Prenatal Programming of Hypertension and Cardiovascular Disease

Barker and colleagues focused on the association of small-for-gestational-age infants and cardiovascular outcomes in several studies. These investigators focused on cities in England where there were excellent birth and death records. In Hertfordshire there were 5654 male births between 1911 and 1930. The death rates from cardiovascular disease in the ensuing 70 years was greatest in men who were less than 5.5 pounds at birth. There appeared to be a protective effect of being a large baby. At 1 year of age, those children who were less than 18 pounds had a threefold greater risk of death from ischemic heart disease than those greater than 27 pounds. A similar relationship between low birth weight and death from cardiovascular disease, but not other diseases, was also found for women born in Hertfordshire during this time. In studies performed in Preston, England, examining the relationship between birth weight and blood pressure measured in men and women aged 46 to 54 found that infants born less than 5.5 pounds had on average an 11 mm Hg greater blood pressure than those born at 7.5 pounds or more. The effect of being small for gestational age on blood pressure is even apparent in children. For every 1-kg decrease in birth weight, there is a 1.3-mm Hg increase in systolic blood pressure and 0.7-mm Hg increase in diastolic blood pressure at 7 years of age.

It does not appear that one can compensate for fetal undernutrition with postnatal caloric supplementation. Nutrient supplementation in small-for-gestational-age neonates by increasing the protein content of their formula resulted in higher blood pressures when measured at 6 to 8 years of age compared with nonsupplemented infants. Another study also found that accelerated early postnatal weight gain resulted in an increase in blood pressure in small-for-gestational-age infants when blood pressure was measured as adults. The greatest risk for hypertension is in those who are small at birth and who have rapid catch-up growth in childhood.

Perhaps the best studied population is adult offspring whose mothers were exposed to the Dutch famine that occurred between late November 1944 to early May 1945. During World War II the Germans cut the daily rations on the Dutch population from 1600 Kcal to 400 to 800 Kcal, in retaliation for a railroad strike that was imposed by the Dutch government in exile to prevent the Germans from reenforcing their troops after the invasion at Normandy. The offspring of the mothers who endured the famine have been compared with those whose mothers were pregnant before or after the famine. Approximately 50 years after the famine the offspring were found to have an increased likelihood of glucose intolerance, microalbuminuria (a harbinger of renal disease), an atherogenic lipid profile, and obesity. Although not all neonates exposed prenatally to the famine were small for gestational age, those who were of low birth weight had a 2.7-mm Hg increase in blood pressure for every 1-kg decrease in birth weight. Thus, in total, exposure to the Dutch famine increased the likelihood of having the metabolic syndrome, although the severity of the organs involved was dependent on the time during gestation when maternal exposure to the famine occurred. Importantly, those with prenatal exposure to the Dutch famine had a threefold increase in death from coronary artery disease. Similarly, those exposed to the Holocaust during early life have an increased incidence of congestive heart failure, angina, dyslipidemia, diabetes mellitus, and malignancy compared with those who were not Holocaust survivors. The observation that small-for-gestational-age infants are at risk for hyperlipidemia, hypertension, noninsulin-dependent diabetes mellitus, and cardiovascular disease has been extended to insults from multiple different causes and in diverse ethnic populations.

Prenatal Programming of Nephron Endowment: Chronic Kidney Disease and Hypertension

The dogma that there are 1 million nephrons in each kidney is likely not correct. Determining nephron endowment with any precision is technically challenging and complicated by the fact that there is a loss of glomeruli as we age. It is apparent that the number of glomeruli is affected by sex, race, age, and importantly by birth weight. Rather than the 1 million nephrons in each kidney in every human, as reviewed by Hoy et al., there is a tremendous variability in the number of nephrons per kidney, ranging from just over 225,000 to more than 2 million. For each kilogram increase in birth weight, there is approximately a 250,000 increase in glomerular number. A study examined the effect of birth weight on kidney development in neonates born after 36 weeks of gestation, a point when nephrogenesis is complete, in infants who died within the first 2 weeks after birth from nonrenal causes. As shown in Fig. 9.1 , there was a direct relationship between glomerular number and birth weight and an inverse relationship between birth weight and glomerular volume. The increase in glomerular capillary surface area and glomerular volume in the remaining nephrons compensates for the reduction in nephron mass.

