 6. Table 6 Age Distribution by Town Age in Years 14-14.9 15-15.9 16-15.9 Total N Low Sodium Group 11 10 12 33* High Sodium Group 13 9 10 33 * One missing value 38 Stage I Analysis To test the f i r s t hypothesis that there would be higher blood pressure values in the high sodium group, mean systo lic and mean d iasto lic pressures were analyzed by V te s t fo r significance of d if fe r ­ ence between the town means. Table 7 presents the results of that analysis. There was no s ign ificant difference between the groups for either systolic or d ias to lic mean blood pressures when the groups consisted of males and females together. There was a s ligh t upswing in the mean systolic value in the high sodium group and even less difference was found in the mean d ias to lic value. For females, the systolic mean was 4.7 mm higher in the experimental group; the d ias to lic mean was 5.7 mm higher and was the only s ta tis ­ t ic a lly s ign ifican t difference demonstrated between the towns. Male blood pressures were elevated by 2.1 mm in the high sodium community over the control group but were 3.8 mm lower than the control group fo r d ias to lic mean. The second hypothesis predicted proportionately higher between town difference in th is study than in Tu th ill and Calabrese's 1979 study to show a dose-response relationship with higher levels of sodium in the drinking water. In that study, male systo lic was 3.57 mm higher. Tab le 7 Differences in Blood Pressure Means Low Sodium Group Mean £ High Sodium Group Mean £ Difference of Means t value df £ Value* Sample systolic 108.726 34 111.434 33 +2.708 1.24 65 .110 Sample d iasto lic 67.716 34 68.667 33 + .951 0.48 65 .315 I Female systolic 104.762 14 109.5 16 +4.738 1.48 28 < .075 Female d iasto lic 64.952 14 70.667 16 +5.715 2.30 28 .014* Male systolic 111.747 19 113.583 16 +2.109 0.71 33 .242 Male d iasto lic 70.474 19 66.667 16 -3.807 -1.33 33 .097 * £ = < .05 for one-tailed test 40 d ias to lic was 2.69 mm higher in the high sodium c ity which were sign ifican t differences for the large sample size of th e ir study. In order to show a d irec tly proportional dose-response relationship , male systolic in th is study would have been 12.85 mm higher in the high sodium town, which did not occur. For male d ia s to lic , the dose proportionate level would be 9.68 mm higher, but th is study produced a finding of lower d iasto lic in males in the experimental group. Female systolic would have been 18.4 mm higher i f the second hypothesis held true; d iasto lic also would have been that high. This study used three grades of student; Tu th ill and Calabrese used only sophomores. Sophomore males in th is study showed no difference between towns in systolic pressures and a 5 mm lower d iasto lic mean in the high sodium group. For female sophomores, the results came closer to showing a dose-response re lationship . Female systolic fo r sophomores was 8.587 mm higher in the high sodium town. D iastolic showed a more dramatic increase, 14.221 mm. The numbers of sophomores was too few, however, fo r meaningful analysis by gender. The de fin ition of adolescent hypertension was individuals • with higher than 140 mm systolic or 90 mm d ias to lic blood pressures over three repetitions . Only one subject even came close to presenting a blood pressure that high, and her high reading was only in the f i r s t of the three recordings. That subject was eliminated from the sample because of diabetes and was from the high sodium community. Thus, 41 there was no incidence of adolescent hypertension in e ither community. Stage 2 Analysis Because of the large number of variables and small number of cases, analysis of the intervening variables was done by studying each variable for e ffec t on blood pressure means and/or differences in responses between towns. On the basis of those studies, a l i s t of 45 variables was reduced to nine for further analysis by analysis of covariance and computation of an adjusted mean for each group. Family history of hypertension was expected to be an important intervening variable with the potential of confounding study results. I f sodium had an e ffec t in elevating blood pressures, the high sodium town could be expected to have more adults with diagnosed hypertension Subjects in that town could have been expected to l i s t more family members with hypertension. There was l i t t l e difference between the - two towns in response to those questions. More subjects in the high sodium group had larger numbers of family members with hypertension, but there was l i t t l e difference in numbers of subjects with some family history versus those with no family history of hypertension. Table 8 displays the frequency of number of family members with hypertension. Ten students in each group did not know or did not answer the questions on family history even with parental help in 42 completing the questionnaire. Table 8 Family History of Hypertension Number of Family Low Sodium High Sodium With Hypertension Group Responses Group Responses N Freq. N Freq. 0 5 20.8% 5 21.7% I 12 50.0% 10 43.5% 2 6 25.0% 3 13.0% 3 0 4 17.4% 4 0 I 4.3% 5 I 4.2% O Missing Values = 10 10 Adjusted frequency calculated without missing cases Pearson correlation coefficients were computed to determine the association between subject's blood pressure values and the number in the family with hypertension. Positive correlation was present in each town sample but none reached the level of significance. The low sodium town systolic values ( r = .2847, £ = .089) were more 43 correlated than the high sodium group's systolic (r_= .2813, £= .097). A s light opposite trend occurred in d iasto lic pressures. The control group ( r = .1211, £= .286) had less correlation than the high sodium group (r.