Relationship between weight at birth and glomerular number (A) and glomerular volume (B) in infants of various weights born after 36 weeks’ gestation when nephrogenesis is complete.

(From Manalich R, Reyes L, Herrera M, Melendi C, Fundora I. Relationship between weight at birth and number and size of renal glomeruli in humans: a histomorphometric study. Kidney Int. 2000;58:770-773.)

A decreased nephron number has been postulated to be a harbinger for hypertension and a progressive fall in glomerular filtration rate. The Brenner hypothesis is depicted in Fig. 9.2 , which shows that a reduction in nephron number from any means, including a reduction of nephron endowment from low birth weight, sets up a progressive process leading to further nephron loss. The mechanism for the effect of renal mass on glomerular hemodynamics has been studied in rats because some strains have surface glomeruli that are amenable to micropuncture. A reduction in nephron number has been shown to increase glomerular capillary pressure, resulting in proteinuria and eventually glomerular sclerosis. Glomerular sclerosis leads to a further fall in glomerular capillary surface area, leading to a vicious cycle resulting in progressive decline in glomerular filtration rate. The remaining unaffected glomeruli hypertrophy, which is a temporizing adaptation to maintain the glomerular filtration rate, but these glomeruli develop an increase glomerular capillary pressure and will eventually undergo glomerular sclerosis as well.

Pathogenesis of chronic kidney disease and hypertension as a result of reduced glomerular number.

(Adapted from Brenner BM, Garcia DL, Anderson S. Glomeruli and blood pressure. Less of one, more the other? Am J Hypertens. 1988;1:335-347.)

There is substantive evidence that low birth weight infants are at risk for chronic kidney disease. With a reduction in nephron mass from any cause, there is an increase in glomerular capillary pressure resulting in proteinuria or albuminuria as the first sign of progressive renal disease. Prenatal exposure to the Dutch famine resulted in an increased risk of microalbuminuria when studied at approximately 50 years of age compared with age-matched controls (12% vs. 7%). In a study comparing males with a birth weight of less than 2500 g to those with a birth weight of 2500 to 4499 g, low birth weight was associated with an increased likelihood of developing chronic kidney disease defined as a glomerular filtration rate of less than 60 mL/min per 1.73 m ² or a urinary albumin to creatinine ratio of ≥30 mg/g. There was no association of birth weight and the risk of chronic kidney disease for females in this study. In the southeastern United States a study examined the birth weight of adult dialysis patients and compared the weight with age-, sex-, and race-matched controls. A birth weight less than 2.5 kg was associated with a 1.4-fold greater risk of developing end-stage renal disease compared with those weighing 3 to 3.5 kg at birth. The results did not depend on the cause for the end-stage renal disease. In Norway the relative risk for having end-stage renal disease was 1.7 (confidence interval 1.4 to 2.2) for those below the 10th percentile compared with those born between the 10th and 90th percentile. This was true for both men and women. A meta-analysis examining 31 studies found that low birth weight was associated with a 70% greater risk of developing chronic kidney disease.

According to the Brenner hypothesis the decline in glomerular filtration rate should be progressive with age. However, there is some evidence that a decrease in renal function may be apparent in children. In African-American children but not in non–African-American children, with an average age of 1.5 years, there was a lower estimated glomerular filtration rate in children of low birth weight (82 vs. 95 mL/min per 1.73 m ² ). Other studies have found an association between low birth weight and a decrease in estimated glomerular filtration rate in white school-age children as well.

There is also evidence that a low nephron endowment is a risk factor for developing hypertension. A study examined the number of glomeruli in 10 people age 35 to 59 who died in car accidents who had hypertension and/or left ventricular hypertrophy with age-matched individuals who were normotensive. The hypertensive group had an average of approximately 700,000 glomeruli per kidney, whereas the normotensive group had twice that number. In addition, there was evidence of compensatory glomerular hypertrophy in the hypertensive group. Similar findings were shown in Aboriginal people, a population with a high prevalence of chronic kidney disease. Aborigines had approximately 200,000 fewer glomeruli per kidney than non-Aboriginal people. The Aboriginal population with hypertension had approximately 250,000 fewer glomeruli than those with normal blood pressures. Another population with a high prevalence of hypertension and end-stage renal disease is African Americans. A study from the southeastern United States did not find a difference in glomerular number comparing hypertensive individuals to controls. This same study examined whites and found a reduction in glomerular number in the hypertensive group, but it was only approximately 20% less than controls. These data are consistent with an association between glomerular number and hypertension in some populations but not others. The significance of the reduction in glomerular number in mediating the increase in blood pressure is not clear because the hypertension could be a factor in medicating the reduction in nephron number.