= .1294, £= .278). Whether subjects in e ither town drank softened water was of importance to the study since water softening adds about 100 mg of sodium to each l i t e r of water. Only one subject in the sample consumed softened water at home. That student was in the low sodium group. The presence of that one case may have had some influence in elevation of the mean pressure in the low sodium group but the influence would be s ligh t. The following variables were found to be of no significance in one-way analysis of variance with either systolic or d iasto lic subject's pressures: With whom subject liv es , p rio r night's sleep, occupation of the head of the household, and amount of added food sa lt. Ten of the personality factors also had nonsignificant main e ffec t in relationship to systolic or d iasto lic pressure. These were: friends get angrier than subject in sim ilar situations; subject is tense and nervous;: subject is seen as good looking; subject gets even rather than forgiving; subject is not restless; others overlook dangers the subject is concerned about; the subject is under pressure; the subject is c r it ic a l of others; the subject is more anxious; and the subject is comfortable with sex ro le (no students answered "no" to that 44 question). The £ values for those variables studied by analysis of variance which were not s ign ifican tly associated with blood pressure ranged from .007 for the variable about good looks to 2.390 fo r d iasto lic pressure with adequacy of the prior night's sleep. Three other variables were eliminated early in the analysis. The subject's pulse rate showed only a weak nonsignificant correlation with blood pressures and l i t t l e difference in mean pulse rate between towns was present. Mean pulse in the control group was 67.485. In the high sodium group, the mean pulse rate was 68.875. The difference between the two group pulse means para lle ls the difference in blood pressure and is affected by many of the same variables. Data on grade in school was collected fo r the purpose of comparing sophomores with Tu th ill and Calabrese's study but was shown to be nonsignificant in association with blood pressure. The variable , length of time lived in the town,did not bear any relationship to blood pressure in e ither sex or within the high sodium town by gender. Many subjects in both towns lived there a ll th e ir lives . The relationships of other variables to subject's systo lic and d iasto lic pressures was analyzed by Pearson Correlation Coeffic ient. The following variables had nonsignificant correlation with systolic or d iasto lic pressures in e ither town: Subject's age had weak, positive correlation in the control group, weak negative correlation with the high sodium group. The variab le , number of soft drinks per 45 week had the same between group difference in d irection of correlation. No subjects in the high sodium town ate black lico rice but that variable was positive ly correlated with pressures in the low sodium town. The number of water-based drinks such as tea , coffee, frozen ju ice and water correlated in a negative direction in both groups. The number of beers per week also had negative correlation with blood pressure. None of the above variables' co rre la tion with blood pressure was s ta t is t ic a lly s ign ifican t. The r values ranged from +.2464 to -.2043. Ordinal level variables were analyzed with Spearman's rho for correlation with systolic and d iasto lic pressures. The variables, subject's general health se lf ra ting , and physicial a c tiv ity se lf rating were not s ign ifican tly correlated with blood pressure (r. values ranged from +.1269 to -.2541). General health of the subject was rated from excellent to poor and varied in negative correlation with blood pressure in both groups. Physical a c tiv ity was rated from much less active to much more active than friends. As the physical ac tiv ity rating indicated more a c tiv ity , the blood pressure was higher (positive corre la tion ). S ignificant Variables Variables with s ignificant correlation with blood pressure were fu rther analyzed fo r male-female differences. 46 Height and weight were both sign ificant in correlation with blood pressure. Since weight is to some extent a function of height, a height/weight ra tio was computed fo r further analysis between towns and sexes by dividing height in inches by weight in pounds. A t - te s t of height/weight difference in means between groups showed no difference in between town height/weight. The control group mean height/weight ra tio was .5216; the high sodium group mean was .5270 (£= .737). The mean height/weight indicates subjects in the experi­ mental group were s ligh tly thinner than the control group. Pearson correlation was run for each sex in each town. Results are displayed in Table 9. Table 9 Height/Weight Ratio Correlation With Blood Pressures Males Females Low Sodium High Sodium Low Sodium High Sodium Group Group Group Group J i = 19 ji = 17 Iid E= 17 Syst r -.2992 r -.4324 r -.3906 E -.4656 N H O ^s l £= .042* E = .084 E= .030* Dias r -.3315 r -.4901 r -.1461 r -.5361 CO8■I E = .023* CTl OCOIlQl E= .013* * £ = .05 47 Subjects weight for height was s ign ifican t only in the high sodium group, and s ign ifican tly so in both sexes in that group. Negative correlations meant that the thinner the subjects, the lower the ir blood pressure and conversely, the heavier the subjects, the higher th e ir blood pressure. The same correlation direction was present in the control group but not to a s ign ifican t extent. Three personality variables demonstrated s ign ifican t differences between group pressure means when analysis of variance was carried out for each of the personality variables. Displayed in Table 10 are the results of that analysis. There was no two-way interaction between water sodium group and any of the personality variables. Table 10 Significant Personality Variables With Blood Pressure, Total Sample Systolic______ ‘ _____Diastolic ANOVA £ Sig. of £_ JF Sig. of F D islike Center of Attention 9.953 .002* 2.116 .151 Calms Fast 6.150 .016* 1.966 .166 Low Surroundings Awareness 4.073 .048* 2.416 .125 £ = . 05 48 Five of the personality variables which had shown a sign ifican t or near s ign ifican t main e ffec t w ithin each sodium group with blood pressures were further analyzed fo r effects by gender. That analysis identified the variables scapegoating and hides anger as demonstrating a s ign ifican t main e ffec t fo r females. Those two variables were used in analysis of covariance to determine adjusted between town pressure differences for females. Analysis of variance for males singled out the variables "d is like center of attention" and "calms fast" for consideration in the fin a l analysis of covariance to determine pressures fo r each town adjusted fo r s ign ifican t intervening variables. Other variables varied s ign ifican tly with blood pressure by group and by sex within group. The educational level of the head of the household, subject's general health and time since the las t meal variables had demonstrated s ign ifican t correlation with blood pressure when run for each group separately. Those variables were further analyzed using one way analysis of covariance by sodium level groups with the three variables for each sex. For mean systolic pressure in males, the length of time since the las t meal was the most powerful covariate with an IF value of 4.569 (ja= .043). With that combination of variables, the main effec t of the water sodium groups was also s ign ifican t in males (£= 5.296, £= .031). The other two covariates were nonsignificant. For male d ias to lic , none of the three covariates reached the level of significance 49 and the main e ffec t of the water sodium groups was s ign ifican t (R= .729). For females, neither the covariates nor water sodium groups were significant fo r mean systo lic . Female d ias to lic also was nonsigni­ ficant fo r each of the covariates but the grouping by water sodium level was s ign ifican t (£= .013) with an F value of 7.440. Six more variables had shown significant varia tion in e ither systolic or d ias to lic pressures for each group. They were further analyzed by two-way analysis of covariance using sa lty snacks recoded as I = 0-10 per week, 2= 11-25 per week and water sodium group as factors with the covariates weight gain, weight loss, height/weight ra tio , smoking and Vitamin C p i l l use. For male systo lic , none of the covariates was s ig n if ic an t,n o r were the effects of sa lty snack intake or water sodium. There was no two-way interaction between the factors. Male d ias to lic was affected by the covariates to a large extent. The variables weight gain (£_= 9.268, £= .006), smoking (£_= 4.701, £= .041), and Vitamin C p il l (£_= 5.465, £= .029) were significant in male d iasto lic varia­ tion . Vitamin C p i l l use was negatively correlated with mean diasto lic in the control group but did not reach the level of signi­ ficance in the high sodium town. More subjects in the low sodium town took no Vitamin C p i l ls . Water sodium group and salty snacks were nonsignificant but more of the variation from the mean was produced by the salty snacks variable. 50 A good portion of the between town difference in females' systolic and d ias to lic pressures was explained by the covariates, weight loss, height/weight ra tio and smoking. From the l i s t of variables which had shown gender differences in effects on blood pressure, variables were chosen to be combined fo r each sex in a fin a l three-way analysis of covariance. Mean male systo lic and d iasto lic pressures were controlled fo r the interaction of three factors: water sodium group, calms fas t and center of a tten tion , with time since las t meal, weight gain, smoking, height/weight ra tio and Vitamin C p i l l as covariates in three-way analysis of covariance. For females, the factors , water sodium groups, scapegoating and hides anger were analyzed with the covariates, time since las t meal, smoking, weight loss, height/weight ra tio and Vitamin C p i l l . Because the number of degrees of freedom was reduced to one by the combination of such a large number of variables, the amount of variation required to produce significance was higher than fo r the original V te s t of means. The purpose of th is portion of the analysis was to compute a mean fo r each of the water sodium groups a fte r the influence of the s ign ifican t intervening variables had been controlled. The multiple c lass ifica tion analysis s ta tis tic chosen produced a grand mean systolic and d iasto lic value a fte r adjustment by independent variables and covariates. The amount of variation produced by the 51 water sodium group was then added or subtracted from the grand mean to achieve an adjusted mean systolic and d ias to lic for males and females. Tables 11 and 12 show the results of that computation. Table 11 Adjusted Female Mean Blood Pressures Systolic_______ ______Diastolic Orig. Value3 Adj.Valu^3 Orig. Value3 Adj. Valueb Low Sodium Group 104.762 104.18 64.952 63.71 High Sodium Group 109.5 110.10 70.667 71.75 Difference Between Means +4.738 +5.38 +5.714* +8.04* * jD = .05 in t - t e s ta or analysis of covariance^ When the effects of s ign ifican t variables fo r females were removed, the between group differences of means was augmented for systolic and even more so for d ias to lic . That combination of variables used provided 50.4% of the variance explanation in systolic and 61% in d ias to lic (multiple £ squared). Although a fte r adjustment of mean systolic the high sodium group systolic was 5.38 mm higher than the control group mean, that result did not reach significance because the number of degrees of freedom was reduced. 