Thrifty Phenotype Hypothesis

Why would a prenatal insult leading to small-for-gestational-age infants result in changes that would predispose to early death from cardiovascular or renal disease? One possible teleologic explanation lies in the thrifty phenotype hypothesis, which states that adaptation at one point in time may lead to changes that are maladaptive at a later time point. This phenomenon occurs because of plasticity, which allows us to adapt to our environment during specific periods of time, usually during prenatal or early neonatal development. These early adaptive changes result in permanent phenotypic changes despite the fact that our genotype does not change. The changes are advantageous for the time surrounding when they are programmed but may become disadvantageous in later life. Thus, if a neonate is born small for gestational age and then becomes an obese adult, the settings established as a neonate or fetus are maladaptive to the adult. The thrifty phenotype hypothesis is depicted in Fig. 9.3 . As an example, if there is a prenatal insult such as fetal malnutrition, one may shift resources from nephron development to brain and cardiovascular development, which are vital at the time. During this point in life there is an adequate number of nephrons for fluid and electrolyte homeostasis, but as we age the paucity of nephrons at birth may lead to progressive chronic kidney disease in later life. The same argument can be made for the pancreas, where the number of islets may be compromised during development if nutrition is inadequate. Indeed, low birth weight infants who become obese as adults are also at substantial risk for the development of type 2 diabetes as adults. The mechanism whereby one can alter the phenotype without a change in the DNA sequence is due to epigenetic changes that affect transcription and translation. During discrete times during development, changes in the environment can cause changes in methylation of CpG islands in the promoter of some genes affecting transcription and there can also be environmental changes in histones and micro-RNAs. These changes can have permanent effects on the phenotype of the organism and in some cases can be passed on to the next generation.

Pathogenesis of the metabolic syndrome according to the thrifty phenotype hypothesis.

(From Hales CN, Barker DJP. The thrifty phenotype hypothesis. Br Med Bull. 2001;60:5-20.)

Animal Models of Prenatal Programming

The previous human epidemiologic studies show an association between small-for-gestational-age infants and the development of hypertension and cardiovascular and chronic kidney disease. However, although the previous data are very compelling, epidemiologic studies are often complicated by potential confounding variables. In addition, although they show an association, they do not prove causality or define mechanisms. Animal studies using rats, mice, and sheep have cemented the association between maternal insults resulting in small-for-gestational-age neonates and cardiovascular and renal disease in mature offspring. Studies have predominantly used prenatal insults that simulate those experienced by humans, which include maternal low calorie or low protein diet, uterine-placental insufficiency, and prenatal exposure to glucocorticoids. In studies using rats, these prenatal insults have been shown to result in small-for-gestational-age neonates that are born at term. In most studies, unless the prenatal insult is very severe, males are affected by programming more severely than females.

Several studies have assessed the effect of a prenatal insult on glomerular number, because a reduction in nephron number could potentially be the harbinger of hypertension and chronic kidney disease. A maternal low protein diet has been shown to result in a decrease in nephron number in adult offspring compared with offspring that were fed a normal protein intake. Similarly, prenatal administration of glucocorticoids produces a reduction in nephron number. In these studies, there is concordance between a reduction in nephron number and the development of hypertension as adults.

Theoretically, a reduction in glomerular number can be the cause for hypertension. If the reduction in glomerular number is severe enough to cause a decrease in glomerular filtration rate, then the impaired glomerular filtration rate could limit sodium excretion, causing hypertension. However, there are a number of issues that preclude the hypertension being mediated by a reduction in nephron number. In almost all of the studies examining the effect of a low protein diet or prenatal administration of dexamethasone on glomerular number, there is only a 20% to 30% reduction in glomeruli, which is far too low to cause impaired sodium excretion. Furthermore, in some studies examining programming, there are rats that are hypertensive but do not have a decrease in glomerular number. Some studies have found a significant reduction in nephron number by perinatal programming, but that the rats and mice did not develop hypertension. Direct measurement of glomerular filtration rate has been found to be normal in several studies at a time when programmed rats are hypertensive. Other studies using different models have also demonstrated that there is a poor correlation between glomerular number and the development of hypertension.