52 Table 12 Adjusted Male Mean Blood Pressures Systolic Adj. Value Orig. Value Adj. Value Orig. Value Adj. Value Low Sodium Group 111.474 111.6 70.474 69.35 High Sodium Group 113.583 113.21 66.667 67.08 Difference Between Means +2.109 +1.61 -3.807 -2.27 The combination of variables used for analysis of covariance reduced the difference in male systolic pressure between the two towns. The d ias to lic pressure difference was reduced as w e ll. There was no s ign ifican t difference between means in e ither the orig inal or adjusted values. That combination of variables produced a multiple r_ squared value of .461 or 46.1% of the variance explanation for systolic and 50.1% of the variance explanation fo r d ias to lic . Inclu­ sion of salty snacks and scapegoating may have reduced the difference in means for males even fu rther, but the ir inclusion was not possible with the small sample size. CHAPTER 5 CONCLUSIONS Based on the findings from the study, support was demonstrated fo r an association between high sodium levels in drinking water and increased d iasto lic pressure fo r female adolescents when compared with a sim ilar group. The to ta l sample groups of 14 to 16 year old rural high school students did not support such an association, nor did between group differences fo r males. The f i r s t hypothesis proposed that there would be an increase in mean systolic and mean d ias to lic blood pressure values in the high sodium group over the low sodium group. There was an increase of 2.708 mm for systolic and .951 mm d ias to lic but that increase was not s ta tis t ic a lly s ign ifican t, thus, the f i r s t hypothesis was rejected. When the sample was divided by gender, female systolic and d iasto lic pressures were increased by 4.738 mm and 5.715 mm respectively;, the d iasto lic pressure difference was the only one s ta t is t ic a lly s ig n if i­ cant. Male systolic pressure varied in the direction hypothesized but male d iasto lic pressure varied in a direction opposite the hypothesis. Neither systolic nor d ias to lic mean differences were s ign ifican t for males. 54 When between group variation fo r s ign ifican t intervening variables was controlled, between group differences fo r both systolic and d iasto lic pressures for females became larger but d iasto lic pressure was the only measure which had s ta t is t ic a lly s ign ifican t results (8.04 mm difference between groups). The main source of variation for females was the height/weight ra tio fo r systolic pressure and the length of time since the las t meal fo r d ias to lic pressure. Control of intervening variables for males reduced the difference between group means for both systolic and d ias to lic . The main source of variation for males was produced by a personality fac to r, the student's a b il ity to calm down quickly a fte r anger (systo lic ) and recent weight gain (d ias to lic ) also had a s ign ifican t e ffec t. The second hypothesis stated that there would be a proportional increase in pressures from findings of Tu th ill and Calabrese's (1979) study of sophomores. Dose-response relationships fo r drugs are rarely present in a d irec t proportion so th is hypothesis and study design was inadequate to address that question. When sophomores from th is study were compared to sophomores in the Tu th ill and Calabrese study,, a marked upward sh ift was shown in the town with 378 mg/1 over Tu th ill and Calabrese's community water level of 107 mg/1, but again only for females. The sample size of th is study was too small fo r va lid comparisons between sophomores. Comparison of the between town differences fo r the two studies are 55 displayed in Table 13. The second hypothesis was rejected on the basis of these findings. Table 13 Comparison of Findings With Those of Tu th ill and Calabrese (1979) Na = 378 mg/1* (14-16 yr olds) Na = 107 mg/ (sophomores) Systolic Male +2.109 +3.58 Female +4.738 +5.11 Diastolic Male -3.807 +2.69 Female +5.715 +5.11 a Author's findings for difference between group means k Tu th ill and Calabrese findings for difference between group means Since no subject met the study de fin ition of adolescent hyper­ tension, the th ird hypothesis that there would be higher numbers of students with hypertension in the high sodium group was rejected. 56 Discussion Blood pressure curves for age (National High Blood Pressure Education Program, 1977) show the 50th percentile mean for th is age group to be 122/76 for males, 119.67/76 fo r females. The adjusted means fo r subjects in the study were 113/68 fo r males and 110/72 fo r females in the high sodium town which illu s tra te s that although there was an upward trend in blood pressure in the high sodium town when compared with the low sodium town, the study populations had lower pressures than the national figures. Lower pressures than national figures and those of Tu th ill and Calabrese's study (1979) could be a re flec tion of rural-urban differences or the res tric tion of th is study population to subjects with no history of heart or kidney disease, diabetes, or medication use. There was weak negative correlation between age and blood pressure in the high sodium group which did not f i t with national findings of an upswing in blood pressure from 14 up to 16 year olds (National Heart, Lung and Blood In s titu te , 1977). Since mean pressures of the combined sexes in the high sodium group were elevated over those in the low sodium group, the negative correlation with age may indicate that younger subjects in the high sodium group had higher pressures than older subjects. The poss ib ility of gradual adaptation 57 to high sodium levels is raised by that find ing, but not supported by correlation of blood pressure with the length of time lived in the town. More females in the high sodium sample than the control group exceeded the 50th percentile figures (six to one) which may suggest confirmation of studies which implicate variations in ind ividual's sa lt sens itiv ity . Those subjects would appear to be a t greater risk for the future development of essential hypertension. The influence of personality factors in blood pressure of adolescents was an interesting finding. Factors of attention seeking, anger arousal, anger suppression and punishment avoidance were associated with blood pressure changes in d iffe ren t ways between groups and between males and females. The instrument used fo r measure­ ment of personality factors was not tested for r e l ia b i l i t y and v a lid ity with an adolescent study population. Future studies of th is type should include such testing fo r development of an instrument appro­ priate to the age group being studied. Another variable found to have a powerful influence in th is study was the length of time since the subject's las t meal. Some subjects had not eaten fo r over 20 hours. Tu th ill and Calabrese (1979) also found that variable to be of greatest influence on blood pressure and its between-group differences. Other variables with effect in th is study were: height/weight ra t io , smoking, weight gain 58 of 20 or more pounds in the las t year (males), amount of Vitamin C p il l intake, and weight loss of 10 or more pounds in the las t year (females). The rationale fo r variation in blood pressure from Vitamin C p i l l intake is unclear as the amount of sodium contained in such tablets is not l ik e ly to cause a sodium-based e ffec t. No lite ra tu re was found to support th is finding and i t may be a spurious one. A possible explanation for study findings is that differences between groups on intake of potassium, calcium and magnesium were not studied. Water content of those ions was higher in the high sodium community. I f those ions tend to lower blood pressure as some have noted, dietary and water intake differences between the two groups may have dampened the blood pressure response to sodium in the high sodium group. Further studies should take those ions into account when controlling for between group varia tion . Limitations A major lim ita tion to th is study was inadequate sample size to allow for n-way analysis of covariance for the large number of variables which have some e ffec t on blood pressure. The approach taken, to individually analyze the variables and combine those found to be s ign ifican tly related to blood pressure, contained a factor of 59 error poss ib ility that would not be present with larger sample size and analysis procedures. Precautions were taken to reduce bias in variable selection but the possib ility of error remains. Participation in the study was very high from the high sodium town. The questionnaires were returned at a rate of 62.9% from the control group; the response rate in the high sodium community was 94.9%. Such a high response rate is unusual fo r studies of th is kind and may indicate e ither that greater pressure fo r questionnaire return was exerted by the school in that community or that subjects were aware of the meaning of the study for the ir community. Either poss ib ility may have had an influence on blood pressures of the subjects in that town so findings of this study must be interpreted in that l ig h t. The resu lt of the differences in response rate between the two towns did allow nearly equal sample size , however. R e lia b ility and v a lid ity testing was not done on the instrument fo r data co llection on the intervening variables. Se lf-report data on the variables smoking and beer drinking seemed to be the least dependable measures in the instrument. There were low numbers of positive response for each of those variables which seemed inconsistent with the peer group nature of smoking and beer drinking in the age group studied. The high sodium group had only one smoker, the low sodium group had three. The relationship between drinking water sodium levels demonstrated 60 for females could possibly be explained on the basis of hormonal relationships with sa lt intake. Study of ther menstrual cycle and use of b irth control p i l ls was not possible because of school restraints on study approval. Further studies of the interaction between sa lt intake, blood pressure and hormonal influences could further c la r ify the role of drinking water sodium in community populations. Recommendations This study does lend some support to requests for national standards for drinking water sodium levels . The problem could be approached from a perspective of iden tifica tion of those individuals who are sa lt sensitive instead of reducing the sodium in community drinking water. That approach, however, is only preventative fo r problems related to hypertension development. Many other problems associated with high levels of sodium in drinking water would remain. Those with established congestive heart fa ilu re , edema from other causes, as well as in fants, sick or w e ll, athletes and the elderly would remain a t r isk for exacerbation of health problems by high levels of water sodium. Community education about the risks associated with high sodium intake and ways to reduce dietary intake seems imperative since water treatment, i f a standard were imposed. 61 remains years away. There are d is tin c t lim itations to the community education approach, pa rticu la rly in the area of compliance achievement. Reduction in sa lt intake is d i f f ic u l t a t best for those persons who have serious disease and can see the results of fa ilu re to reduce th e ir sodium intake. Chances fo r a high success rate fo r community education programs would surely be reduced fo r those who are unable to see any clear benefit to th e ir immediate health. In the long run, a water sodium standard and community water treatment to remove excessive levels may be indicated. Studies are needed which document the magnitude of health problems associated with high water sodium content. Many of those problems are known to ex ist and th e ir presence suggests the need to mobilize education programs in the state to provide information on communities with sodium levels over 100 mg/1 to health care providers in those communities as well as providers in health care re fe rra l centers in the state . Maps of sodium levels should be distributed to