However, the reduction in nephron endowment with programming could be a predisposing factor for chronic kidney disease. Studies looking at the effect of prenatal programming by administration of dexamethasone to rats did not show a decrease in glomerular filtration rate when compared with vehicle-treated controls at 2 months and at 6 to 9 months of age. In rats whose mothers had surgically induced uteroplacental insufficiency, there was an increase in blood pressure at 3 months but no decrease in glomerular filtration rate. Several studies have examined the effect of a maternal low protein diet on glomerular filtration rate. Compared with offspring whose mothers were fed a normal protein diet, offspring whose mothers were fed a low protein diet did not have a reduction in glomerular filtration rate when corrected for body weight at 4 to 5 months of age. There are compensatory mechanisms that maintain a normal glomerular filtration rate with a loss in nephron number comparable with that seen with programming. The reduction in glomerular filtration rate may only manifest when the rats age. Two studies have examined the effect of maternal low protein diet on glomerular filtration rate in rats at 1.5 years of age, and both found an approximately 50% reduction in glomerular filtration rate. There was no reduction in glomerular filtration rate in rats whose mothers were fed a low protein diet compared with controls at 3 months of age. Thus these studies solidify the epidemiologic studies and show a direct effect of programming on the kidney that results in a progressive decrease in glomerular filtration rate.

Because a reduction in glomerular number does not seem to be the predominant factor mediating the hypertension, investigators have looked at other potential factors that could mediate the increase in blood pressure in programmed animals. Several studies have examined the role of the renin-angiotensin system in the prenatal programming of hypertension. The renin-angiotensin system can act to cause vasoconstriction and an increase in tubular transport, which can both affect blood pressure. Many studies have examined the effect of programming on the systemic renin-angiotensin system, and the results have been reviewed elsewhere. To summarize, although an increase in renin-angiotensin-aldosterone activity could explain the hypertension with programming, the results are quite variable, depending on the cause of the prenatal insult resulting in small for gestational age, the age and sex of the offspring studied, and which components of the system were assayed. Dysregulation of the systemic renin-angiotensin system, although clearly present in some studies, does not provide a unifying cause for hypertension with programming.

An increase in salt transport could explain the hypertension with prenatal programming. In a study comparing renal transporter protein abundance in several nephron segments, rats whose mothers were fed a low protein diet had an increase in NKCC2 and NCC protein abundance. These are the apical membrane sodium transporters in the thick ascending limb and distal convoluted tubule, respectively. More direct assessment of transport was performed using isolated perfused tubules, where sodium transport was measured in programmed and control rat nephron segments. Rats whose mothers were administered dexamethasone had an increase in proximal tubule and thick ascending limb sodium transport compared with vehicle-treated controls. A maternal low protein diet also programmed an increase sodium transport in the thick ascending limb and cortical collecting duct compared with offspring of rats whose mothers were fed a normal protein diet.

The kidney is richly innervated, and sodium transport in several nephron segments is regulated by renal sympathetic nerve activity. Renal denervation normalized the blood pressure in programmed rats to control levels resulting from both maternal uteroplacental insufficiency and prenatal administration of dexamethasone, whereas denervation did not affect the blood pressure of control rats. Prenatal administration of dexamethasone increased proximal tubule NHE3 (the apical membrane Na ⁺ /H ⁺ exchanger responsible for the majority of proximal tubule sodium transport), the thick limb NKCC2, and the distal convoluted tubule NCC protein abundance. The abundance of these transporters was normalized to control levels with denervation, whereas denervation had no effect on the expression of the transporters in offspring of mothers who received that vehicle. Subsequent studies measured renal sympathetic nerve activity directly, which demonstrated an increase in renal sympathetic nerve activity with leg muscle contraction. These results were consistent with prenatal programming increasing transport via an increase in sodium transport mediated by an increase in sympathetic nerve activity. Other factors, such as reactive oxygen species, the systemic and intrarenal renin-angiotensin system, and vascular changes, may contribute to programming of hypertension. These have been reviewed elsewhere.