physicians, d ie tic ians , nurses and hospitals. Such maps and supporting lite ra tu re could begin to reduce the risks fo r patients with health problems exacerbated by high drinking water sodium intake. There is a need to demonstrate dose-response relationships for sodium's e ffec t on blood pressure. Well designed studies should include matching samples fo r disease history and other variables with a large sample size. I t is d i f f ic u l t , i f not impossible, however, to 62 iden tify a community with a large population base which homogeneously consumes drinking water with sodium levels over 300 mg/1. The search fo r dose-response to sodium may be more f r u i t fu l ly carried out under experimental conditions rather than epidemiological ones. Further studies of relationships between variables and blood pressure are needed. Studies of hypertension in association with any set of variables should be analyzed with males and females as separate groups. Pooling data from both sexes cannot c learly explain variable relationships. Further studies of personality variables and establishment of cause-effect relationships with blood pressure are indicated. Because the major finding of th is study was elevated d iasto lic pressures for females, in teraction between water sodium levels and hormonal influences, such as menstrual cycle, b irth control p i l ls , and blood pressure should be studied fu rther or controlled in future studies of the effects of drinking water sodium levels and blood pressure. Studies of th ird graders, conducted by Tu th ill and Calabrese, may more c learly demonstrate the e ffec t of sodium on blood pressure levels when published. Stimulant and depressant drug use would also be controlled through study of th ird graders. • In conclusion, th is study has shown a difference in mean dia­ s to lic blood pressure means fo r adolescent females between two rural towns which had marked differences in drinking water sodium levels. No other results or measures were s ign ifican t. The results must be interpreted with consideration for the problems associated with small sample size and the differences in response rates between the two study groups. Study recommendations provided consideration fo r other health problems which may be affected by high levels of water sodium and ways health risks could be reduced. Recommendations fo r further study were also made. 63 REFERENCES 65 REFERENCES CITED Abramowicz, M. (Ed .). Salt res tric tion to prevent hypertension. The Medical Letter on Drugs and Therapeutics, 1980, 22 (3 ) , (Issue 550), 14-16. Baer, P. E ., Collins, F. H ., Bourianoff, G. G ., and Ketchel, M. F. Assessing personality factors in essential hypertension with a b rie f se lf-report instrument. Psychosomatic Medicine, 1979, 41 (4 ) , 321-329. — Bianchi, G., Fox, U ., DiFrancesco, G. F ., Giovannetti, A. M., and Pagetti, D. Blood pressure changes produced by kidney cross­ transplantation between spontaneously hypertensive rats (SHR) and normotensive rats (NR). C lin ical Science of Molecular Medicine, 1974, 47, 435-438. Bing, R. F ., Thurston, H ., and Swales, 0. D. Salt intake and d iu re tic treatment of hypertension. Lancet, 1979, 2, 121. Borhani, N. 0. Mortality trend in hypertension. United States, 1950- 1976. Proceedings of the Conference on the Decline in Coronary Heart Disease M o rta lity , U. S. Department of Health, Education and Welfare, National Ins titu tes of Health, May, 1979, 218-234. Calabrese, E. J. and T u th i l l , R. W. Sources of elevated sodium levels in drinking water...and recommendations fo r reduction. Journal of Environmental Health, 1978, 41_ (3 ) , 151-155. Coleman, T. G ., Samar, R. E ., and Murphy, W. R. Autoregulation versus other vasoconstrictors in hypertension - a c r it ic a l review. Hypertension, 1979, 1_, 324. (Abstract) Corbet, W. T ., Kuller, L. H ., Blaine, E. H ., and Damico, F. J. U tiliza tio n of swine to study the risk factor of an elevated s a lt d ie t on blood pressure. The American Journal of C lin ical Nu trition , 1979, 32, 2068-2075. Dahl, L. K. Salt and hypertension. American Journal of C lin ical Nutrition , 1972, 25, 231-234. • Edmondson, R. P ., Thomas, R. D ., H ilton , P. J . , Patrick, J . , and Jones, N. F. Abnormal leucocyte composition and sodium transport in essential hypertension. Lancet, 1975, 1_, 1003-5. 66 Groer, M. E ., and Shekleton, M. E. Basic Pathophysiology - A Conceptual Approach. St. Louis: Mosby, 1979. Guyton, A. C. Textbook of Medical Physiology. Philadelphia: Saunders. 1976. Haddy, F. J. Local control of vascular resistance as related to hyper­ tension. Archives of Internal Medicine, 1974, 133, 916-931. Lu ft, F. C ., Grim, C. E ., Fineberg, N ., and Weinberger, M. D. Effects of volume expansion and contraction in normotensive whites, blacks, and subjects of d iffe ren t ages. C ircu lation , 1979, 59, (4 ) , 643-650. (a) Lu ft, F. C ., Grim, C. E ., Fineberg, N ., and Weinberger, M .C., Norepinephrine values with sa lt loading. Hypertension, 1979, I , 262-265. (b) “ Madden, J. A ., Smith, G. A ., and Llaurado, J. G. Sodium d istribu tion in mesenteric a rte r ia l wall of rats with hypertension induced by drinking saline. C lin ical Science, 1979, 56, 471-478. McClelland, D. C. Inhibited power motivation and high blood pressure in men. Journal of Abnormal Psychology, 1979, 88, (2 ) , 182-190. Montana V ita l S ta tis tic s . State Department o f Health and Environmental Sciences, Bureau of Records and S ta tis tic s , 1976. Montana Water Quality Bureau. Water quality parameters and th e ir significance. No date given. National Academy of Sciences. Drinking Water and Health. Washington, D.C.: National Academy of Sciences, 1977. National Institu tes of Health. Proceedings of the Conference on the Decline in Coronary Heart Disease M o rta lity . Washington, D.C.: U. S. Department of Health, Education and Welfare, 1979. Nekrasova, A. A ., Gazaryan, G. A ., Fatula, M. I . , Suvorov, Y. I . , Zhemerikina, E. U ., and Chernova, N. A. Humoral systems in healthy subjects and hypertensive patients with prolonged excessive consumption of sa lt. F irs t Joint U.S.-U.S.S.R. Symposium on Hypertension, 1978, 283-294. N eri, L. C. and Johansen, H. L. Water hardness and cardiovascular m orta lity . Annals of the New York Academy of Sciences, 1978, 304, 203-221. 67 Pickering, G. Salt intake and essential hypertension. Cardiovascular Reviews and Reports, 1930, 1_, 13-17. Romolt , M. S ., Keusch, G ., Campese, V. M., e t a l. Effect of sodium intake on plasma catecholamines in normal subjects. Journal of Clinical Endocrinology and Metabolism, 1979, 413, ( I ) , 26-31. Simpson, F. D., Waal-Manning, H. J . , BolI i , P. et a l. Relationship of blood pressure to sodium excretion in a population survey. Clinical Science of Molecular Medicine, 1978, 55, 373. (Cardio­ vascular Reviews and Reports, 1980, %, 13-17. Slonin, N. B. (ed .) Environmental Physiology. St. Louis: C. V. Mosby, 1974. State Water Quality Bureau. Water Analysis #77W2676. Montana State Department of Health.and Environmental Sciences, 1977. Thomas, G. W., Ledingham, J. G. G ., B e ilin , L. J . , e t a l. Reduced renia in hypertension. Kidney I n t . , 1978, 1_3, 513. (Cardiovas­ cular Reviews and Reports, 1980, 1_, 13-17.) Tobian, L ., and Binion, J. Tissue cations and water in a rte ria l hypertension. C ircu la tion , 1952, 754-758. Tobian, L ., Lange, J . , Azar, S ., Iw a i, J . , Koop, D. and Coffee, K. Reduction of in tr in s ic na triu re tic capacity in kidneys of Dahl hypertension-prone ra ts . Abstracts C ircu la tion , 1977, 55 and 56, I I I-2 4 0 . (Abstract) Toutounchian, S. P. The e ffec t of electro lytes in drinking water on various parameters of the cardiovascular system of man. Dissertation Abstracts In ternationa l, 1977, 38 (11 ), 5305B. T u th il l , R. W. and Calabrese, E. J. Elevated sodium levels in the public drinking water as a contributor to elevated blood pressure levels in the community. Archives of Environmental Health, 1979, 34, (4 ) , 197-203. U. S. Bureau of the Census. Census of Population and Housing Advance Reports, 1980. U. S. Environmental Protection Agency. A Manual fo r Evaluating Drinking Water Supplies. Washington, D .C ., O ffice of Water Supply, 1971. U. S. Environmental Protection Agency. National Interim Primary Drinking Water Regulations. Washington, D.C.: O ffice of Water Supply, 1977. Walker, W. G., Whel ton, P. K ., Sai to , H ., Russel I , R. P ., and Herman, J. Relation between blood pressure and ren ia , renia substrate, angiotensin I I , aldosterone and urinary sodium and potassium in 574 ambulatory subjects. Hypertension, 1979, 1_, (3) 287-291. Waltz, C. and Bousell, R. B. Nursing Research Design, S ta tis tics and Computer Analysis. Philadelphia: J. A. Davis Company, 1981. Weidman, W. H. High blood pressure (hypertension) and the school-age ch ild . The Journal of School Health, 1979, 49, (4 ) , 213-214. 68 Your 500 Milligram Diet. New York: American Heart Association, 1957. APPENDICES 70 APPENDIX A. PARENTAL PERMISSION SLIP TEENAGE BLOOD PRESSURE SURVEY PARENTAL PERMISSION SLIP I hereby give my consent fo r the partic ipation of my child in a research project on (name) rural adolescent blood pressures to be carried out in the school on _____________________ . I understand that my child w ill have (date) his/her blood pressure taken and answer a two-part questionnaire. The f i r s t part of the questionnaire is attached and w ill be completed a t home. The second part w ill be completed when the blood pressures are taken and includes questions on smoking, prior night's sleep and response to stresses. A copy of those questions is in the school o ffice should I wish to review them. My ch ild 's height and weight w ill be measured. The information collected w ill only be iden tified by assigned number; my ch ild 's name w ill not appear on any of the data analysis or reports. Should my ch ild 's blood pressure be found to be elevated, the information should be reported to: ________. ____________________________________________________ and myself. (doctor's name) Signed: (Parent or guardian) Date: __________________________________ I do not want my child to partic ipate : _______(check) I f permission is not granted, please return the blank form to the school,. 71 APPENDIX B. TEENAGE BLOOD PRESSURE SURVEY - PART ONE TEENAGE BLOOD PRESSURE SURVEY - PART I You are being asked to vo luntarily partic ipate in a study of factors which influence blood pressure levels in young people. Patterns of blood pressure in young people have been shown to con­ tinue into adulthood. I f factors affecting blood pressure at a young age can be id en tified , health measures can be developed to decrease problems of high blood pressure which may occur in la te r l i f e . Please answer the questions care fu lly . Your response to these questions w ill be handled with con fiden tia lity . The data w ill be punched onto computer cards without your name,and identifying inform­ ation w ill be destroyed. All of the questions included have been shown in isolated studies to have a possible e ffec t on blood pressure and that is the only reason for th e ir inclusion. Completion of the questionnaire indicates consent by the student to partic ipate in the study. Parental consent must be completed on a separate form. Parents may assist with th is form. 1 - 4 ___________ I.D . Number (Leave Blank) Full Name________________________________________ (Student) Address ___________________ ________________________________ 5 I Male (C ircle Number) 6-7 Grade __________ 2 Female 8 I Mother With whom do you live? 2 Father (Student) 3 Both parents 4 Other (specify) ________________________________________ 72 9 What is the major occupation of the head o f your household? Job T it le _________________________________________________ What are the job duties? __________________________________ 10 How much schooling did the head of your household complete? I None 6 Some College or Tech School 2 Grade School 7 College Graduate 3 Junior High 8 More than College Graduate 4 5 Some High School High School Graduate 9 Unknown 11-14 Your date of b i r t h _______________Age?________and_______ (Student's) Years Months 15-16 How many years have you lived in this town? 17 1 Excellent 2 Good 3 Fair 4 Poor Compared to others of your age and sex, would you say your health is ........ (CIRCLE NUMBER) 18 1 Much less active 2 Somewhat less active 3 About the same 4 Somewhat more active 5 Much more active Compared to others of your age and sex, how would you rate yourself in the amount of physical a c tiv ity you get? 19 I Yes 2 No Have you gained more than twenty pounds in the las t 12 months? 20 I Yes 2 No Have you lost more than ten pounds in the la s t 3 months? Why? ____________________ 73 21 I Yes Have you been on a low sa lt d ie t prescribed 2 No by a doctor? 22 I Yes Did a doctor ever te l l you that you had 2 No diabetes? (Sugar in your urine?) 23 I Yes Has a doctor ever told you of any trouble 2 No with your kidneys? (Not the same as a bladder in fec tion ). IF YES, what was wrong? 24 I Yes Have e ither of your natural 2 No Mother parents been told by a doctor 3 Don't know that they have high blood pressure? 25 I Yes 2 No Father 3 Don't know Write "Yes" or "No" i f any of the following in your family have been told they have high blood pressure. Grandparent Brother Grandparent Brother Sister Sister Grandparent Sister 27 I Yes 2 No Did a doctor ever te l l you that you had high blood pressure? 28 I Yes 2 No Have you ever taken high blood pressure medicine prescribed by a doctor? What medicine?29 I f Yes: 74 30 1 Yes 2 No Have you ever been to ld by a doctor that you had anything wrong with your heart? 31 I f Yes: What was wrong? 32 1 Yes 2 No Are you currently taking any medication or drugs? 33-35 I f Yes: Please l i s t medications: 36 1 Yes 2 No Do you take Vitamin C p i l ls daily? 37 I f Yes: How many a day? 38 1 None When you eat your meals, how much 2 A l i t t l e b it sa lt do you add to your food? 3 Quite a b it 39-41 For the do you I snack items indicated below, how many to ta l times eat these items? Potato chips, corn chips, Doritos, corn curls , salted nuts, sunflower seeds, popcorn, pretzels, french fr ies _______Total # times/week Other salty snacks_________ _______ (Specify) _______Total # times/week 42-43 How many bottles or cans of soft drinks do you consume each week? _______Number/week 75 44-45 How many sticks or pieces of black lico rice do you eat each week? _______Number/week 46-47 How many cups of coffee or tea do you drink each week? _______Number/week 48-49 How many glasses of water from the faucet do you drink each week? _______Number/week 50-51 How many glasses of frozen ju ice or Tang do you drink each 52-53 How many times do you drink from the school drinking . fountain each week? _______Number/week week? Number/week 54 1 Yes 2 No Does your fam ily drink bottled water? 55 1 Yes 2 No Is there a water softening unit fo r your home drinking water supply? 56 I City water supply Where do you get your home water supply? 2 Well (less than 100 fee t) 3 Well (more than 100 fee t) 4 Spring 5 Other (Specify) I .D . NUMBER 76 TEENAGE BLOOD PRESSURE SURVEY - PART I I 56 I Yes Did you have a good night's sleep 2 No las t night? I f no: How many hours of sleep did you get? Was being upset about something the reason you d idn 't sleep well? Yes No 57 I Never smoked 2 Tried smoking but d idn 't continue 3 Used to smoke regular 4 Now smoke regularly 58-59 I f you NOW smoke, how many months have you been a regular smoker? Months 60-61 I f you USED TO smoke, how many months ago did you stop? Months 62 I Yes My friends seem to get angrier than I 2 No in sim ilar situations. 63 I Yes When I'm mad, I calm down faster than • 2 No my friends do. 64 I Yes Some of my friends view me as tense 2 No and nervous. 65 I Yes Others would say I am rather good 2 No looking or a ttrac tive . 77 66 I Yes People are usually angrier than they 2 No are w illing to admit. 67 I Yes I am not as aware of things around me 2 No as others seem to be. 68 I Yes I sometimes try to get even, rather than 2 No forgive and forget. 69 I Yes Most of the time I am comfortable in my 2 No role as a male/female. 70 I Yes I would not consider myself somewhat 2 No restless. 71 I Yes I c learly do not lik e to be the center 2 No of attention . 72 I Yes People tend to overlook situations I 2 No consider dangerous. 73 I Yes I would never think of le ttin g someone else 2 No be punished fo r my wrongdoings. 74 I Yes I often feel under pressure even though 2 No others do not notice that I feel that way. 75 I Yes I am secretly quite c r it ic a l of others. 2 No 76 I Yes I am more anxious or nervous than the 2 No average person my age. 77 I drink number of beers or wine each week. 78-81 The la s t meal I a te was ____ hours and _____m inu tes ago. APPENDIX B. SCREENING BOOKLET 78 xxxxxxxxxxxxxxxxxxxx ________________________________________________________________________ xxxxxxxxxxxxxxxxxxxx Height Height xxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxx ______________________________ xxxxxxxxxxxxxxxxxxxx Are xxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxx xxxxxxxxxxxxxxxxxxxx I . D . N u m b e r X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X Sys to l ic mi Hg D ia s to l ic Nurse t □ ' C Sys to l ic era Hg D ia s to l ic | I I e ra Hg :I Nurse I □ I B I I X X X X X X X X X X X X X X X X X X X X I X X X X X X X X X X X X X X X X X X X X I t X X X X X X X X X X X X X X X X X X X X I X X X X X X X X X X X X X X X X X X X X I . xxxxxxxxxxxxxxxxxxxx 1 xxxxxxxxxxxxxxxxxxxx 1 X X X X X X X X X X X X X X X X X X X X I xxxxxxxxxxxxxxxxxxxx 1 xxxxxxxxxxxxxxxxxxxx 1 xxxxxxxxxxxxxxxxxxxx ' Size Ch i ld I I I Cuff Adu lt I 2 I Pulse Rate I per Bin Systo lIc era Hg D ia s to l ic I I I Itm Hg □ !A Nurse I *@*TA* STATl UWIVEItSTT LtMAtrs